Operation assistance system and method Number:7,386,371 from the United States Patent and Trademark Office (PTO) owispatent A method and system for providing automatic assistance in operating a machine. A current operation performed by an operator of the machine is detected, and data related to an estimated intention of the operator is generated based on the detected operation. A state of the estimated intention of the operator is then determined. The state indicates, for example, the reliability or strength of the estimated intention. The operation of the machine is altered based on the estimated intention and the state of the estimated intention.<H assistance, Operation, Operation, assis, 7386371.html, Patent, pto, 7386371

Title: Operation assistance system and method

Abstract: A method and system for providing automatic assistance in operating a machine. A current operation performed by an operator of the machine is detected, and data related to an estimated intention of the operator is generated based on the detected operation. A state of the estimated intention of the operator is then determined. The state indicates, for example, the reliability or strength of the estimated intention. The operation of the machine is altered based on the estimated intention and the state of the estimated intention.

Patent Number: 7,386,371 Issued on 06/10/2008 to Kuge, et al.

Inventors: Kuge; Nobuyuki (Kanagawa-ken, JP), Yamamura; Tomohiro (Yokohama, JP)
Assignee: Nissan Motor Co., Ltd. (Kanagawa-Ken, JP)

Appl. No.: 11/012,208

Filed: December 16, 2004

Foreign Application Priority Data


Dec 16, 2003 [JP]			2003-417744
Dec 16, 2003 [JP]			2003-417746
Dec 16, 2003 [JP]			P2003-417745

Current U.S. Class: 701/1 ; 180/168; 701/117; 701/28; 701/29; 701/36

Field of Search: 701/1,23,28,41,42,88,93,117,29 180/167,168,169

References Cited [Referenced By]

U.S. Patent Documents


5189621	February 1993	Onari et al.
5546305	August 1996	Kondo
5911771	June 1999	Reichart et al.
6049749	April 2000	Kobayashi
6185492	February 2001	Kagawa et al.
6240357	May 2001	Bastian
2002/0013650	January 2002	Kusafuka et al.
2003/0060936	March 2003	Yamamura et al.
2003/0233902	December 2003	Hijikata
2003/0236608	December 2003	Egami
2004/0172185	September 2004	Yamamura et al.

Foreign Patent Documents


19620929	Nov., 1997	DE
1347214	Feb., 2002	EP
1 197 683	Apr., 2002	EP
1375280	Jun., 2002	EP
1285842	Aug., 2002	EP
1426230	Nov., 2002	EP
1357013	Mar., 2003	EP

Other References

US. Appl. No. 11/012,158, filed Dec. 16, 2004, Kuge et al. cited by other .
U.S. Appl. No. 11/012,165, filed Dec. 16, 2004, Yamamura et al. cited by other.

Primary Examiner: Jeangla; Gertrude Arthur
Attorney, Agent or Firm: McDermott Will & Emery LLP

Claims

What is claimed is:

1. An operation assistance method comprising the steps of: detecting an operation performed by a real operator of a machine, wherein the operation may correspond to multiple possible intentions; generating data related to one or more imaginary operators based on the detected operation of the real operator; and determining an estimated intention of the real operator based on the generated data related to the one or more imaginary operators.

2. The method of claim 1 further comprising the steps of: generating data related to an estimated period of time that the real operator retains the estimated intention; and generating a state of the estimated intention of the real operator based on the estimated period of time, wherein the state of the estimated intention indicates a degree of reliability of the estimated intention of the real operator.

3. The method of claim 2, wherein the step of generating data related to an estimated period of time comprises the steps of: generating data related to an estimated intention of the real operator at a specific point in time and at least one past estimated intention of the real operator before the specific point in time; and generating data related to the estimated period of time based on the data related to the estimated intention of the real operator at the specific point in time and the at least one past estimated intention of the real operator before the specific point in time.

4. The method of claim 1, wherein: the machine is a vehicle; and the real operator is a driver of the vehicle.

5. The method of claim 1, wherein: the detecting step detects an operation of the real operator at each one of different points in time; and the step of generating data related to the one or more imaginary operators comprises the steps of: providing data related to a plurality of imaginary operators, each of the plurality of imaginary operators associated with a sequence of intentions corresponding to the different points in time, wherein each of the sequence of intentions is associated with an operation; for each imaginary operator, calculating a likelihood value based on partial likelihood values of each imaginary operator corresponding to the different points in time, wherein each of the partial likelihood values is respectively associated with each of the sequence of intentions and the respective operation at each one of the different points in time, and is generated based on the respective detected operation of the real operator at each one of the different points in time and the respective operation of the respective one of the plurality of imaginary operators at each one of the different points in time; selecting one of the plurality of imaginary operators based on the likelihood value of each one of the imaginary operators; and generating the estimated intention of the real operator based on a chosen intention of the selected one of the imaginary operators.

6. The method of claim 5, wherein the step of generating data related to one or more imaginary drivers comprises the steps of: determining the point in time corresponding to the chosen intention; for the selected one of the plurality of imaginary operators, determining the point in time corresponding to the last intention in the sequence of intentions that is different from the chosen intention; and determining a period of time that the chosen intention is retained by the selected one of the plurality of imaginary operators, based on the point in time corresponding to the chosen intention and the point in time corresponding to the last intention in the sequence of intentions that is different from the chosen intention; and generating the state of the estimated intention based on the determined period of time.

7. The method of claim 4 further comprising the steps of: generating a state of the estimated intention based on the estimated period of time, wherein the state of the estimated intention indicates a degree of reliability of the estimated intention; calculating a risk potential associated with the vehicle; calculating a reaction force based on the calculated risk potential, the estimated intention and the state of the estimated intention; and applying the reaction force to a vehicle control device of the vehicle.

