Driving intention determination method, device and equipment and readable storage medium

Abstract

The invention discloses a driving intention determination method, device and equipment and a readable storage medium, and is applied to the technical field of intelligent driving, and the method comprises the steps: determining a turning membership degree corresponding to each piece of turning intention information and a lane changing membership degree corresponding to each piece of lane changing intention information based on a fuzzy membership degree method; performing fuzzy relation combination by using a defuzzification function based on all the turning membership degrees, and determining a left turning probability and a right turning probability; the defuzzification function is a function for determining an intention probability based on a membership degree and a weight coefficient; carrying out fuzzy relation combination by utilizing a defuzzification function based on all the lane changing membership degrees, and determining a left lane changing probability and a right lane changing probability; and determining a target driving intention based on the left turning probability, the right turning probability, the left lane changing probability and the right lane changing probability. According to the method, the membership degrees corresponding to the multiple types of intention information are determined, all the membership degrees are combined based on the defuzzification function, the target driving intention is determined, and the accuracy of driving intention determination is improved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent driving technology, and in particular to a driving intention determination method, device, equipment and readable storage medium. Background Art

[0002] The current behavior of evaluating driving intention mainly qualifies the driver's driving intention based on several indicators or a single indicator. This method of measuring driver behavior has great uncertainty. Because in the actual driving process, each driver's driving behavior is very different and irregular, such as changing lanes and turning without turning on the turn signal, turning on the left turn signal but changing lanes to the right, turning on the turn signal but not turning, etc., making it difficult to accurately identify driving intention.

[0003] It can be seen that how to improve the accuracy of determining driving intention is a technical problem that technical personnel in this field urgently need to solve. Summary of the invention

[0004] In view of this, an object of the present invention is to provide a driving intention determination method, device, equipment and readable storage medium, which solve the technical problem of inaccurate driving intention determination in the prior art.

[0005] In order to solve the above technical problems, the present invention provides a method for determining driving intention, comprising:

[0006] Determining the turning membership corresponding to each turning intention information and the lane changing membership corresponding to each lane changing intention information based on a fuzzy membership method; wherein the turning intention information and the lane changing intention information include vehicle motion state information and active intention information; the fuzzy membership method is a method for determining the membership corresponding to each intention information based on the set first membership information and second membership information corresponding to each intention information;

[0007] Based on all the turning memberships, a defuzzification function is used to combine fuzzy relations to determine the probability of a left turn and the probability of a right turn; wherein the defuzzification function is a function that determines the intention probability based on the membership and the weight coefficient;

[0008] Based on all lane change memberships, the fuzzy relationship combination is performed using the defuzzification function to determine the probability of a left lane change and the probability of a right lane change;

[0009] A target driving intention is determined based on the left turn probability, the right turn probability, the left lane change probability, and the right lane change probability.

[0010] Optionally, determining the turning membership corresponding to each turning intention information and determining the lane changing membership corresponding to each lane changing intention information based on the fuzzy membership method includes:

[0011] If the current intention information is an intention information whose value is larger and whose membership is closer to zero, the membership corresponding to the current intention information is determined based on the fuzzy small membership method; wherein the input of the fuzzy small membership method is the value x of the current intention information, the first membership parameter M1 and the second membership parameter M2; when the input x is greater than M2, the membership is 0; when the input x is less than M1, the membership is 1; when the input is between M1 and M2, the membership is (M2-x) / (M2-M1), M1 <M2;

[0012] If the current intention information is an intention information whose value is larger and whose membership is closer to one, the membership corresponding to the current intention information is determined based on the fuzzy large membership method; wherein the input of the fuzzy large membership method is the value y of the current intention information, the third membership parameter M3 and the fourth membership parameter M4, when the input y is less than M3, the membership is 0; when y is greater than M4, the membership is 1; when y is between M3 and M4, the membership is (y-M3) / (M4-M3), M3 <M4;

[0013] If the current turning intention information is a turn signal switch, it is determined that the membership degree corresponding to when the turn signal is turned on is 1, and the membership degree corresponding to when the turn signal is turned off is 0;

[0014] If the current steering intention information is the brake pedal of the vehicle, it is determined that the corresponding membership when the pedal is braking is 1, and the corresponding membership when the pedal is not braking is 0.

[0015] Optionally, determining the turning membership corresponding to each turning intention information and determining the lane changing membership corresponding to each lane changing intention information based on the fuzzy membership method includes:

[0016] If the current turning intention information is any one of the steering wheel angle, the yaw rate of the vehicle, the curvature of the vehicle trajectory, the longitudinal acceleration of the vehicle, the accelerator pedal opening of the vehicle, and the duration of the vehicle turn signal on, the membership corresponding to the current turning intention information is determined based on the fuzzy large membership method;

[0017] If the current turning intention information is the vehicle speed signal of the own vehicle, when calculating the membership of the vehicle speed signal of the own vehicle, the lowest level speed, the medium speed and the highest level speed are distinguished, and the membership corresponding to the current turning intention information is determined based on the fuzzy small membership method; wherein the highest level speed is greater than the medium speed, and the medium speed is greater than the lowest level speed.

[0018] Optionally, determining the turning membership corresponding to each turning intention information and determining the lane changing membership corresponding to each lane changing intention information based on the fuzzy membership method includes:

[0019] If the current lane change intention information is any one of a steering wheel angle, a turn signal switch signal, an angle of human eye rotation, a rate of change of the distance from the vehicle to the lane boundary, a curvature of the vehicle trajectory, and a turn signal on time, determining the membership corresponding to the current lane change intention information based on the fuzzy large membership method;

[0020] If the current lane change intention information is the standard deviation of the rate of change of the distance from the vehicle to the lane boundary, the membership corresponding to the current lane change intention information is determined based on the fuzzy small membership method.

[0021] Optionally, the method of combining fuzzy relations based on all turn memberships using a defuzzification function to determine the left turn probability and the right turn probability includes:

[0022] Based on all the turning memberships, the fuzzy relationship combination is performed using a fuzzy and defuzzification function and a fuzzy or defuzzification function to determine the left turn probability and the right turn probability;

[0023] The inputs of the fuzzy and defuzzified functions are the first membership L1, the second membership L2 and the weight coefficient K1. When L1 is less than or equal to L2, the probability value is K1*L1+(1-K1)*((L1+L2) / 2); when L1 is greater than L2, the probability value is K1*L2+(1-K1)*((L1+L2) / 2);

[0024] The input of the fuzzy or defuzzified function is the first membership L1, the second membership L2 and the weight coefficient K1. When L1 is less than or equal to L2, the probability value is K1*L2+(1-K1)*((L1+L2) / 2); when L1 is greater than L2, the probability value is K1*L1+(1-K1)*((L1+L2) / 2).

[0025] Optionally, the method of combining fuzzy relations based on all turn memberships using a defuzzification function to determine the left turn probability and the right turn probability includes:

[0026] Based on the fuzzy and defuzzification functions, fuzzy relationship combination is performed on the turning membership corresponding to the first type of turning intention information to obtain an initial left turn probability and an initial right turn probability;

[0027] Based on the fuzzy or defuzzified function, the initial turning probability and the initial right turn probability, the turning membership corresponding to the second type of turning intention information is fuzzily combined to obtain the final left turn probability and right turn probability; wherein the second type of turning intention information includes the longitudinal acceleration of the vehicle and the brake pedal of the vehicle, and the first type of turning intention is the turning intention information other than the second type of turning intention information.

[0028] Optionally, based on all lane change memberships, the fuzzy relationship combination is performed using the defuzzification function to determine the left lane change probability and the right lane change probability, including:

[0029] Determine whether there is a change in lane sequence value;

[0030] When there is a change in the lane sequence value, if the self-lane sequence value switches from the middle lane sequence value to the left lane sequence value, the left lane change probability is determined to be 1; if the self-lane sequence value switches from the middle lane sequence value to the right lane sequence value, the right lane change probability is determined to be 1;

[0031] When there is no change in the lane sequence value, the step of determining the left lane change probability and the right lane change probability by combining fuzzy relations based on all lane change memberships using the defuzzification function is performed.

[0032] Optionally, determining the target driving intention based on the left turn probability, the right turn probability, the left lane change probability, and the right lane change probability includes:

[0033] Determine the maximum probability among the left turn probability, the right turn probability, the left lane change probability and the right lane change probability;

[0034] Determining whether the maximum probability is greater than a minimum probability threshold;

[0035] When the maximum probability value is greater than the minimum probability threshold, determining the intention corresponding to the maximum probability value as the target driving intention;

[0036] When the maximum probability value is not greater than the minimum probability threshold value, it is determined that the target driving intention is going straight.

[0037] The present invention also provides a driving intention determination device, comprising:

[0038] A membership determination module, used for determining a turning membership corresponding to each turning intention information and a lane change membership corresponding to each lane change intention information based on a fuzzy membership method; wherein the turning intention information and the lane change intention information include vehicle motion state information and active intention information; the fuzzy membership method is a method for determining the membership corresponding to each intention information based on first membership information and second membership information set corresponding to each intention information;

[0039] A turning probability determination module, used to determine the probability of a left turn and the probability of a right turn by combining fuzzy relations based on all turning memberships using a defuzzification function; wherein the defuzzification function is a function for determining the intention probability based on the membership and the weight coefficient;

[0040] A lane change probability determination module, configured to determine a left lane change probability and a right lane change probability by combining fuzzy relations based on all lane change memberships using the defuzzification function;

[0041] A driving intention determination module is used to determine a target driving intention based on the left turn probability, the right turn probability, the left lane change probability and the right lane change probability.

[0042] The present invention also provides a driving intention determination device, comprising:

[0043] Memory for storing computer programs;

[0044] A processor is used to execute the computer program to implement the steps of the driving intention determination method as described above.

[0045] The present invention also provides a readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the driving intention determination method as described above are implemented.

[0046] The present invention also provides a computer program product, comprising a computer program / instruction, which implements the steps of the driving intention determination method as described above when the computer program / instruction is executed by a processor.