8. The method of claim 7, wherein the vehicle control device is an acceleration pedal or a steering wheel.

9. The method of claim 7, wherein the step of calculating a reaction force comprises the steps of: modifying the risk potential based on the state of the estimated intention; and calculating the reaction force based on the modified risk potential and the estimated intention.

10. The method of claim 7, wherein the step of calculating a reaction force comprises the steps of: calculating the reaction force based on the calculated risk potential and the estimated intention; and modifying the calculated reaction force based on the state of the estimated intention.

11. The method of claim 1 further comprising the steps of: generating a state of the estimated intention based on the estimated period of time, wherein the state of the estimated intention indicates a degree of reliability of the estimated intention; and modifying the operation of the machine based on the state of the estimated intention of the real operator.

12. An operation assistance system comprising: a first device configured to detect an operation performed by a real operator of a machine, wherein the operation may correspond to multiple possible intentions; a second device configured to generate data related to one or more imaginary operators based on the detected operation of the real operator; and a third device configured to determine an estimated intention of the real operator based on the generated data related to the one or more imaginary operators.

13. The method of claim 12, wherein the third device is configured to generate data related to an estimated period of time that the real operator retains the estimated intention, and to generate a state of the estimated intention based on the estimated period of time.

14. The method of claim 13, wherein the third device is configured to generate data related to an estimated intention of the real operator at a specific point in time and at least one past estimated intention of the real operator before the specific point in time, and to generate data related to the estimated period of time based on the data related to the estimated intention of the real operator at the specific point in time and the at least one past estimated intention of the real operator before the specific point in time.

15. The system of claim 12, wherein: the machine is a vehicle; and the real operator is a driver of the vehicle.

16. The system of claim 12, wherein: the first device is configured to detect an operation of the real operator at each one of different points in time; and the second device is configured to perform the steps of: providing data related to a plurality of imaginary operators, each of the plurality of imaginary operators associated with a sequence of intentions corresponding to the different points in time, wherein each of the sequence of intentions is associated with an operation; for each imaginary operator, calculating a likelihood value based on partial likelihood values of each imaginary operator corresponding to the different points in time, wherein each of the partial likelihood values is respectively associated with each of the sequence of intentions and the respective operation at each one of the different points in time, and is generated based on the respective detected operation of the real operator at each one of the different points in time and the respective operation of the respective one of the plurality of imaginary operators at each one of the different points in time; selecting one of the plurality of imaginary operators based on the likelihood value of each one of the imaginary operators; and generating the estimated intention of the real operator based on a chosen intention of the selected one of the imaginary operators.

17. The system of claim 16, wherein second device is configured to perform the steps of: determining the point in time corresponding to the chosen intention; for the selected one of the plurality of imaginary operators, determining the point in time corresponding to the last intention in the sequence of intentions that is different from the chosen intention; and determining a period of time that the chosen intention is retained by the selected one of the plurality of imaginary operators, based on the point in time corresponding to the chosen intention and the point in time corresponding to the last intention in the sequence of intentions that is different from the chosen intention; generating a state of the estimated intention based on the determined period of time.

18. The system of claim 15 further comprising: a device configured to determine a state of the estimated intention based on the data related to one or more imaginary operators; a risk calculation device configured to calculate a risk potential associated with the vehicle; a reaction force calculation device configured to calculate a reaction force based on the calculated risk potential, the estimated intention and the state of the estimated intention; and applying the reaction force to a vehicle control device of the vehicle.

19. The system of claim 18, wherein the vehicle control device is an acceleration pedal or a steering wheel.

20. An operation assistance system comprising: means for detecting an operation performed by a real operator of a; machine, wherein the operation may correspond to multiple possible intentions; means for generating data related to one or more imaginary operators based on the detected operation; and means for determining an estimated intention of the real operator based on the generated data related to one or more imaginary operators.

21. A machine-readable medium bearing instructions for assisting operation of a machine, the instructions, upon execution by a data processing system, causing the data processing system to perform the steps of: receiving data related to an operation performed by a real operator of the machine, wherein the operation may correspond to multiple possible intentions; generating data related to one or more imaginary operators based on the detected operation; and determining an estimated intention of the real operator based on the generated data.

22. A vehicle comprising: a first device configured to detect an operation performed by a driver of the vehicle, wherein the operation may correspond to multiple possible intentions; a second device configured to generate data related to one or more imaginary drivers based on the detected operation, wherein each of the imaginary drivers is associated with a pre-assigned operation; and a third device configured to determine an estimated intention of the driver based on the generated data related to the one or more imaginary drivers.

23. The vehicle of claim 22 further comprising a control device to determine a state of the estimated intention of the driver and control the operation of the vehicle based on the identified state of the estimated intention of the driver.

24. An operation assistance method comprising the steps of: detecting an operation performed by an operator of a machine, wherein the operation may correspond to multiple possible intentions; estimating an intention of the operator based on the detected operation; generating data related to the estimated intention of the operator based on the detected operation; and identifying a state of the estimated intention of the operator based on the generated data.

25. An operation assistance system comprising: a first device configured to detect an operation performed by an operator of a machine, wherein the operation may correspond to multiple possible intentions; a second device configured to estimate an intention of the operator based on the detected operation; a third device configured to generate data related to the estimated intention of the operator based on the detected operation; and a fourth device configured to identify a state of the estimated intention of the operator based on the generated data.

26. An operation assistance system comprising: means for detecting an operation performed by an operator of a machine, wherein the operation may correspond to multiple possible intentions; means for estimating an intention of the operator based on the detected operation; means for generating data related to the estimated intention of the operator based on the detected operation; and means for identifying a state of the estimated intention of the operator based on the generated data.

27. A computer readable storage medium having stored therein data representing instructions executable by a computer to assist operation of a machine, comprising: instructions to detect an operation performed by an operator of a machine, wherein the operation may correspond to multiple possible intentions; instructions to estimate an intention of the operator based on the detected operation; instructions to generate data related to the estimated intention of the operator based on the detected operation; and instructions to identify a state of the estimated intention of the operator based on the generated data.