[0047] It can be seen that the present invention determines the turning membership corresponding to each turning intention information and the lane changing membership corresponding to each lane changing intention information based on the fuzzy membership method; wherein the turning intention information and the lane changing intention information include vehicle motion state information and active intention information; the fuzzy membership method is a method for determining the membership corresponding to each intention information based on the set first membership information and second membership information corresponding to each intention information; based on all the turning memberships, the fuzzy relationship combination is performed using the defuzzification function to determine the probability of turning left and the probability of turning right; wherein the defuzzification function is a function for determining the probability of intention based on the membership and the weight coefficient; based on all the lane changing memberships, the fuzzy relationship combination is performed using the defuzzification function to determine the probability of changing lane left and the probability of changing lane right; the target driving intention is determined based on the probability of turning left, the probability of turning right, the probability of changing lane left and the probability of changing lane right.

[0048] The beneficial effect of the present invention is that compared with the current method of measuring driving intention directly based on a single type of variable, the present invention uses a fuzzy membership method to determine the turning intention information based on the vehicle motion state information and the active intention information, the corresponding turning membership, and uses the fuzzy membership method to determine the lane change intention information based on the vehicle motion state information and the active intention information, the corresponding lane change membership, so as to determine the left turn probability and the right turn probability based on a plurality of turning memberships by fuzzy relationship combination, determine the left lane change probability and the right lane change probability based on a plurality of lane change memberships by fuzzy relationship combination, and finally determine the target driving intention based on the left turn probability, the right turn probability, the left lane change probability and the right lane change probability. Since the present application uses multiple types of turning intention information and lane change intention information in the process of determining the target driving intention, the subsequent fuzzy membership method designed based on various types of intention information in the present application performs more accurate membership calculation, and performs more accurate fuzzy relationship combination based on the defuzzification function, thereby improving the accuracy of determining the target driving intention.

[0049] In addition, the present invention also provides a driving intention determination device, equipment and readable storage medium, which also have the above-mentioned beneficial effects. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0051] Figure 1 A flowchart of a method for determining driving intention provided by an embodiment of the present invention;

[0052] Figure 2 A logical schematic diagram of a fuzzy small membership method provided by an embodiment of the present invention;

[0053] Figure 3 A logic diagram of a fuzzy large membership method provided by an embodiment of the present invention;

[0054] Figure 4 A schematic diagram of a membership degree determination logic for the duration of turning lights on a vehicle provided by an embodiment of the present invention;

[0055] Figure 5 A trend of a vehicle trajectory during a lane change process provided by an embodiment of the present invention;

[0056] Figure 6 A schematic diagram of a lane sequence value provided by an embodiment of the present invention;

[0057] Figure 7 A schematic diagram of a blurring and defuzzification function provided by an embodiment of the present invention;

[0058] Figure 8 A schematic diagram of a blurring or defuzzifying function provided by an embodiment of the present invention;

[0059] Fig. 9 An example flowchart of a method for determining driving intention provided by an embodiment of the present invention;

[0060] Fig.10 A schematic diagram of the structure of a driving intention determination device provided by an embodiment of the present invention;

[0061] Fig.11 A schematic diagram of the structure of a driving intention determination device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0062] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0063] Please refer to Figure 1 , Figure 1 A flowchart of a method for determining driving intention provided by an embodiment of the present invention. The method may include:

[0064] S101, determining the turning membership corresponding to each turning intention information and the lane changing membership corresponding to each lane changing intention information based on a fuzzy membership method; wherein the turning intention information and the lane changing intention information include vehicle motion state information and active intention information; the fuzzy membership method is a method for determining the membership corresponding to each intention information based on a first membership parameter and a second membership parameter set corresponding to each intention information.

[0065] The execution subject of this embodiment is an electronic device, and this embodiment does not limit the specific electronic device. The electronic device in this embodiment can be a computer, a tablet, etc. The types of turning intention information in this embodiment include vehicle operation state information related to turning, and active turning intention information corresponding to the driver's active operation information. The lane change intention information in this embodiment includes vehicle motion state information related to lane change, and active lane change intention information corresponding to the driver's active operation. This embodiment does not limit the specific turning intention information. For example, the vehicle motion state information in the turning intention information in this embodiment can be at least one of the vehicle speed signal, the yaw angular velocity of the vehicle, the curvature of the vehicle trajectory, the longitudinal acceleration of the vehicle, and the duration of the turn signal of the vehicle. The active intention information in the turning intention in this embodiment can be at least one of the steering wheel angle, the brake pedal of the vehicle, the accelerator pedal of the vehicle, and the turn signal switch. Among them, the steering wheel angle refers to the angle at which the driver turns the steering wheel, not the front wheel angle (the steering wheel angle divided by the transmission ratio of the steering mechanism is the front wheel angle), the brake pedal of the self-vehicle is whether the pedal is braked, that is, whether the driver has stepped on the brake pedal; the throttle pedal of the self-vehicle refers to the throttle pedal opening. When the throttle pedal is fully open, the throttle pedal opening value is 100%, and when the throttle pedal is half open, it is 50%, without units; the turn signal switch is the signal of the left turn signal and the signal of the right turn signal; the vehicle speed signal of the self-vehicle refers to the vehicle's own driving speed information; the yaw rate of the self-vehicle refers to the rotation speed of the vehicle around its vertical axis (i.e., the vertical line passing through the center of gravity of the vehicle) when turning. This speed describes the speed of the overall rotation of the vehicle and is an important indicator for measuring vehicle stability; the curvature of the self-vehicle trajectory refers to the rotation rate of the tangent direction angle at a certain point on the vehicle's driving trajectory to the arc length; the longitudinal acceleration refers to the acceleration of forward driving.

[0066] The lane change intention information in this embodiment includes vehicle motion state information related to lane change and active lane change intention information. In this embodiment, the vehicle motion state information related to lane change can be at least one of the change rate of the distance from the vehicle to the lane boundary, the standard deviation of the change rate of the distance from the vehicle to the lane boundary, the curvature of the vehicle trajectory, and the turn signal on time. In this embodiment, the active lane change intention information includes at least one of the steering wheel angle, the turn signal switch signal, and the angle of human eye rotation. Among them, the turn signal switch information is a left turn signal signal or a right turn signal; the human eye rotation angle is the left turn angle or the right turn angle; the change rate of the distance from the vehicle to the lane boundary includes the change rate of the distance from the vehicle to the left boundary of the lane and the change rate of the distance from the vehicle to the right boundary of the vehicle; the standard deviation of the change rate of the distance from the vehicle to the lane boundary includes the standard deviation of the change rate of the distance from the vehicle to the left boundary of the lane and the standard deviation of the change rate of the distance from the vehicle to the right boundary of the vehicle; the lane sequence value, for example, the left lane is 0, the middle lane is 1, and the right lane is 2; the turn signal on time includes the left turn signal on time and the right turn signal on time.

[0067] The fuzzy membership method in this embodiment is a method for determining the membership corresponding to each intention information based on the first membership parameter and the second membership parameter set corresponding to each intention information. The first membership parameter in this embodiment is the minimum value corresponding to the current intention information, and the second membership parameter is the maximum value corresponding to the current intention information. The fuzzy membership method in this embodiment is a method for comparing the current intention information with the first membership parameter and the second membership parameter, and determining the specific membership calculation method based on the current comparison result according to the comparison result. The comparison result includes that the current intention information is less than the first membership parameter, the first membership parameter is greater than the second membership parameter, and the current intention information is between the first membership parameter and the second membership parameter.

[0068] It should be further explained that, in order to determine the accuracy of the membership based on the fuzzy membership method, the above-mentioned determination of the turning membership corresponding to each turning intention information and the determination of the lane changing membership corresponding to each lane changing intention information based on the fuzzy membership method may include:

[0069] S1012, if the current intention information is an intention information whose value is larger and whose membership is closer to zero, the membership corresponding to the current intention information is determined based on the fuzzy small membership method; wherein the input of the fuzzy small membership method is the value x of the current intention information, the first membership parameter M1 and the second membership parameter M2; when the input x is greater than M2, the membership is 0; when the input x is less than M1, the membership is 1; when the input is between M1 and M2, the membership is (M2-x) / (M2-M1), M1 <M2。

[0070] In this embodiment, the first membership information is the first membership parameter, and the second membership information is the second membership parameter. Figure 2 , Figure 2 A logical schematic diagram of a fuzzy small membership method provided by an embodiment of the present invention.

[0071] S1013, if the current intention information is an intention information whose value is larger and whose membership is closer to one, the membership corresponding to the current intention information is determined based on the fuzzy large membership method; wherein the input of the fuzzy large membership method is the value y of the current intention information, the third membership parameter M3 and the fourth membership parameter M4, when the input y is less than M3, the membership is 0; when y is greater than M4, the membership is 1; when y is between M3 and M4, the membership is (y-M3) / (M4-M3), M3 <M4。

[0072] The first membership information in this embodiment is the third membership parameter, and the second membership information in this embodiment is the fourth membership parameter. For ease of understanding, please refer to Figure 3 , Figure 3A logical schematic diagram of a fuzzy large membership method provided by an embodiment of the present invention.

[0073] S1014: If the current turning intention information is a turn signal switch, determine that the membership corresponding to when the turn signal is turned on is 1, and the membership corresponding to when the turn signal is turned off is 0.

[0074] S1015, if the current steering intention information is the brake pedal of the vehicle, determine that the corresponding membership when the pedal is braking is 1, and the corresponding membership when the pedal is not braking is 0.

[0075] This embodiment provides a specific method for determining the membership corresponding to various intention information, thereby improving the accuracy of membership determination.

[0076] It should be further explained that the above-mentioned determination of the turning membership corresponding to each turning intention information and the determination of the lane change membership corresponding to each lane change intention information based on the fuzzy membership method may include: if the current turning intention information is any one of the steering wheel angle, the yaw angular velocity of the vehicle, the curvature of the vehicle trajectory, the longitudinal acceleration of the vehicle, the accelerator pedal opening of the vehicle and the turn signal on time of the vehicle, the membership corresponding to the current turning intention information is determined based on the fuzzy large membership method; if the current turning intention information is the vehicle speed signal of the vehicle, when calculating the membership of the vehicle speed signal, the lowest level speed, the medium speed and the highest level speed are distinguished, and the membership corresponding to the current turning intention information is determined based on the fuzzy small membership method; wherein the highest level speed is greater than the medium speed, and the medium speed is greater than the lowest level speed.