28. A vehicle comprising: a first device configured to detect an operation performed by an operator of the vehicle, wherein the operation may correspond to multiple possible intentions; a second device configured to estimate an intention of the operator based on the detected operation; a third device configured to generate data related to the estimated intention of the operator based on the detected operation; and a fourth device configured to identify a state of the estimated intention of the operator based on the generated data.

Description

RELATED APPLICATIONS

The present application claims the benefit of priority from Japanese patent application No. 2003-417744, filed Dec. 16, 2003; Japanese patent application No. 2003-417745, filed Dec. 16, 2003; and Japanese patent application No. 2003-417746, filed Dec. 16, 2003, all of which are incorporated herein by reference in their entireties.

FIELD OF THE DISCLOSURE

The present disclosure relates to system and method for providing assistance based on an operator's intention, and more particularly, to driving assistance system and method for determining a state of an estimated intention of an operator and providing operation assistance accordingly.

BACKGROUND OF THE DISCLOSURE

A number of methods and systems have been proposed for providing assistance in operating a device, system or machine, such as a vehicle. For example, several driving assistance systems were disclosed in U.S. Published Patent Application Nos. 20030060936 A1, published Mar. 27, 2003, and 20040172185 A1, published Sep. 2, 2004. In order to enhance performance, some driving assistance systems may require estimation of a driver's intention in driving a vehicle. A system for estimating a driver's intention may collect estimates of the driver's intention using movement of the driver's eyeballs. For example, directions to which the driver's eyeballs turn are projected onto a plane divided into a number of regions, for calculating a distribution of projected directions over the divided regions to estimate the driver's intention. However, such type of systems lacks accuracy because the driver's eyeballs move all the time and do not always relate to a "driving" intention of the driver.

Therefore, there is a need for reliable intention estimation systems that can estimate an operator's intention with satisfactory accuracy. There is also a need for determining how reliable or how strong an estimated intention is, such that operation assistance can be provided accordingly.

SUMMARY OF THE DISCLOSURE

This disclosure presents system, control process and method that provide effective estimation of an operator's intention in operating a device, system or machine, and indicate a state of the estimation of the operator's intentions, which may indicate the strength of the estimated intention. Operation assistance may be provided based on the state of the estimated intention of the operator. The advantages, operations and detailed structures of the disclosed methods and systems will be appreciated and understood from the descriptions provided herein.

An exemplary system and method according to this disclosure detect an operation performed by an operator of a machine, and generate data related to one or more imaginary operators based on the detected operation. An estimated intention of the operator is then determined based on the generated data related to the one or more imaginary drivers. In one aspect, a state of the estimated intention may determined. The state may indicate the reliability or strength of the estimated intention. For instance, the state may represent a period of time that the operator has retained the estimated intention. The longer the operator retains the intention, the more determined the operator is to perform an action or operation according to the estimated intention.

In one embodiment, an estimated intention of the operator at a specific point in time, and at least one past estimated intention of the operator before the specific point in time are determined. An estimated period of time that the operator retains an estimated intention is determined based on the estimated intention of the operator at the specific point in time and the at least one past estimated intention of the operator before the specific point in time. In one aspect, the machine is a vehicle and the operator is a driver of the vehicle.

According to another embodiment, the estimated intention of the operator is determined by comparing an operation of the operator with reference data, such as data related to a plurality of imaginary operators. Each of the plurality of imaginary operators is associated with a sequence of intentions corresponding to different points in time, wherein each of the sequence of intentions is associated with an operation. For each imaginary operator, a likelihood value is calculated based on partial likelihood values of each imaginary operator corresponding to the different points in time, wherein each of the partial likelihood values is respectively associated with each respective intention and the respective operation at each one of the different points in time, and is generated based on the respective detected operation of the real operator at each one of the different points in time and the respective operation of the respective one of the plurality of imaginary operators at each one of the different points in time. One of the plurality of imaginary operators is selected to approximate to the real operator based on the likelihood value of each one of the imaginary operators. A chosen intention of the selected imaginary operator is used to approximate the intention of the real operator.

In order to determine a state of the estimated intention, such as how long the real operator have retained the estimated intention, the point in time corresponding to the chosen intention and the most recent point in time that the selected imaginary operator retaining an intention different from the chosen intention are determined. The period between the two points in time is used to indicate the state of the estimated intention. The longer this period of time is, the stronger the estimated intention is.

According to still another embodiment, the operation of the machine is adjusted based on the state of the estimated intention. For instance, in a vehicle, a risk potential associated with the vehicle and the driver's intention are constantly monitored and determined. A reaction force to be applied or being applied to a vehicle control device is determined based on the calculated risk potential, the estimated intention and the state of the estimated intention. The vehicle control device may be any device that a driver manipulates to control the operation of the vehicle, such as an acceleration pedal or a steering wheel, or any device that can provide a haptic feedback to the operator. In one embodiment, the risk potential is modified based on the state of the estimated intention, and the reaction force is calculated based on the modified risk potential and the estimated intention. In another embodiment, the reaction force is calculated based on the calculated risk potential and the estimated intention, and is then modified based on the state of the estimated intention.

Additional advantages of the present disclosure will become readily apparent to those skilled in this art from the following detailed description, wherein only the illustrative embodiments are shown and described, simply by way of illustration of the best mode contemplated. As will be realized, the disclosure is capable of other and different embodiments, and its several details are capable of modifications in various obvious respects, all without departing from the disclosure. Accordingly, the drawings and description are to be regarded as illustrative in nature, and not as restrictive.

BRIEF DESCRIPTION OF THE DRAWINGS

The present disclosure is illustrated by way of example, and not by way of limitation, in the figures of the accompanying drawing and in which like reference numerals refer to similar elements and in which:

FIG. 1 is a block diagram illustrating an exemplary implementation of a driving assistance system according to the present disclosure.

FIG. 2 is a flow chart illustrating the operation of the driving assistance system illustrated in FIG. 1.

FIG. 3 illustrates calculations of an operation amount for an imaginary driver.

FIG. 4 is an exemplary illustration of generating a series of lane-keeping intentions retained by a parent imaginary driver and derivative lane-change intentions retained by additional imaginary drivers.

FIG. 5(a) shows a rule for applying to the generation of data related to imaginary drivers as illustrated in FIG. 4.

FIG. 5(b) is an illustration of another rule for applying to the generation of data related to imaginary drivers as illustrated in FIG. 4.

FIG. 6 is an exemplary illustration of the a first series of intentions (SERIES L1) corresponding to a first imaginary driver having a lane-change intention to the right (LCR), a second series of intentions (SERIES L2) corresponding to a second imaginary driver having a lane-change intention to the right (LCR), and a third series of intentions (SERIES L3) corresponding to a third imaginary driver having a lane-change intention to the right (LCR).

FIG. 7 is a block diagram illustrating another exemplary implementation of a driving assistance system according to the present disclosure.

FIG. 8 is a perspective view of a vehicle in the form of a vehicle incorporating the driving assistance system described herein.

FIG. 9 is an illustration of an exemplary vehicle control device in the form of an accelerator pedal.

FIG. 10 is a flow chart illustrating operation of the driving assistance system illustrated in FIG. 7.

FIG. 11 illustrates of the relationship between a reaction force increment and different values of risk potential (RP).

FIG. 12 illustrates characteristics of a time constant Tsf_etlc relative to different values of time etlc.

FIG. 13(a) shows a traffic scene in which a vehicle changes lanes to pass the preceding vehicle.

FIG. 13(b) illustrates a corrected accelerator pedal reaction force instruction value FAc in response to the estimated driver's lane-change intention.

FIG. 14 is a flow chart showing the steps of the exemplary implementation illustrated in FIGS. 7 to 13(b).

FIG. 15 illustrates characteristics of coefficient Ksf relative to different values of elapsed time etlc greater than a predetermined value etlc0.

FIG. 16 is a block diagram of another exemplary implementation of a driving assistance system according to the present disclosure.

FIG. 17 is a flow chart showing the operation of the driving assistance system illustrated in FIG. 16.

FIG. 18 illustrates different values of time constant Tsp_etlc relative to different values of elapsed time etlc.

FIG. 19(a) illustrates a traffic scene in which a vehicle changes lanes to pass the preceding vehicle.

FIG. 19(b) illustrates an accelerator pedal reaction force instruction value FA in response to the estimated driver's lane-change intention.

FIG. 20 is a flow chart of the exemplary implementation illustrated in FIGS. 16 to 19(b).

FIG. 21 illustrates characteristics of coefficient Ksp relative to different values of elapsed time etlc greater than a predetermined value etlc0.

ILLUSTRATIVE EMBODIMENTS OF THE DISCLOSURE

In the following description, for the purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the present disclosure. It will be apparent, however, to one skilled in the art that the present method and system may be practiced without these specific details. In other instances, well-known structures and devices are shown in block diagram form in order to avoid unnecessarily obscuring the present disclosure. For illustration purpose, the following examples describe the operation of an exemplary tester used for evaluating a circuit of an automotive vehicle. It is understood that the use of tester is not limited to vehicle circuits. The tester also can be used in other types of electrical circuits.

Referring to FIG. 1, an exemplary system 1 according to this disclosure includes a vehicle's environment detector 10, a vehicle's status detector 20, a real driver's operation detector 30, an imaginary driver's intention generating section 40, an imaginary driver's operation calculator 50, a likelihood P(j)ids calculator 60, a driver's intention estimator 70, and a section 80 to determine a state of an estimated driver's intention, for instance, by determining how long the estimated driver's intention has been retained. The vehicle's environment detector 10 detects a state of environment within a field around the vehicle. The vehicle's status detector 20 detects an operation status of the vehicle. The real driver's operation detector 30 detects an operation amount of a real driver in driving the vehicle.

The driver's intention estimating system 1 has access to reference data, such as data related to a plurality of imaginary drivers. Each of the imaginary drivers is designed to perform an operation of the vehicle according to an associated intention. Examples of the intention may include a lane-keeping intention (LK), a lane-change intention to the right (LCR), and a lane-change intention to the left (LCL).

As will be described in detail in connection with FIGS. 4, 5(a), and 5(b), the imaginary driver's intention generating section 40 continuously generates a lane-keeping intention (LK) at every point in time to form a series of intention for a parent imaginary driver. Furthermore, the imaginary driver's intention generating section 40 generates data related to at least one additional imaginary driver based on the intention of the parent imaginary driver. In one embodiment, the imaginary driver's intention generating section 40 generates data related to two additional imaginary drivers, each has one of two derivative lane-change intentions (LCR) and (LCL) based on a lane-keeping intention (LK) of the parent imaginary driver at an immediately preceding pint in time. In another embodiment, the imaginary driver's intention generating section 80 applies special rules in generating series of intentions for the additional imaginary drivers.

The imaginary driver's intention generating section 40 allows a parent imaginary driver to retain a lane-keeping intention (LK) at every point in time. Further, at every point in time with the parent imaginary driver having a lane-keeping intention (LK), the imaginary driver's intention generating section 40 generates data related to two additional imaginary drivers having lane-change intentions to the right (LCR) and to the left (LCL), respectively, for the next point in time. In one embodiment, an additional imaginary driver generated at a specific point of time assumes at least some of the intentions for all points of time preceding the specific point in time, from the parent imaginary driver.