[0077] For ease of understanding, this embodiment provides a method for determining the corresponding membership of various turning intention information. When the turning intention information is the steering wheel angle, the membership when turning left and the membership of the steering wheel angle when turning right to the turning intention are respectively:

[0078] A(x_alpWheeLAnglekorrLeft=FuzzyLarge(m alpWheelAnelekorr , PagwheeLAngleLeft1, PagwheeLAngleLeft2) and

[0079] A(x_alpWheeLAnglekorrRight=FuzzyLarge(m alpWheelAnglekorr,PagwheeLAngleRight1,PagwheeLAngleRight2); wherein PagwheelAngleLeft1PagwheelAngleLeft2 represent the first steering wheel angle threshold and the second steering wheel angle threshold when turning left (the reference thresholds are 0.0125rad and 0.0325rad respectively); PagwheelAngleRight1, PagwheelAngleRight2 represent the first steering wheel angle threshold and the second steering wheel angle threshold when turning right (the reference thresholds are 0.0125rad and 0.0325rad respectively). For example, if the current steering wheel angle is 0.01rad, the membership is 0; if the steering wheel angle is 0.04rad, the membership is 1; if the steering wheel angle is 0.015, the calculated membership is: (0.015-0.0125) / (0.0325-0.0125)=0.125.

[0080] When the turning intention information is the turn signal switch, the method for determining the membership of left turn is The method to determine the membership of right turn is In this embodiment, the turn signal switch is a special fuzzy element. When the turn signal is turned on, the membership of the turn signal switch is 1, and when the turn signal is turned off, the membership of the turn signal switch is 0.

[0081] When the turning intention information is the vehicle speed, considering that the vehicle speed has a wide range of variation, three speed segments are distinguished when calculating the membership of the vehicle speed: low speed (corresponding to the lowest level speed mentioned above), medium speed and high speed (corresponding to the highest level speed mentioned above). For the medium speed segment, its membership is calculated as follows:

[0082] A(egpVelocityMedium)=FuzzySmall(megoVelocity, PAGeoVelocityMedium1, PAGeoVelocityMedium2); wherein PAGeoVelocityMedium1 and PAGeoVelocityMedium2 represent the first and second speed thresholds of medium speed respectively (reference thresholds are 15m / s and 16.7m / s). According to the definition of the fuzzy small membership method function, if the current vehicle speed is 14m / s, the calculated membership is 1; if the current vehicle speed is 18m / s, the calculated membership is 0; if the current vehicle speed is 15.8m / s, the calculated membership is: (16.7-15.8) / (16.7-15)=0.53. For high speeds, considering the turning radius and driving safety, it is considered that the turning intention is not obvious or there is no turning intention at all, so its membership can be considered to be equal to zero or there is no turning intention. A(egpVelocityHigh)=FuzzySmall(megoVelocity, PAGeoVelocityHigh1, PAGeoVelocityHigh2); wherein PAGeoVelocityHigh1 and PAGeoVelocityHigh2 represent the first speed threshold and the second speed threshold of high speed respectively (the reference threshold is 8.3m / s and 11.1m / s). For low speed, its membership is calculated as follows: A(egpVelocityLow)=FuzzySmall(megoVelocity, PAGeoVelocityLow1, PAGeoVelocityLow2); wherein PAGeoVelocityLow1 and PAGeoVelocityLow2 represent the first speed threshold and the second speed threshold of low speed respectively (the reference threshold is 3.5m / s and 5.0m / s).

[0083] When the turning intention information is the yaw rate of the vehicle, the method for determining the membership degree is: A(yawrateLeft) = FuzzyL arge(m yawrate , PAGyawrateLeft1, PAGyaw rateLeff2), A(yawrateRight)=FuzzyLarge(m yawrate, PAGyawrateRight1, PAGyawrateRigh2); PAGyawrateLeft1, PAGyawrateLeft2 represent the first yaw rate threshold and the second yaw rate threshold for left turn respectively; PAGyawrateRight1, PAGyawrateRight2 represent the first yaw rate threshold and the second yaw rate threshold for right turn respectively (reference thresholds are 0.1rad / s and 0.25rad / s). For example, when the yaw rate is 0.15rad / s, the method for calculating the membership is: (0.15-0.1) / (0.25-0.1)=0.33.

[0084] When the turning intention information is the curvature of the vehicle trajectory, the method for calculating the membership degree is:

[0085] A(kapTrajectoryLeft)=FuzzyLarge(m kapTrajectory ,PAGKapTrajectoryLeft1,PAGKapTrajectoryLeft2);

[0086] A(kapTrajectoryRight)=FuzzyLarge(m kapTrrajectory , PAGKapTrajectoryRight1, PAGKapTrajectoryRight2). Among them: PAGKapTrajectoryLeft1, PAGKapTrajectoryLeft2 represent the first threshold and the second threshold of the curvature of the ego vehicle trajectory when turning left; PAGKapTrajectoryRight1, PAGKapTrajectoryRight2 represent the first threshold and the second threshold of the ego vehicle trajectory when turning right (the reference thresholds are 0.02 / m and 0.06 / m). For example: when the curvature of the ego vehicle trajectory is 0.01, the membership is 0; when it is 0.08, the membership is 1; when it is 0.04, the membership is (0.04-0.02) / (0.06-0.02)=0.5.

[0087] When the turning intention information is the longitudinal acceleration of the ego vehicle, the method for determining the membership degree is: A(egoAccelerate) = FuzzyLarge(m Accelerate, PAGegoAccelerate1, PAGegoAccelerate2); where AGegoAccelerate1 and AGegoAccelerate2 represent the first longitudinal acceleration threshold and the second longitudinal acceleration threshold respectively (the reference thresholds are -0.98m / s2 and 0.98m / s2). According to the definition of the fuzzy large function, when the longitudinal acceleration of the vehicle is -1.2m / s 2 When the longitudinal acceleration of the vehicle is 1.2 m / s 2 When the longitudinal acceleration of the vehicle is 0.5 m / s 2 When , the membership degree is calculated as: (0.98-0.5) / (0.98-(-0.98))=0.24.

[0088] When the turning intention information is the brake pedal of the vehicle, the method for determining the membership degree is: Among them, when the driver steps on the brake pedal, the membership of the brake pedal is 1, and when the driver does not step on the brake pedal, the membership of the brake pedal is 0.

[0089] When the turning intention information is the ego vehicle's accelerator pedal opening, A(egoGasPedalPosition)=FuzzyLarge(m GasPedalPostion , PAGegoGasPedalPostion1, egoGasPedalPostion2); where, m GasPedalPostion Indicates the accelerator pedal opening of the vehicle. PAGegoGasPedalPosition1 and PAGegoGasPedalPosition2 represent the first accelerator pedal position threshold and the second accelerator pedal position threshold respectively (the reference thresholds are 0.075 and 0.125). The accelerator pedal refers to the accelerator pedal opening. When the accelerator pedal is fully open, the accelerator pedal opening value is 100%, and when the accelerator pedal is half open, it is 50%. There is no unit.

[0090] When the turning intention information is the duration of the vehicle's turn signal being on, the method for determining the degree of membership is as follows: the turn signal on duration signal is calculated based on the state of the turn signal being on, and its calculation function FuzzyCounter is as follows: Get parameters x, B1, B2, K1, K2, where x represents the state of the turn signal being on, B1 represents the state of the turn signal being on at the last moment, B2 is a flag, K1 represents the time the turn signal is on, and K2 represents the time threshold. The calculation logic of this function is as follows: If the turn signal state at the last moment was not on and the turn signal state at the current moment is on or the turn signal on time is greater than zero, and the flag is false, when none of these conditions are met, the calculation result is zero. When all of these conditions are met, continue to judge the following conditions: If the turn signal state at the last moment was not on and the turn signal state at the current moment is on, then the calculation result is the running cycle of the current program, if the current turn signal state is on or the current turn signal on time is less than the time threshold, then for each running cycle, the calculation result increases by one timing cycle, otherwise the calculation result is zero. For ease of understanding, please refer to Figure 4 , Figure 4 A schematic diagram of the logic for determining the duration of the vehicle's turn signal on provided by an embodiment of the present invention. Among them, t_cycletime represents the timing cycle (the time for each program to run is 0.02s). The turn signal on time of the left and right turn signals is calculated according to the turn signal on state.

[0091] m_timeofIndLeff=FuzzyCounter(egoLeftInd, m timeofIndLeftK1 ,false,m timeofIndLeft , 0); where egoLeftInd represents the state of the left turn signal, m timeofIndLeffK1 Indicates the left turn signal status of the previous cycle m_timeofIndRight=FuzzyCounter(egoRightInd,m timeofIndRightK1 ,false,m timeofIndRight ,0)

[0092] Among them, egoRightInd indicates the state of the right turn signal, m timeofIndRightK1 Indicates the state of the right turn signal in the previous cycle. The membership degrees of the left and right turn signal opening durations obtained above are:

[0093] A(timeLeftCounter)=FuzzyLarge(m timeofindLeft, PAGLeftCounterTime1, PAGLeftCounterTime2); wherein, PAGLeftCounterTime1 and PAGLeftCounterTime2 represent the first left turn signal on time threshold and the second left turn signal on time threshold respectively (the reference thresholds are 0.1 and 1).

[0094] A(timeRightCounter)=FuzztyLarge(m timeofindRightt ,PAGRightCounterTime1,PAGRightCounterTime2)

[0095] . Among them, PAGRightCounterTime1 and PAGRightCounterTime2 represent the first right turn signal on time threshold and the second right turn signal on time threshold respectively (reference thresholds are 0.1 and 1). For example: the turn signal switch is turned on in the current cycle, and the turn signal is turned off in the previous cycle. The condition for calculating the turn signal on time is met, and the next condition calculation will be performed. If this condition continues to be met, it means that the turn signal is turned on once, and the calculation time is only 0.02s. If this condition is not met, it means that the turn signal was turned on in the previous cycle, that is, the turn signal has been turned on since the previous cycle, and a 0.02s will be superimposed, and the judgment of the next cycle will be performed in turn, thereby superimposing multiple calculation cycles; if the timing result is less than a certain value such as 0.12s, it is considered that the turn signal on time cannot represent the valid time; or if the turn signal is turned off, the calculation result will be directly cleared to zero.