Furthermore, the imaginary driver's intention generating section 40 determines whether or not an imaginary driver retaining one of the derivative lane-change intentions to exist at the next point in time should be allowed to continue to exist, by applying one or more rules. For instance, an exemplary rule allows the parent imaginary driver to retain a lane-keeping intention (LK) at every point in time, and generates data related to two additional imaginary drivers having lane-change intentions (LCR) and (LCL), respectively, at the next point in time. According to another exemplary rule, an imaginary driver is allowed to retain a lane-change intention to the right (LCR) at the next point in time if it is determined that the real driver continues to retain a lane-changing intention at the present point in time. On the other hand, if it is determined that at a specific point in time, the real driver no longer wants to change lanes or has just changed lanes, an imaginary driver is not allowed to retain a lane-change intention to the right (LCR) at the next point in time. This is equally applicable to a lane-change intention to the left (LCL). Accordingly, an imaginary driver having a lane-change intention to the left (LCL) at a specific point in time is allowed to retain a lane-change intention to the left (LCL) at the next point in time upon determination that a lane change continues, but the imaginary driver is not allowed to continue to retain a lane-change intention to the left (LCL) at a specific point in time upon failure to determine that the lane change continues. Therefore, an imaginary driver that has one of the derivative lane-change intentions (LCR) and (LCL), is allowed to retain the derivative lane-change intention at the next point in time upon determination that a lane change continues.

At any point in time, each of the imaginary drivers has an associated operation corresponding to an intention retained by that imaginary driver. The process for determining an operation associated with each intention is described below. The vehicle's environment detector 10 provides information on a state of environment around the vehicle to the imaginary driver's operation calculator 50. Examples of such information include a lateral distance y of the vehicle from a centerline within a lane, and a yaw angle .psi. of the vehicle with respect a line parallel to the centerline. The vehicle's status detector 20 provides information on a status of the vehicle to the imaginary driver's operation calculator 50. Examples of such information include a vehicle speed of the vehicle and a steering angle of the vehicle.

The imaginary driver's operation calculator 50 calculates operation amounts Oid of the imaginary drivers in a manner that will be described in detail in connection with FIG. 3. In order to reduce the computation load, certain rules are applied to determine whether an existing additional imaginary driver retaining one of the derivative lane-change intentions should be allowed to exist at the next point. In other words, if a predetermined condition established by the rules is not met by an additional imaginary driver at a specific point in time, that additional imaginary driver is terminated or eliminated. Since it is not necessary to calculate operation amounts Oid of the eliminated imaginary drivers, the computation load is reduced.

The imaginary driver's operation calculator 50 provides the calculated operation amounts Oid of the imaginary drivers to the likelihood value P(j)ids calculator 60. For comparison with each of the calculated operation amounts Oid of the imaginary drivers, the real driver's operation detector 30 provides a detected operation amount Ord to the likelihood value P(j)ids calculator 60. An example of the operation amount to be detected is a steering angle of the vehicle.

The likelihood value P(j)ids calculator 60 calculates a likelihood value Pid(j)(t) of an imaginary driver, based on the associated operation amounts Oid and the detected operation amount Ord. The calculated likelihood values Pid(j)(t) are stored in data storage devices, such as memory or hard disk. For each imaginary driver, the data storage device stores the most recent likelihood value Pid(j)(t) after shifting the previously calculated likelihood value. The stored likelihood values may be represented in the form of Pid(j)(t), Pid(j)(t-1), . . . , Pid(j)(t-m+1), which correspond to likelihood values calculated at different points in time ranging from the present point in time (t) back to (t-m+1). The m (a natural number) points in time are arranged at regular intervals and define a predetermined period of time.

The likelihood value P(j)ids calculator 60 calculates a collective likelihood value P(j)ids for each imaginary driver j based on likelihood values Pid(j)(t), Pid(j)(t-1), . . . , Pid(j)(t-m+1)and provides the calculated series-likelihood values P(j)ids for processing at the driver's intention estimator 70.

In one embodiment, the driver's intention estimator 70 selects one of the imaginary drivers to approximate behaviors of the real driver based on the calculated collective likelihood values P(j)ids. An intention of the selected imaginary driver, such as the most recent intention retained by the selected imaginary driver, is set as an estimated driver's intention .lamda.rd of the real driver.

In order to determine a state of an estimated driver's intention .lamda.rd, section 80 is provided. Detailed operation of section 80 will be described later.

With continuing reference to FIG. 1 the vehicle's environment detector 10 includes a front camera that covers a field of front view and a yaw angle sensor. The front camera acquires images of road conditions, within the field of front view. The vehicle's environment detector 10 detects a lateral distance y of the vehicle from a centerline within a lane, and a yaw angle .psi. of the vehicle with respect a line parallel to the centerline. The vehicle's environment detector 10 is equipped with an image processor that processes the acquired image. The vehicle's status detector 20 includes a vehicle speed sensor for detecting a speed of the vehicle. The real driver's operation detector 30 includes a sensor to detect an operation performed by the driver. Detector 30 may be a steering angle sensor that detects a steering angle of the vehicle. Other types of sensors can be used, such as acceleration sensor or brake sensor.

As shown in FIG. 1, the exemplary system 10 includes imaginary driver's intention generating section 40, imaginary driver's operation calculator 50, likelihood value P(j)ids calculator 60, driver's intention estimator 70, and section 80. Some or all of these elements are implemented using one or more microcomputers or microcontrollers, such as a central processor unit (CPU), executing microcode, software programs and/or instructions. The microcode and/or software reside in volatile and/or non-volatile data storage devices and/or machine-readable data storage medium such as read only memory (ROM) devices, random access memory (RAM) devices, SRAM, PROM, EPROM, CD-ROM, disks, carrier waves, etc.

As described before, the imaginary driver's intention generating section 40 continuously generates data related to imaginary drivers. Each of the imaginary drivers retains a series of intentions over a period of time. The number of the imaginary drivers and types of intentions retained by the imaginary drivers are dynamic, and may change over time.