[0096] It should be further explained that, in order to improve the accuracy of determining the lane change membership, the above-mentioned determination of the turning membership corresponding to each turning intention information based on the fuzzy membership method, and determination of the lane change membership corresponding to each lane change intention information, may include: if the current lane change intention information is any one of the steering wheel angle, the turn signal switch signal, the angle of human eye rotation, the rate of change of the distance from the vehicle to the lane boundary, the curvature of the vehicle trajectory and the turn signal on time, determining the membership corresponding to the current lane change intention information based on the fuzzy large membership method; if the current lane change intention information is the standard deviation of the rate of change of the distance from the vehicle to the lane boundary, determining the membership corresponding to the current lane change intention information based on the fuzzy small membership method.

[0097] In this embodiment, when the lane change intention information is the rate of change of the distance from the vehicle to the lane boundary, the corresponding lane change trend is as follows: Figure 5 As shown, Figure 5The embodiment of the present invention provides a trend of the trajectory of the vehicle during the lane change process, and judges whether the vehicle has a lane change trend according to the difference in distance between the vehicle and the left and right lane edges twice before and after. LeftBorderDescend=(m LeftBorder -mLeftBorderk1) / t_cycletime;

[0098] RightBorderDescend=(m RightBorder -mRightBorderk1) / t_cycletime; where m LeftBorder is the distance between the current vehicle and the left lane edge, mLeftBorderk1 is the distance between the vehicle and the left lane edge in the previous cycle, m RightBorder is the distance between the current vehicle and the right lane edge, and mRightBorderk1 is the distance between the vehicle and the right lane edge in the previous cycle. The membership degrees of the distance between the vehicle and the lane boundary during the left and right lane changes are calculated as follows:

[0099] A(LeftBorderDescend)=FuzzyLarge(LeftBorderDescend,PAGLeftBorder1,PAGLeftBorder2); A(RightBorderDescend)=FuzzyLarge(RightBorderDescend,PAGRightBorder1,PAGRightBorder2); wherein, PAGLeftBorder1 and PAGLeftBorder2 represent the thresholds of the rate of change of the distance from the first vehicle to the left lane edge and the rate of change of the distance from the second vehicle to the left lane edge when changing lanes to the left, respectively; PAGRightBorder1 and PAGRightBorder2 represent the thresholds of the rate of change of the distance from the first vehicle to the right lane edge and the rate of change of the distance from the second vehicle to the right lane edge when changing lanes to the right (the reference thresholds are 0.15m / s and 0.33m / s).

[0100] When the current lane change intention information is the standard deviation of the distance from the vehicle to the lane boundary, the calculation formula for determining the membership degree is: A(LeftBorderStd) = FuzzySmall(m stdLeftBorder, PAGLeftBorderStd1,LeftBorderStd2), A(RightBorderStd)=FuzzySmall(m stdRightBorder,PAGRightBorderStd1, RightBorderStd2); where PAGLeftBorderStd1 and PAGLeftBorderStd2 represent the standard deviation of the rate of change of the distance from the first vehicle to the left lane edge and the standard deviation of the rate of change of the distance from the second vehicle to the left lane edge when turning left; PAGRightBorderStd1 and PAGRightBorderStd2 represent the standard deviation of the rate of change of the distance from the first vehicle to the right lane edge and the standard deviation of the rate of change of the distance from the second vehicle to the right lane edge when turning right (the reference thresholds are 0.4 and 1.44).

[0101] When the lane change intention information is the curvature of the vehicle trajectory, the method for determining the membership degree is:

[0102] A(kap TrajectoryChangeLeft)=FuzzyLarge(m kapTrajectory ,PAGChang eLeft3,PAGChangeLeft4);

[0103] A(kapTrajectoryChangeRight)=FuzzyL arge(m kapTrajectory , PAGChangeRight3,PAGChangeRight4); wherein, PAGChangeLeft3 and PAGChangeLeft4 represent the third and fourth curvature thresholds of the vehicle when changing lanes to the left, respectively; PAGChangeRight3 and PAGChangeRight4 represent the third and fourth curvature thresholds of the vehicle when changing lanes to the right, respectively (the reference thresholds are 0.015 and 0.016).

[0104] When the lane change intention information is the lane sequence value of the vehicle, the change of the lane sequence value can also directly reflect the driver's driving intention. By detecting the switch between the current lane sequence value and the lane sequence value of the previous cycle, it is determined whether the lane is changing. Figure 3 As shown in , 0 represents the left lane, 1 represents the current lane, and 2 represents the right lane. If the vehicle was in lane 1 in the previous cycle and is currently in lane 0, it means that the vehicle is changing lanes to the left; if the vehicle is currently in lane 2 and was in lane 1 in the previous cycle, it means that the vehicle is changing lanes to the right. Figure 6 As shown, Figure 6 A schematic diagram of a lane sequence value provided by an embodiment of the present invention.

[0105]

[0106] When the lane change intention information is the steering wheel angle, the method for determining the membership degree is:

[0107] A(LeftWheelAngle)=FuzzyLarge(m alpWheelAnglekorr ,PAGWheelAngleLeft3,PAGWheelAngleLeft4);

[0108] A(RightWheelAngle)=FuzzyLarge(m alpWheelAnglekorr , PAGWheelAngleRight3,PAGWheelAngleRight4). Among them, PAGWheelAngleLeft3 and PAGWheelAngleLeft4 respectively represent the third steering wheel angle threshold and the fourth steering wheel angle threshold when changing lanes to the left; PAGWheelAngleRight3 and PAGWheelAngleRight4 respectively represent the third steering wheel angle threshold and the fourth steering wheel angle threshold when changing lanes to the right (the reference thresholds are 0.0125rad and 0.0325rad). For example: if the current direction angle is 0.01rad, the membership is 0; if the current steering wheel angle is 0.04rad, the calculated membership is 1; if the steering wheel angle is 0.02rad, the membership is: (0.02-0.0125) / (0.0325-0.0125)=0.375.

[0109] When the lane change intention information is the turn signal switch, the method for determining the membership degree is:

[0110] A(leftIndicator)=FuzzyLarge(m_leftIndicator,PAGleffIndicator3,PAGleffIndicator4);

[0111] A(rightIndicator)=FuzzyLarge(m_rightIndicator,PAGrightIndicator3,PAGrightIndicator4). Among them, PAGleffIndicator3 and PAGleftIndicator4 respectively represent the third turn signal switch threshold and the fourth turn signal threshold when changing lanes to the left; PAGrightIndicator3 and PAGrightIndicator4 respectively represent the third turn signal switch threshold and the fourth turn signal threshold when changing lanes to the right. (The reference threshold is 1)

[0112] When the lane change intention information is the turn signal on time, the method for determining the membership degree is: according to the turn signal on state, the turn signal on time of the left and right turn signals is calculated respectively,

[0113] m_timeofindChangeLeft=FuzzyCountel'(egoLeftInd,m timeoflndLeftKl ,false,m timeoflndLeft , 0).

[0114] Among them, egoLeftInd indicates the state of the left turn signal, m timeoflndLeftK1 Indicates the left turn signal status in the previous cycle.

[0115] m_timeofindChangeRight=FuzzyCounter(egoRightlnd, m timeofIndRightK1 ,false,m timeoflndRight , 0).

[0116] Among them, egoRightlnd indicates the state of the right turn signal, m timeofIndRightK1 Indicates the state of the right turn signal in the previous cycle. The membership degrees of the left and right turn signal opening durations obtained above are:

[0117] A(timeChangeLeftCounter)=FuzzvLarge(m timeoflndChangeLeft ,PAGleftCounterTime3,PAGleftCounterTime4);

[0118] Among them, PAGleftCounterTime3 and PAGleftCounterTime4 respectively represent the third left turn signal on time threshold and the fourth left turn signal on time threshold.

[0119] A(timeChangeRigilttCounter)=FuzzvLarge(m timeoflndCh a ngeRight ,PAGRightCounterTime3,PAGRightCounterTime4).

[0120] Among them, PAGrightCounterTime3 and PAGrightCounterTime4 represent the third right turn signal on time threshold and the fourth right turn signal on time threshold (reference thresholds 2.0s and 2.5s), respectively. For example, when the turn signal is on for 1.5s, the membership is 0; if the turn signal is on for 3s, the membership is 1; if the turn signal is on for 2.2s, the membership is: (2.2-2) / (2.5-2)=0.4.

[0121] When the lane change intention information is the angle of eye rotation, the membership includes the left turning angle m_eyeRotationLeft and the right turning angle m_eyeRotationRight.

[0122] A(rotationLeft)=FuzzyLarge(rotationAngle, RotationLeff1, RotationLeft2);

[0123] A(rotationRight)=FuzzyLarge(-rotationAngle, RotationRight1, RotationRight2).

[0124] Among them, RotationLeft1 and RotationLeft2 respectively represent the first eye rotation angle threshold and the second eye rotation angle threshold when changing lanes to the left; RotationRight1 and RotationRight2 respectively represent the first eye rotation angle threshold and the second eye rotation angle threshold when changing lanes to the right (reference thresholds 0.69rad and 1.39rad).

[0125] S102, based on all the turning memberships, a defuzzification function is used to combine fuzzy relations to determine the left turn probability and the right turn probability; wherein the defuzzification function is a function that determines the intention probability based on the membership and the weight coefficient.

[0126] The defuzzification function in this embodiment may include a fuzzy or defuzzification function and a fuzzy and defuzzification function, and the specific defuzzification function corresponding to the current turning intention may be determined. For example, the method for combining the fuzzy relationship between the brake pedal of the vehicle and other turning intentions (or the probability of the turning intention after the combination) is a fuzzy or defuzzification function, and the method for combining the fuzzy relationship between the turn signal and other turning intentions is a fuzzy and defuzzification function. The weight coefficient in this embodiment is a value set based on demand, for example, the value may be 0.9, 0.8, etc.