As described before, the imaginary driver's operation calculator 50 calculates operation amounts Oid of the imaginary drivers associated with different intentions that are determined by the imaginary driver's intention generating section 40.

Referring to FIGS. 2 and 3, the operation of the exemplary system 1 is explained. The flow chart in FIG. 2 illustrates the operation of a driver's intention estimation processing program. Execution of this program is repeated at a regular interval of .DELTA.T, for example, .DELTA.T=50 milliseconds.

At step S101, the microcomputer reads in data related to a lateral position y of the vehicle within a lane (or track) and a yaw angle .psi. of the vehicle. As shown in FIG. 3, the lateral position y is a distance of a center O of the vehicle from the centerline of the lane, and the yaw angle .psi. is an angle through which the vehicle is turned relative to a specific reference, such as a line parallel to the centerline of the lane.

At step S102, the microcomputer calculates an operation Oid of each of a plurality of imaginary drivers. In this example, the plurality of imaginary drivers are variable in number and includes an imaginary driver A designed to behave as directed by the latest intention of a series of a lane-keeping intention (LK). The remaining of the plurality of imaginary drivers consists of at least one imaginary driver B designed to behave as directed by a lane-change intention to the right (LCR), and at least one imaginary driver C designed to behave as directed by a lane-change intention to the left (LCL). The microcomputer calculates an operation amount Oid, by which each of these three imaginary drivers A, B and C would operate a vehicle control device, such as a steering wheel or an acceleration pedal, in driving the vehicle as directed by the intention. In the exemplary implementation, the vehicle control device is a steering system of the vehicle. In this case, the operation amount Oid is a steering angle .theta.id. The microcomputer calculates a steering angle .theta.id, which each of the three imaginary drivers A, B and C would perform to manipulate a steering wheel in driving the vehicle as directed by the intention. The following descriptions describe how a steering angle .theta.id associated with an imaginary driver is calculated.

(1) Imaginary Driver A Having a Lane-keeping Intention (LK):

Steering angle .theta.id.lk represents an angle that an imaginary driver A having a lane-keeping intention (LK) would manipulate a steering wheel in driving the vehicle. The microcomputer sets at least one reference point LK(i) in front on a longitudinal centerline of the vehicle at a distance px(i) from the center O of the vehicle, and calculates a lateral position p_lk(px(i)) of the reference point LK(i) from a centerline of a lane. At least one reference point LK(i) includes any desired number of reference points LK(i). In this example, as shown in FIG. 3, two reference points LK(1) and LK(2) are set on the longitudinal centerline of the vehicle at different distances px(1) and px(2) from the center O of the vehicle, wherein the distance px(1)=10 m and the distance px(2)=30 m. The distance px(i) may vary with different vehicle speeds.

A lateral distance lat_pos(px(i)) of the reference point LK(i) from the centerline of the lane is dependent on, and is thus determined by, the yaw angle .psi.and the distance px(i), which may be determined, for example, by processing the acquired image from the front camera. Thus, the lateral position p_lk(px(i) of the reference point LK(i) may be expressed as: p.sub.--lk(px(i)=lat_pos(px(i)) i={1, . . . ,n} (Eq. 1) The number n is equal to 2 (n=2) in the example shown in FIG. 3.

Using the lateral position p_lk(px(i)), the steering angle .theta.id.sub.13lk may be expressed as: .theta.id.sub.--lk=.SIGMA.{a(i)p.sub.--lk(px(i))} (Eq. 2)

where: a(i) is an appropriately determined coefficient weighting the lateral position p_lk(px(i)), and is determined based on characteristics of vehicles, such as the gear ratio of a vehicle implementing the system disclosed herein.

(2) Imaginary Driver B Having a Lane-change Intention to the Right (LCR):

Steering angle .theta.id_lcr represents an angle that an imaginary driver B having a lane-change intention to the right (LCR) would manipulate a steering wheel in driving the vehicle as directed by the lane-change intention to the right (LCR). The microcomputer sets at least one reference point LCR(i) which may include any desired number of reference points LCR(i). In the example shown in FIG. 3, two reference points LCR(1) and LCR(2) are set.

A lateral position p_lcr(px(i)) of the reference point LCR(i) may be given as a sum of lat_pos(px(i)) and a predetermined offset lc_offset_lcr. Lateral position p_lcr(px(i)) can be expressed as: p.sub.--lcr(px(i)=lat_pos(px(i))+lc_offset.sub.--lcr i={1, . . . ,n} (Eq. 3) The number n is equal to 2 (n=2) in the example shown in FIG. 3. The predetermined offset lc_offset_lcr is an appropriately determined value for giving the lateral position p_lcr(px(i)) of the reference point LCR(i). In this example, the offset lc_offset_lcr is equal to -1.75 (lc_offset_lcr=-1.75).

Using the lateral position p_lcr(px(i)), the steering angle .theta.id_lcr may be expressed as: .theta.id.sub.--lcr=.SIGMA.{a(i)p.sub.--lcr(px(i))} (Eq. 4)

where: a(i) is an appropriately determined coefficient weighting the lateral position p_lcr(px(i)), and is determined based on characteristics of vehicles, such as the gear ratio of a vehicle implementing the system disclosed herein.

(3) Imaginary Driver C Having a Lane-change Intention to the Left (LCL):

Steering angle .theta.id_lcl represents an angle by which an imaginary driver C having a lane-change intention to the left (LCR) would manipulate a steering wheel in driving the vehicle as directed by the lane-change intention to the left (LCR). The microcomputer sets at least one reference point LCL(i) which may include any desired number of reference points LCL(i). In the example shown in FIG. 3, two reference points LCL(1) and LCL(2) are set.