[0127] It should be further explained that, in order to improve the accuracy of determining the turning probability, the above-mentioned determination of the left turn probability and the right turn probability by using the defuzzification function to perform fuzzy relationship combination based on all the turning memberships may include: based on all the turning memberships, using the fuzzy and defuzzification function, the fuzzy or defuzzification function to perform fuzzy relationship combination to determine the left turn probability and the right turn probability; wherein the input of the fuzzy and defuzzification function is the first membership L1, the second membership L2 and the weight coefficient K1, When L1 is less than or equal to L2, the probability value is K1*L1+(1-K1)*((L1+L2) / 2); when L1 is greater than L2, the probability value is K1*L2+(1-K1)*((L1+L2) / 2); the input of the fuzzy or defuzzified function is the first membership L1, the second membership L2 and the weight coefficient K1. When L1 is less than or equal to L2, the probability value is K1*L2+(1-K1)*((L1+L2) / 2); when L1 is greater than L2, the probability value is

[0128] K1*L1+(1-K1)*((L1+L2) / 2). The two main functions of the custom defuzzification of the present invention are: FuzzyAnd (fuzzy and defuzzification function) and FuzzyOr (fuzzy or defuzzification function). Among them, the input of the defuzzification function FuzzyAnd is X2 (corresponding to L1 above) and X3 (corresponding to L2 above), and K1 is the weight coefficient. The calculation process is as follows: when X2 is less than or equal to X3, the calculation result is K1*X2+(1-K1)*((X2+X3) / 2); when X2 is greater than X3, the calculation result is K1*X3+(1-K1)*((X2+X3) / 2). Figure 7 A schematic diagram of a fuzzy and defuzzification function provided by an embodiment of the present invention; the input of the defuzzification function FuzzyOr is X2 and X3, and K1 is the weight coefficient. The calculation process is as follows: When X2 is less than or equal to X3, the calculation result is

[0129] K1*X3+(1-K1)*((X2+X3) / 2); when X2 is greater than X3, the calculation result is K1*X2+(1-K1)*((X2+X3) / 2). Figure 8A schematic diagram of a fuzzy or defuzzified function provided for an embodiment of the present invention. The choice of whether to select a fuzzy or fuzzy or defuzzified function is mainly based on the contribution of describing the turning intention or lane changing intention from different dimensions. For example, a large steering wheel angle and the driver stepping on the brake pedal can better characterize that the driver is currently turning. This combination is mainly adjusted based on actual driving habits and experience. When this embodiment is combined, it can be grouped and combined based on the type of intention information, and then combined as a whole, that is, the membership corresponding to the active intention information can be combined first, and then the membership corresponding to the vehicle motion state information can be combined, and finally the two results can be combined to obtain the probability of turning left and turning right or the probability of changing lanes left and changing lanes right.

[0130] It should be further explained that the above-mentioned fuzzy relationship combination based on all turning memberships using a defuzzification function to determine the probability of left turn and the probability of right turn can include: fuzzy relationship combination of the turning memberships corresponding to the first type of turning intention information based on fuzzy and defuzzification functions to obtain the initial left turn probability and the initial right turn probability; fuzzy relationship combination of the turning memberships corresponding to the second type of turning intention information based on the fuzzy or defuzzification function, the initial turning probability and the initial right turn probability to obtain the final left turn probability and the right turn probability; wherein the second type of turning intention information includes the longitudinal acceleration of the vehicle and the brake pedal of the vehicle, and the first type of turning intention is the turning intention information other than the second type of turning intention information. For ease of understanding, this embodiment gives the specific process of fuzzy relationship combination: Combination of memberships corresponding to active intention information: ① Fuzzy relationship combination of the steering wheel angle and the turn signal on time to obtain the corresponding probabilities of left turn and right turn intentions:

[0131] m probIntentionLeft1 =FuzzyAnd(p0,A(timeLeftCounter),A(X_alpWheelAnglekorrLeft));

[0132] m probIntentionRight1=FuzzyAnd(p0, A(timeRightCounter), A(X_alpWheelAnglekorrRight)). Among them, p0 is the first weight coefficient, and the reference value is 0.5. A(timeLeftCounter) represents the membership of the left turn signal on time, A(X_alpWheelAnglekorrLeft) represents the membership of the left steering wheel angle, A(timeRightCounter) represents the membership of the right turn signal on time, and A(X_alpWheelAnglekorrRight) represents the membership of the right steering wheel angle. For example, the membership of the turn signal on time is 0.4, and the membership of the steering wheel angle is 0.3. According to the definition of the defuzzification function, the probability of obtaining the intention is 0.5*0.3+(1-0.5)*((0.4+0.3) / 2)=0.325. ② Combine the probability obtained by the above combination ① with the turn signal for a second time through fuzzy relationship to obtain the corresponding probability of left turn and right turn intention: m probIntentionLeft2 =FuzzyAnd(p0,A(dirIndicatorLeft),m probIntentionLeft1 );m probIntentionRight2 =FuzzyAnd(p0,A(dirIndicatorRight),m probIntentionRight1 ). Among them, A(dirIndicatorLeft) represents the membership of the left turn signal, and A(dirIndicatorRight) represents the membership of the right turn signal. ③ Combine the probability obtained by the above ② combination with the vehicle speed (lowest level speed, medium speed) for the third time to obtain the corresponding probabilities of left turn and right turn intentions:

[0133] m probIntentionLeft3 =FuzzyAnd(p0,A(egoVelocityMedium),m probIntentionLeft2 );

[0134] m probIntentionRight3 =FuzzyAnd(p0,A(egoVelocityMedium),m probIntentionRight2 ).in,

[0135] A(egoVelocityMedium) represents the degree of membership corresponding to the ego vehicle speed. It is understandable that when the vehicle speed exceeds a certain value (the threshold of this speed is 16.7m / s (60km / h)), the evaluation of the driver's intention in the first part is no longer performed. Because according to daily driving experience, when the ego vehicle turns left and right, on the premise of ensuring safe driving, the vehicle will maintain a relatively reasonable medium and low speed. Too high a speed loses the meaning of evaluation. For example, when the degree of membership of the vehicle speed is 0.8, the probability of the turn signal on time and the steering wheel angle after defuzzification is 0.325, and the probability of the intention after superposition is: 0.5*0.325+(1-0.5)*

[0136] ((0.8+0.325) / 2)=0.44375. The combination of vehicle motion state information corresponding to the degree of membership: ④ The yaw rate and the curvature of the vehicle trajectory are combined for the fourth time to obtain the corresponding probability of left turn and right turn intention: m probIntentionLeft4 =FuzzyAnd(p0,A(kapTrajectory),A(yawrateLeft));m probIntentionRight4 =FuzzyAnd(p0, A(kapTrajectory), A(yawrateRight)). Among them, A(kapTrajectory) represents the membership corresponding to the curvature of the vehicle's trajectory, A(yawrateLeft) represents the membership of the yaw rate of the left turn; A(yawrateRight) represents the membership of the yaw rate of the right turn. For example, if the membership of the yaw angle is 0.25 and the membership of the curvature of the vehicle's trajectory is 0.5, then the probability of the yaw angle and the curvature of the vehicle's trajectory after defuzzification is: 0.5*0.25+(1-0.5)*((0.25+0.5) / 2)=0.5. ⑤ Combine the low speed and the accelerator pedal opening of the vehicle for the fifth time to obtain the probability of low-speed driving:

[0137] m probIntention5 =FuzzyAnd(p0, A(egoGasPedalPositon), A(egoVelocityLow)); where A(egoGasPedalPositon) represents the membership of the ego vehicle's accelerator pedal opening, and A(egoVelocityLow) represents the membership of low equal speed. For example, the membership of low equal speed is 0.35 and the membership of accelerator pedal opening is 0.4. Then the defuzzified probability obtained by superimposing speed and accelerator pedal opening is 0.5*0.35+(1-0.5)*((0.35+0.4) / 2)=0.55. ⑥ Combine the probability obtained in ⑤ above with the ego vehicle's brake pedal for the sixth time to obtain the probability of low speed driving: m probIntention6=FuzzyOr(p0,A(brakePedal),m probIntention5 ); where A(brakePedal) represents the membership of the brake pedal of the vehicle. For example, the membership of the brake pedal is 0.1, and the probability calculated in ⑤ is 0.55. The probability of the defuzzified superimposed brake pedal is: 0.5*0.1+(1-0.5)((0.1+0.55) / 2)=0.2125. ⑦ Combine the probability obtained in ⑥ and the turn signal on time for the seventh time, and the corresponding probability of left and right turns is:

[0138] m probIntentionLeft7 =FuzzyAnd(p0,A(timeLeftCounter),mprob Intention6 );m probIntentionRight7 =FuzzyAnd(p0,A(timeRightCounter),m probIntention6 ). Among them, A(timeLeftCounter) represents the membership of the left turn signal on time, and A(timeRightCounter) represents the membership of the right turn signal on time. The membership of the turn signal on time is 0.6, and the probability calculated in ⑥ is 0.2125. The superposition of the turn signal on time gives the defuzzified probability: 0.2125*0.5+(1-0.5)*((0.2125+0.6) / 2)=0.309. ⑧ The probability obtained in ⑦ and the curvature of the vehicle trajectory are combined for the eighth time to obtain the corresponding probabilities of left and right turns:

[0139] m probIntentionLeft8 =FuzzyAnd(p0,A(kapTrajectoryLeft),m probIntentionLeft7 );

[0140] m probIntentionRight8 =FuzzyAnd(p0,A(kapTrajectoryRight),m probIntentionTight7 ). Among them, A(kapTrajectoryLeft) represents the degree of membership of the curvature of the ego trajectory when turning left; A(kapTrajectoryRight) represents the degree of membership of the curvature of the ego trajectory when turning right. For example, the degree of membership of the curvature of the ego trajectory is 0.8 and the probability calculated in ⑦ is 0.309, thus the curvature of the ego trajectory is superimposed, and the defuzzified probability is: 0.5*0.309+(1-0.5)*((0.8+0.309) / 2)=0.432. ⑨ The probability obtained in the above ⑧ and the steering wheel angle of the ego vehicle are combined for the ninth time, and the corresponding probabilities of left and right turns are obtained:

[0141] m probIntentionLeft9=FuzzyAnd(p0,A(X_alpWheelAnglekorrLeft),m prob I ntentionLeft8 );

[0142] m probIntentionRight9 =FuzzyAnd(p0,A(X_alpWheelAnglekorrRight),m probIntentionRight8 ). For example, the membership of the steering wheel angle is 0.7 and the probability calculated in ⑧ is 0.432, so the steering wheel angle is superimposed, and the defuzzified probability is: 0.432*0.5+(1-0.5)*((0.432+0.7) / 2)=0.499. ⑩ The probability obtained in ⑨ and the probability obtained in ④ are combined for the tenth time, and the corresponding probability of left and right turns is: m probIntentionLeft10 =FuzzyOr(p0,m probIntentionLeft4 , m probIntentionLeft9 ), m probIntentionRight10 =FuzzyOr(p0,m probIntentionRight4 , m probIntentionRight9 ). For example, the probability calculated by ⑨ is 0.499 and the probability calculated by ④ is 0.5, and the probability of defuzzification obtained by superimposing ⑨ and ④ is: 0.5*0.499+(1-0.5)*((0.5+0.499) / 2)=0.4992. Determination of the driver's turning intention: Combine the probabilities calculated by ⑩ and ③ above to get the final probability of turning left or right:

[0143] m probIntentionLeft =FuzzyAnd(p0,m probIntentionLeft3 , m probIntentionLeft10 ), m probIntentionRight =FuzzyAnd(p0,m probIntentionRight3 , m probIntentionRight10 ). For example, the result calculated in ⑩ and the result calculated in ③, the probability of the defuzzified turning intention after superposition is: 0.444*0.5+(1-0.5)*((0.444+0.4992) / 2)=0.46. It is not specified here whether it is the probability of turning left or right, which should be determined according to the actual situation.

[0144] S103, based on all lane change memberships, a defuzzification function is used to perform fuzzy relationship combination to determine the probability of a left lane change and the probability of a right lane change.

[0145] It should be further explained that, in order to improve the accuracy of determining the probability of lane change, the above-mentioned fuzzy relationship combination based on all lane change memberships using the defuzzification function to determine the probability of left lane change and the probability of right lane change may include: S1031, judging whether there is a change in the lane sequence value; S1032, when there is a change in the lane sequence value, determining that if the lane sequence value switches from the middle lane sequence value to the left lane sequence value, the probability of left lane change is 1; if the lane sequence value switches from the middle lane sequence value to the right lane sequence value, the probability of right lane change is 1; S1033, when there is no change in the lane sequence value, determining to perform the step of combining the fuzzy relationship based on all lane change memberships using the defuzzification function to determine the probability of left lane change and the probability of right lane change. It can be understood that, in general, the lane sequence value switching intuitively reflects the driver's driving intention, such as the lane sequence value switching from 1 to 0 indicates a left lane change, and the lane sequence value switching from 1 to 2 indicates a right lane change (i.e., the probability of right lane change is 100%).

[0146] The process of determining the lane change intention in this embodiment may include: combining the membership degree of the vehicle motion state information, combining the curvature of the vehicle trajectory and the rate of change of the distance between the vehicle and the edge line, and obtaining the corresponding probability of changing lanes left or right:

[0147] m proIntentionChangeLeft1 =FuzzyAnd(p1,A(kapTrajectoryChangeLefi),A(lefiBorderDescend)),

[0148] m probltentionChangeRight1 =FuzzyAnd(p1, g(kapTrajectoryChangeRight), A(RightBorderDescend)). Where A(kapTrajectoryChange) represents the degree of membership corresponding to the curvature of the vehicle trajectory, and A(BorderDescend) represents the degree of membership corresponding to the rate of change of the distance from the vehicle to the border. Where p1 is the second weight coefficient, and the reference coefficient is 1.0. ② The probability calculated above is combined with the standard deviation of the rate of change of the distance from the vehicle to the border for the second time, and the corresponding probability of changing lanes left or right is obtained:

[0149] m proIntentionChangeLeft2 =FuzzyOr(p1,A(lefiBorderStd),m proIntentionChangeLeft1 ), m probmmtionChangeRight2 =FuzzyOr(p1,A(RightBorderStd),m proIntentionChangeRight1); A(BorderStd) represents the degree of membership corresponding to the standard deviation of the rate of change of the distance between the vehicle and the border line. Active lane change intention: ③ Combine the steering wheel angle and the turn signal on time to obtain the corresponding probability of changing lanes left or right: m probaentionChangeLqft3 =FuzzyAnd(p1,A(timeChangeLeftCounter),A(lefiWheelAngel)),m probntentionChangeRight3 =FuzzyAnd(p1, A(timeChangeRightCounter), A(RightWheeIAngel)); where A(timeChangeLeftCounter) represents the membership corresponding to the time when the left turn signal is turned on, and A(leftWheeMngel) represents the membership corresponding to the left steering wheel angle. ④ Combine the probability calculated in ③ above with the turn signal signal to obtain the corresponding probability of changing lanes left or right:

[0150] m probltentionChangeLeft4 =FuzzyAnd(p1,A(lefilndicator),m proIntentionChangeLeft3 ), m proIntentionChangeRight4 =FuzzyAnd(pl,A(Rightlndicator),m prvIntentionChangeRightt3 ). ⑤ Combine the probability calculated in ④ above with the driver's eye rotation angle to obtain the corresponding probability of changing lanes left or right: m probltentionChangeLefi 5=FuzzyAnd(p1,A(rotationlefi),m probztentionChangeLefi4) , m proIntentionChangeRight5 =FuzzyAnd(p1,A(rotationRight),m probttentionChangeRight4 ). Combination of membership degrees corresponding to active lane change intention information: ⑥ Combining the probability calculated in ⑤ above with the probability calculated in ② above, the final corresponding probability of changing lanes left or right is obtained:

[0151] m proIntentionChangeLeft =FuzzyAnd(p1,m proIntentionChangeLqft2 , m proIntentionChangeLqft5 ), m proIntentionChangeRight =FuzzyAnd(p1,m proIntentionChangeRight2 , m proIntentionChangeRight5 ).

[0152] S104, determining a target driving intention based on the left turn probability, the right turn probability, the left lane change probability, and the right lane change probability.

[0153] This embodiment does not limit the specific method for determining the target driving intention based on the left turn probability, the right turn probability, the left lane change probability, and the right lane change probability. For example, this embodiment can determine the maximum probability value among the left turn probability, the right turn probability, the left lane change probability, and the right lane change probability to determine the target driving intention; or this embodiment determines the target driving intention based on the left turn probability, the right turn probability, the left lane change probability, and the right lane change probability when it is determined that there is no lane sequence value change, and determines the target driving intention based on the left turn probability, the right turn probability, the left lane change probability, and the right lane change probability when there is a lane sequence value change.

[0154] When the target driving intention is determined based on the change of the lane sequence value. When there is only a left turn probability in the whole process, the right turn probability is 0; or when there is only a left lane change probability, the right lane change probability is 0. This embodiment can determine the target driving intention based only on the left turn probability and the left lane change probability; or this embodiment can also determine the target driving intention based on the right turn probability and the left lane change probability; or this embodiment can also determine the target driving intention based on the left turn probability and the right lane change probability; or this embodiment can also determine the target driving intention based on the right turn probability and the right lane change probability. The target driving intention in this embodiment mainly includes the driver's driving intention mainly including lane keeping in the same lane, lane changing in adjacent lanes and turning at intersections.

[0155] It should be further explained that, in order to improve the accuracy of determining the target driving intention, determining the target driving intention based on the left turn probability, the right turn probability, the left lane change probability and the right lane change probability may include:

[0156] S1041, determining the maximum probability among the left turn probability, the right turn probability, the left lane change probability and the right lane change probability;

[0157] S1042, determining whether the maximum probability value is greater than the minimum probability threshold value;

[0158] S1043, when the maximum probability is greater than the minimum probability threshold, determining the intention corresponding to the maximum probability as the target driving intention;

[0159] S1044: When the maximum probability value is not greater than the minimum probability threshold value, it is determined that the target driving intention is to go straight.

[0160] The embodiment of the present invention provides a specific method for determining a target driving intention, thereby improving the accuracy of determining the target driving intention.

[0161] A method for determining driving intention provided by an embodiment of the present invention may include: S101, determining the turning membership corresponding to each turning intention information and the lane change membership corresponding to each lane change intention information based on a fuzzy membership method; wherein the turning intention information and the lane change intention information include vehicle motion state information and active intention information; the fuzzy membership method is a method for determining the membership corresponding to each intention information based on a first membership parameter and a second membership parameter set corresponding to each intention information; S102, determining the probability of a left turn and the probability of a right turn by combining fuzzy relations based on all turning memberships using a defuzzification function; wherein the defuzzification function is a function for determining the probability of intention based on a membership and a weight coefficient; S103, determining the probability of a left lane change and the probability of a right lane change by combining fuzzy relations based on all lane change memberships using a defuzzification function; S104, determining the target driving intention based on the probability of a left turn, the probability of a right turn, the probability of a left lane change, and the probability of a right lane change. Compared with the current method of measuring driving intention directly based on a single variable, the present invention determines the membership corresponding to the intention information based on multi-parameter intention information composed of vehicle motion state information and active intention, and finally combines all the memberships based on the defuzzification function to determine the target driving intention, thereby improving the accuracy of driving intention determination.

[0162] The current behavior of evaluating driving intention mainly uses a few single indicators to qualitatively determine the driver's driving intention, such as judging whether the current driver is keeping the lane through the steering wheel angle or hand torque, judging whether the current driver is changing lanes through the turn signal, and judging whether the current driver is turning through the turn signal, steering wheel angle and yaw angle. This method of measuring the driver's behavior through a single quantity has great uncertainty. Because in the actual driving process, each driver's driving behavior has great differences and irregular driving behavior, such as changing lanes and turning without turning on the turn signal, turning on the left turn signal but changing lanes to the right, turning on the turn signal but not turning, etc., which is not conducive to the accurate identification of driving intention.