A lateral position p_lcl(px(i)) of the reference point LCL(i) may be given by a sum of lat_pos(px(i)) and a predetermined offset lc_offset_lcl, and thus expressed as: p.sub.--lcl(px(i))=lat_pos(px(i))+lc_offset.sub.--lcl i={1, . . . ,n} (Eq. 5) The number n is equal to 2 (n=2) in the example shown in FIG. 3. The predetermined offset lc_offset_lcl is an appropriately determined value for giving the lateral position p_lcl(px(i)) of the reference point LCL(i). In this example, the offset lc_offset_lcl is equal to 1.75 (lc_offset_lcr=1.75).

Using the lateral position p_lcl(px(i)), the steering angle .theta.id_lcl may be expressed as: .theta.id.sub.--lcl=.SIGMA.{a(i)p.sub.--lcl(px(i))} (Eq. 6)

where: a(i) is an appropriately determined coefficient weighting the lateral position p_lcl(px(i)), and is determined based on characteristics of vehicles, such as the gear ratio of a vehicle implementing the system disclosed herein.

After calculating the operation amount Oid of each of the imaginary drivers A, B and C at step S102, the logic goes to step S103. At step S103, the microcomputer receives, as an input, an operation amount Ord of a real driver by, in this exemplary implementation, reading in a steering angle .theta.rd detected by the real driver's operation detector 30.

At the next step S104, the microcomputer forms a series of intentions for each of the plurality of imaginary drivers. The types of intentions and the number of the imaginary drivers may change over time. The microcomputer has memory portions for storing the intentions. Each of the memory portions is designed to store m, in number, intentions over a period of time ranging from time (t) back to time (t-m+1). Except for a special memory portion, the microcomputer resets any one of the remaining memory portions upon determination that the memory portion has contained m, in number, intentions of the same kind.

FIG. 4 illustrates data related to a plurality of imaginary drivers generated by the microcomputer. Each imaginary driver retains a series of intentions over, m, in number, points in time ranging from time (t) back to time (t-m+1). Referring to FIG. 4, the microcomputer continuously generates a lane-keeping intention (LK) at every point in time. The lane-keeping intentions form a series of intentions for a parent imaginary driver.

Furthermore, the microcomputer generates data related to at least one additional imaginary driver based on the intention of the parent imaginary driver. In the example shown in FIG. 5(a), the microcomputer generates data related to two additional imaginary drivers, each has one of two derivative lane-change intentions (LCR) and (LCL) based on a lane-keeping intention (LK) of the parent imaginary driver at an immediately preceding pint in time. In addition, the two additional imaginary drivers generated at a specific point of time assumes at least some of the intentions for all points in time preceding the specific point in time, from the parent imaginary driver.

Referring also to FIG. 5(a), the microcomputer applies certain rules in generating series of intentions for existing additional imaginary drivers. For instance, the microcomputer determines whether an imaginary driver retaining one of the derivative lane-change intentions may continue to exist at the next point in time, by applying one or more rules. An exemplary rule allows the parent imaginary driver to retain a lane-keeping intention (LK) at every point in time, and generates data related to two additional imaginary drivers having lane-change intentions (LCR) and (LCL), respectively, at the next point in time. According to another exemplary rule, an imaginary driver is allowed to retain a lane-change intention to the right (LCR) at the next point in time if it is determined that the real driver continues to retain a lane-changing intention at the present point in time. On the other hand, if it is determined that at a specific point in time, the real driver no longer wants to change lanes or has just changed lanes, an imaginary driver is not allowed to retain a lane-change intention to the right (LCR) at the next point in time. This is equally applicable to a lane-change intention to the left (LCL). Accordingly, an imaginary driver having a lane-change intention to the left (LCL) at a specific point in time is allowed to retain a lane-change intention to the left (LCL) at the next point in time upon determination that a lane change continues, but the imaginary driver is not allowed to continue to retain a lane-change intention to the left (LCL) at a specific point in time upon failure to determine that the lane change continues. Therefore, an imaginary driver that has one of the derivative lane-change intentions (LCR) and (LCL), is allowed to retain the derivative lane-change intention at the next point in time upon determination that a lane change continues.

As described above, a special memory portion is provided for storing intentions of the parent imaginary driver. The intentions include m, in number, lane-keeping intentions (LK), over a period of time ranging from time (t) back to time (t-m+1). Each of the remaining memory portions is provided for storing intentions for one of the additional imaginary drivers. The intentions include lane-change intention (LCR) or (LCL) over a period of time ranging from time (t) back to time (t-m+1). It is now apparent that, except for the special memory portion provided for the parent imaginary driver, the microcomputer resets any memory portions for the additional imaginary drivers upon determination that the memory portion has contained m, in number, lane-change intentions.

Referring to FIG. 4, "SERIES L1" corresponds to a series of intentions of an additional imaginary driver that is generated at time t, and includes a lane-change intention to the right (LCR) at time t. "SERIES L2" includes two lane-change intentions to the right and represents intentions of another additional imaginary driver generated earlier. "SERIES L3" including (m-3), in number, lane-change intentions to the right (LCR) and represents intentions of still another imaginary driver that is generated earlier than "SERIES L1" and "SERIES L2.".

The imaginary driver corresponding to "SERIES L1" retains a lane-keeping intention (LK) at every point in time from (t-m+1) to (t-1), and has a lane-change intention to the right (LCR) at the present point in time (t). The imaginary driver corresponding to "SERIES L2" retains a lane-keeping intention (LK) from (t-m+1) to (t-2), and has a lane-change intention to the right (LCR) at both (t-1) and (t). The imaginary driver corresponding to "SERIES L3" retains a lane-keeping intention (LK) from (t-m+1) to (t-m+2), and a lane-change intention to the right (LCR) at (t-m+3) through (t).

FIGS. 5(a) and 5(b) show rules that the imaginary driver's intention generating section 40 (as shown in FIG. 1) utilizes in determining an intention for each existing driver at each point in time. As mentioned before, the microcomputer allows a parent imaginary driver having a lane-keeping intention (LK) at every point in time. As shown in FIG. 5(a), the microcomputer generates data related to two additional imaginary drivers, each has one of two derivative lane-change intentions (LCR) and (LCL) based on a lane-keeping intention (LK) of the parent imaginary driver at an immediately preceding pint in time.