[0163] In order to make the present invention easier to understand, please refer to Fig. 9 , Fig. 9 A flowchart of a method for determining driving intention provided by an embodiment of the present invention may specifically include:

[0164] S201: Acquire a turning intention information set; wherein the turning intention information set includes vehicle motion state information related to turning, and active turning intention information.

[0165] The turning intention information set in this embodiment includes the steering wheel angle, turn signal switch, vehicle speed signal, vehicle yaw angular velocity, vehicle trajectory curvature, vehicle longitudinal acceleration, vehicle brake pedal, vehicle accelerator pedal opening, and vehicle turn signal on time.

[0166] S202: Acquire a lane change intention information set; wherein the lane change intention information set includes vehicle motion state information related to lane change, and active lane change intention information.

[0167] The lane change intention information set in this embodiment includes the rate of change of the distance from the vehicle to the lane boundary, the standard deviation of the rate of change of the distance from the vehicle to the lane boundary, the curvature of the vehicle trajectory, the lane sequence value of the vehicle, the steering wheel angle, the turn signal switch signal, the turn signal on time, and the angle of human eye rotation.

[0168] S203: Determine the turning membership corresponding to each turning intention information in the turning intention information set based on the fuzzy small membership method and the fuzzy large membership method.

[0169] S204: According to each turn membership, fuzzy relationship combination is performed based on the fuzzy and defuzzification function and the fuzzy or defuzzification function to obtain the left turn probability and the right turn probability corresponding to the turning intention.

[0170] S205: Determine the lane change membership of each lane change intention information in the lane change intention information set based on the fuzzy small membership method and the fuzzy large membership method.

[0171] S206: According to each lane change membership, a fuzzy relationship combination is performed based on a fuzzy and defuzzification function and a fuzzy or defuzzification function to obtain a left lane change probability and a right lane change probability corresponding to the lane change intention.

[0172] S207: Determine a target driving intention based on the left turn probability, the right turn probability, the left lane change probability, and the right lane change probability; wherein the target driving intention is one of going straight, turning, and changing lanes.

[0173] The embodiment of the present invention provides an accurate and quantitative calculation to evaluate the driver's driving intention during driving. Compared with a single variable to measure driving intention, it is relatively stable and reliable. Compared with multi-input neural network model calculation, it can reduce the requirements on the computing power of the domain controller. It is a low-computing-power, low-cost evaluation method.

[0174] The driving intention determination device provided in an embodiment of the present invention is introduced below. The driving intention determination device described below and the driving intention determination method described above can be referenced to each other.

[0175] Please refer to Fig.10 , Fig.10 A schematic diagram of a driving intention determination device provided by an embodiment of the present invention may include:

[0176] The membership determination module 100 is used to determine the turning membership corresponding to each turning intention information and the lane change membership corresponding to each lane change intention information based on a fuzzy membership method; wherein the turning intention information and the lane change intention information include vehicle motion state information and active intention information; the fuzzy membership method is a method for determining the membership corresponding to each intention information based on the first membership information and the second membership information set corresponding to each intention information;

[0177] The turning probability determination module 200 is used to determine the left turn probability and the right turn probability by combining fuzzy relations based on all the turning memberships using a defuzzification function; wherein the defuzzification function is a function for determining the intention probability based on the membership and the weight coefficient;

[0178] A lane change probability determination module 300, configured to determine a left lane change probability and a right lane change probability by combining fuzzy relations based on all lane change memberships using the defuzzification function;

[0179] The driving intention determination module 400 is used to determine the target driving intention based on the left turn probability, the right turn probability, the left lane change probability and the right lane change probability.

[0180] It should be further explained that the above-mentioned membership degree determination module 100 may include:

[0181] A membership determination unit based on a fuzzy small function is used to determine the membership corresponding to the current intention information based on a fuzzy small membership method if the current intention information is the intention information whose value is larger and closer to zero. The input of the fuzzy small membership method is the value x of the current intention information, the first membership parameter M1 and the second membership parameter M2. When the input x is greater than M2, the membership is 0; when the input x is less than M1, the membership is 1; when the input is between M1 and M2, the membership is (M2-x) / (M2-M1), M1 <M2;

[0182] The membership determination unit based on the fuzzy large function is used to determine the membership corresponding to the current intention information based on the fuzzy large membership method if the current intention information is the intention information whose value is larger and closer to one, wherein the input of the fuzzy large membership method is the value y of the current intention information, the third membership information M3 and the fourth membership information M4, when the input y is less than M3, the membership is 0; when y is greater than M4, the membership is 1; when y is between M3 and M4, the membership is (y-M3) / (M4-M3), M3 <M4;

[0183] a turn signal member degree determination unit, for determining, if the current turning intention information is a turn signal switch, that the member degree corresponding to when the turn signal is turned on is 1, and that the member degree corresponding to when the turn signal is turned off is 0;

[0184] The self-vehicle brake pedal membership determination unit is used to determine that if the current steering intention information is the self-vehicle brake pedal, the corresponding membership when the pedal is braking is 1, and the corresponding membership when the pedal is not braking is 0.

[0185] Further, based on the above embodiment, the membership degree determination module 100 may include:

[0186] A first turning membership method determining unit is used to determine the membership corresponding to the current turning intention information based on the fuzzy large membership method if the current turning intention information is any one of the steering wheel angle, the yaw angular velocity of the vehicle, the curvature of the vehicle trajectory, the longitudinal acceleration of the vehicle, the accelerator pedal opening of the vehicle, and the turn signal on time of the vehicle;

[0187] The second turning membership method determination unit is used to distinguish between the lowest level speed, the medium speed and the highest level speed when calculating the membership of the vehicle speed signal if the current turning intention information is the vehicle speed signal of the own vehicle, and determine the membership corresponding to the current turning intention information based on the fuzzy small membership method; wherein the highest level speed is greater than the medium speed, and the medium speed is greater than the lowest level speed.

[0188] Further, based on the above embodiment, the membership degree determination module 100 may include:

[0189] a first lane change membership determination unit, configured to determine the membership corresponding to the current lane change intention information based on the fuzzy large membership method if the current lane change intention information is any one of a steering wheel angle, a turn signal switch signal, an angle of human eye rotation, a rate of change of a distance from the vehicle to a lane boundary, a curvature of the vehicle trajectory, and a turn signal on time;

[0190] The second lane change membership determination unit is used to determine the membership corresponding to the current lane change intention information based on the fuzzy small membership method if the current lane change intention information is the standard deviation of the rate of change of the distance from the vehicle to the lane boundary.

[0191] Further, based on any of the above embodiments, the above turning probability determination module 200 may include:

[0192] a turning probability determination unit, configured to determine the left turn probability and the right turn probability by combining fuzzy relations using a fuzzy and defuzzification function and a fuzzy or defuzzification function based on all turning memberships;

[0193] The inputs of the fuzzy and defuzzified functions are the first membership L1, the second membership L2 and the weight coefficient K1. When L1 is less than or equal to L2, the probability value is K1*L1+(1-K1)*((L1+L2) / 2); when L1 is greater than L2, the probability value is K1*L2+(1-K1)*((L1+L2) / 2);

[0194] The input of the fuzzy or defuzzified function is the first membership L1, the second membership L2 and the weight coefficient K1. When L1 is less than or equal to L2, the probability value is K1*L2+(1-K1)*((L1+L2) / 2); when L1 is greater than L2, the probability value is K1*L1+(1-K1)*((L1+L2) / 2).

[0195] Further, based on the above embodiment, the above turning probability determination module 200 may include:

[0196] an initial turning probability determination unit, configured to perform fuzzy relationship combination on the turning membership corresponding to the first type of turning intention information based on the fuzzy and defuzzification functions to obtain an initial left turn probability and an initial right turn probability;

[0197] The final turning probability determination unit is used to perform fuzzy relationship combination on the turning membership corresponding to the second type of turning intention information based on the fuzzy or defuzzified function, the initial turning probability and the initial right turn probability to obtain the final left turn probability and right turn probability; wherein the second type of turning intention information includes the longitudinal acceleration of the vehicle and the brake pedal of the vehicle, and the first type of turning intention is the turning intention information other than the second type of turning intention information.

[0198] Further, based on any of the above embodiments, the lane change probability determination module 300 may include:

[0199] A judgment unit, used for judging whether there is a change in the lane sequence value;

[0200] A first lane change probability determination unit is used to determine, when there is a change in the lane sequence value, if the lane sequence value switches from the middle lane sequence value to the left lane sequence value, determine that the left lane change probability is 1; if the lane sequence value switches from the middle lane sequence value to the right lane sequence value, determine that the right lane change probability is 1;

[0201] The second lane change probability determination unit is used to determine the step of performing fuzzy relationship combination based on all lane change memberships using the defuzzification function to determine the left lane change probability and the right lane change probability when there is no change in the lane sequence value.

[0202] Further, based on any of the above embodiments, the above driving intention determination device may further include:

[0203] A sequence value determination unit, configured to determine that the left lane change probability is 1 if the self-lane sequence value switches from the middle lane sequence value to the left lane sequence value;

[0204] A 0 probability determination unit, used for determining the left lane change probability and the right lane change probability to be 0 if the lane sequence value of the vehicle does not switch;

[0205] A probability determination unit is used to determine that the probability of right lane change is 1 if the lane sequence value switches from the middle lane sequence value to the right lane sequence value.

[0206] Further, based on any of the above embodiments, the driving intention determination module 400 may include:

[0207] a probability maximum determination unit, used to determine the maximum probability among the left turn probability, the right turn probability, the left lane change probability and the right lane change probability;

[0208] A judgment unit, used to determine whether the maximum probability is greater than a minimum probability threshold;

[0209] A first target driving intention determination unit, configured to determine, when the maximum probability value is greater than the minimum probability threshold value, the intention corresponding to the maximum probability value as the target driving intention;

[0210] The second target driving intention determination unit is used to determine that the target driving intention is going straight when the maximum probability value is not greater than the minimum probability threshold value.