At each point in time, the microcomputer determines whether or not the vehicle's environment allows a lane-change intention to continue to exist at the next point in time by applying certain rules. FIG. 5(b) shows an exemplary rule used by the microcomputer. As shown in FIG. 5(b), an imaginary driver is allowed to retain a lane-change intention to the right (LCR) at the next point in time, if it is determined that the real driver continues to retain a lane-changing intention at the present point in time. On the other hand, if it is determined that at a specific point in time, the real driver no longer wants to change lanes or has just changed lanes, an imaginary driver is not allowed to retain a lane-change intention to the right (LCR) at the next point in time. This is equally applicable to a lane-change intention to the left (LCL). Accordingly, an imaginary driver having a lane-change intention to the left (LCL) at a specific point in time is allowed to retain a lane-change intention to the left (LCL) at the next point in time upon determination that a lane change continues, but the imaginary driver is not allowed to continue to retain a lane-change intention to the left (LCL) at a specific point in time upon failure to determine that the lane change continues. Therefore, an imaginary driver that has one of the derivative lane-change intentions (LCR) and (LCL), is allowed to retain the derivative lane-change intention at the next point in time upon determination that a lane change continues.

In the exemplary implementation, on one hand, the microcomputer determines that the lane-change intention may continue to exist if the vehicle continues to stay in the same lane. On the other hand, the microcomputer determines that the lane-change intention has been realized if the vehicle has changed to a different lane. In other words, the microcomputer fails to determine that the lane-change intention continues. Thus, lane-change intentions (LCR) and (LCL) at a specific point in time are allowed to continue to exist at the next point in time upon determination that the vehicle continues to stay in the same lane. In contrast, lane-change intentions (LCR) and (LCL) are not allowed to continue to exist at the next point in time upon determination that the vehicle has changed to a different lane. As will be understood from the descriptions below, all imaginary drivers (except for the parent imaginary driver) that have at least one derivative lane-change intention (LCR) or (LCL)are terminated and reset upon determination that the vehicle has changed to a different lane.

At step S105, using the calculated operation amount Oid of each imaginary driver (calculated at step S102) and the detected operation amount Ord of the real driver (detected at step S103), the microcomputer calculates a likelihood value Pid indicating how the calculated operation amount Oid of each imaginary driver approximates the detected operation amount Ord of the real driver. For illustration purpose, the likelihood value Pid is used to represent a likelihood value Pid_lk of an imaginary driver having a lane-keeping intention (LK), a likelihood value Pid_lcr of an imaginary driver having a lane-change to the right (LCR), or a likelihood value Pid_lcl of an imaginary driver having a lane-change intention to the left (LCL). In the exemplary implementation, the calculated operation amount Oid of each imaginary driver is expressed by any one of the calculated steering angles .theta.id_lk, .theta.id_lcr, and .theta.id_lcl. For illustration purpose, an imaginary driver's steering angle .theta.id is used to represent any one of these calculated steering angles .theta.id_lk, .theta.id_lcr, and .theta.id_lcl. In the exemplary implementation, the detected operation amount Ord of the real driver is expressed by the detected steering angle .theta.rd performed by the real driver.

Many mathematical calculations can be used to compute the likelihood value Pid. For example, the likelihood value Pid of each imaginary driver is a logarithmic probability of a normalized value of the imaginary driver's steering angle .theta.id relative to a normal distribution, where the mean (e) is the real driver's steering angle .theta.rd and the variance (.sigma.) is a predetermined value .rho.rd such as a standard deviation of steering angles. Generally, the value of .rho.rd depends on characteristics of the vehicle, such as the steering gear ratio, and/or the speed of the vehicle. .rho.rd may range from -15 degrees to +15 degrees, such as between 3 to 5 degrees. Of course, other values of .rho.rd may be used depending on the type and/or characteristics of vehicles.

The likelihood value Pid is expressed as: Pid=log {Probn[(.theta.id -.theta.rd)/.rho.rd]} (Eq. 7)

where Probn is a probability density function that is used to calculate a probability with which a given sample is observed from a population expressed by the normal distribution.

At step S105, using before-mentioned equation Eq. 7, the microcomputer calculates a likelihood value Pid(t) for each of the imaginary drivers of a dynamic family illustrated in FIG. 4. The calculated likelihood values are stored in the memory portions corresponding to each imaginary driver j, and are expressed as Pid(j)(t), where j corresponds to one of the imaginary drivers. Thus, Pid(j)(t) means a calculated likelihood value for an imaginary driver j having an intention at a point in time (t).

At step S106, using the stored likelihood values Pid(j)(t).about.Pid(j)(t-m+1), the microcomputer calculates a collective likelihood value P(j)ids for each imaginary driver j that is designed to behave as directed by intentions associated with each imaginary driver j. The collective likelihood value P(j)ids may be expressed as:

.function..times..times..times..times..times..function..times..times. ##EQU00001## Equation 8 states that the collective likelihood value P(j)ids is the product of the respective calculated likelihood values Pid(j)(t).about.Pid(j)(t-m+1).

At step S107, the microcomputer estimates a real driver's intention .lamda.rd. In this exemplary implementation, the microcomputer chooses one of the imaginary drivers that has the maximum calculated collective likelihood values P(j)ids among all imaginary drivers. The series of intentions corresponding to the chosen imaginary driver is now labeled Lmax. Then, the microcomputer chooses the latest intention of the series Lmax to approximate a real driver's intention .lamda.rd. The real driver's intention .lamda.rd may be expressed as: .lamda.rd=max[Pid(Lmax).sub.--lk(t), Pid(Lmax).s