[0211] It should be noted that the order of the modules and units in the above-mentioned driving intention determination device can be changed without affecting the logic.

[0212] The driving intention determination device provided by the embodiment of the present invention may include: a membership determination module 100, which is used to determine the turning membership corresponding to each turning intention information and the lane change membership corresponding to each lane change intention information based on a fuzzy membership method; wherein the turning intention information and the lane change intention information include vehicle motion state information and active intention information; the fuzzy membership method is a method for determining the membership corresponding to each intention information based on the set first membership information and second membership information corresponding to each intention information; a turning probability determination module 200, which is used to determine the left turn probability and the right turn probability by combining fuzzy relations based on all turning memberships using a defuzzification function; wherein the defuzzification function is a function for determining the intention probability based on the membership and the weight coefficient; a lane change probability determination module 300, which is used to determine the left lane change probability and the right lane change probability by combining fuzzy relations based on all lane change memberships using the defuzzification function; and a driving intention determination module 400, which is used to determine the target driving intention based on the left turn probability, the right turn probability, the left lane change probability and the right lane change probability. Compared with the current method of measuring driving intention directly based on a single variable, the present invention determines the membership corresponding to the intention information based on multi-parameter intention information composed of vehicle motion state information and active intention, and finally combines all the memberships based on the defuzzification function to determine the target driving intention, thereby improving the accuracy of driving intention determination.

[0213] A driving intention determination device provided in an embodiment of the present invention is introduced below. The driving intention determination device described below and the driving intention determination method described above can be referenced to each other.

[0214] Please refer to Fig.11 , Fig.11 A schematic diagram of a driving intention determination device provided in an embodiment of the present invention may include:

[0215] A memory 10, used for storing computer programs;

[0216] The processor 20 is used to execute a computer program to implement the above-mentioned driving intention determination method.

[0217] The memory 10 , the processor 20 , and the communication interface 30 all communicate with each other via a communication bus 40 .

[0218] In the embodiment of the present invention, the memory 10 is used to store one or more programs, and the program may include program code, and the program code includes computer operation instructions. In the embodiment of the present invention, the memory 10 may store programs for implementing the following functions:

[0219] Determining the turning membership corresponding to each turning intention information and the lane changing membership corresponding to each lane changing intention information based on a fuzzy membership method; wherein the turning intention information and the lane changing intention information include vehicle motion state information and active intention information; the fuzzy membership method is a method for determining the membership corresponding to each intention information based on a first membership parameter and a second membership parameter set corresponding to each intention information;

[0220] Based on all the turning memberships, the fuzzy relationship combination is performed using a defuzzification function to determine the probability of a left turn and the probability of a right turn; wherein the defuzzification function is a function that determines the intention probability based on the membership and the weight coefficient;

[0221] Based on all lane change memberships, the fuzzy relationship combination is performed using a defuzzification function to determine the probability of changing lanes to the left and the probability of changing lanes to the right;

[0222] The target driving intention is determined based on the left turn probability, the right turn probability, the left lane change probability and the right lane change probability.

[0223] In a possible implementation, the memory 10 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and an application required for at least one function, etc.; the data storage area may store data created during use.

[0224] In addition, the memory 10 may include a read-only memory and a random access memory, and provide instructions and data to the processor. A portion of the memory may also include an NVRAM. The memory stores an operating system and operating instructions, executable modules or data structures, or a subset thereof, or an extended set thereof, wherein the operating instructions may include various operating instructions for implementing various operations. The operating system may include various system programs for implementing various basic tasks and processing hardware-based tasks.

[0225] The processor 20 may be a central processing unit (CPU), an application specific integrated circuit, a digital signal processor, a field programmable gate array or other programmable logic device, a microprocessor or any conventional processor, etc. The processor 20 may call a program stored in the memory 10 .

[0226] The communication interface 30 may be an interface of a communication module, and is used to connect to other devices or systems.

[0227] Of course, it should be noted that Fig.11 The structure shown does not constitute a limitation on the driving intention determination device in the embodiment of the present invention. In actual applications, the driving intention determination device may include Fig.11More or fewer components than shown, or combinations of certain components.

[0228] The following is an introduction to a readable storage medium provided in an embodiment of the present invention. The readable storage medium described below and the driving intention determination method described above can be referenced to each other.

[0229] The present invention also provides a readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the above-mentioned driving intention determination method are implemented.

[0230] The readable storage medium may include: a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0231] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0232] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0233] Finally, it should be noted that, in this article, relationships such as first and second, etc. are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprises" or any other variations are intended to cover non-exclusive inclusion, so that a process, method, article or device that includes a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0234] The above is a detailed introduction to a driving intention determination method, device, equipment and readable storage medium provided by the present invention. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for a person skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A method for determining driving intention, characterized in that: include: Determining the turning membership corresponding to each turning intention information and the lane changing membership corresponding to each lane changing intention information based on a fuzzy membership method; wherein the turning intention information and the lane changing intention information include vehicle motion state information and active intention information; the fuzzy membership method is a method for determining the membership corresponding to each intention information based on the set first membership information and second membership information corresponding to each intention information; Based on all the turning memberships, a defuzzification function is used to combine fuzzy relations to determine the probability of a left turn and the probability of a right turn; wherein the defuzzification function is a function that determines the intention probability based on the membership and the weight coefficient; Based on all lane change memberships, the fuzzy relationship combination is performed using the defuzzification function to determine the probability of a left lane change and the probability of a right lane change; A target driving intention is determined based on the left turn probability, the right turn probability, the left lane change probability, and the right lane change probability.

2. The driving intention determination method according to claim 1, characterized in that: The method of determining the turning membership corresponding to each turning intention information and the lane changing membership corresponding to each lane changing intention information based on the fuzzy membership method includes: If the current intention information is an intention information whose value is larger and whose membership is closer to zero, the membership corresponding to the current intention information is determined based on the fuzzy small membership method; wherein the input of the fuzzy small membership method is the value x of the current intention information, the first membership parameter M1 and the second membership parameter M2; when the input x is greater than M2, the membership is 0; when the input x is less than M1, the membership is 1; when the input is between M1 and M2, the membership is (M2-x) / (M2-M1), M1 <M2; If the current intention information is an intention information whose value is larger and whose membership is closer to one, the membership corresponding to the current intention information is determined based on the fuzzy large membership method; wherein the input of the fuzzy large membership method is the value y of the current intention information, the third membership parameter M3 and the fourth membership parameter M4, when the input y is less than M3, the membership is 0; when y is greater than M4, the membership is 1; when y is between M3 and M4, the membership is (y-M3) / (M4-M3), M3 <M4; If the current turning intention information is a turn signal switch, it is determined that the membership degree corresponding to when the turn signal is turned on is 1, and the membership degree corresponding to when the turn signal is turned off is 0; If the current steering intention information is the brake pedal of the vehicle, it is determined that the corresponding membership when the pedal is braking is 1, and the corresponding membership when the pedal is not braking is 0.

3. The driving intention determination method according to claim 2, characterized in that: The method of determining the turning membership corresponding to each turning intention information and the lane changing membership corresponding to each lane changing intention information based on the fuzzy membership method includes: If the current turning intention information is any one of the steering wheel angle, the yaw rate of the vehicle, the curvature of the vehicle trajectory, the longitudinal acceleration of the vehicle, the accelerator pedal opening of the vehicle, and the duration of the vehicle turn signal on, the membership corresponding to the current turning intention information is determined based on the fuzzy large membership method; If the current turning intention information is the vehicle speed signal of the own vehicle, when calculating the membership of the vehicle speed signal of the own vehicle, the lowest level speed, the medium speed and the highest level speed are distinguished, and the membership corresponding to the current turning intention information is determined based on the fuzzy small membership method; wherein the highest level speed is greater than the medium speed, and the medium speed is greater than the lowest level speed.

4. The driving intention determination method according to claim 2, characterized in that: The method of determining the turning membership corresponding to each turning intention information and the lane changing membership corresponding to each lane changing intention information based on the fuzzy membership method includes: If the current lane change intention information is any one of a steering wheel angle, a turn signal switch signal, an angle of human eye rotation, a rate of change of the distance from the vehicle to the lane boundary, a curvature of the vehicle trajectory, and a turn signal on time, determining the membership corresponding to the current lane change intention information based on the fuzzy large membership method; If the current lane change intention information is the standard deviation of the rate of change of the distance from the vehicle to the lane boundary, the membership corresponding to the current lane change intention information is determined based on the fuzzy small membership method.

5. The driving intention determination method according to any one of claims 1 to 4, characterized in that: The method of combining fuzzy relations based on all the turning memberships by using a defuzzification function to determine the left turn probability and the right turn probability includes: Based on all the turning memberships, the fuzzy relationship combination is performed using a fuzzy and defuzzification function and a fuzzy or defuzzification function to determine the left turn probability and the right turn probability; The inputs of the fuzzy and defuzzified functions are the first membership L1, the second membership L2 and the weight coefficient K1. When L1 is less than or equal to L2, the probability value is K1*L1+(1-K1)*((L1+L2) / 2); when L1 is greater than L2, the probability value is K1*L2+(1-K1)*((L1+L2) / 2); The input of the fuzzy or defuzzified function is the first membership L1, the second membership L2 and the weight coefficient K1. When L1 is less than or equal to L2, the probability value is K1*L2+(1-K1)*((L1+L2) / 2); when L1 is greater than L2, the probability value is K1*L1+(1-K1)*((L1+L2) / 2).

6. The driving intention determination method according to claim 5, characterized in that: The method of combining fuzzy relations based on all the turning memberships by using a defuzzification function to determine the left turn probability and the right turn probability includes: Based on the fuzzy and defuzzification functions, fuzzy relationship combination is performed on the turn membership corresponding to the first type of turn intention information to obtain an initial left turn probability and an initial right turn probability; Based on the fuzzy or defuzzified function, the initial turning probability and the initial right turn probability, the turning membership corresponding to the second type of turning intention information is fuzzily combined to obtain the final left turn probability and right turn probability; wherein the second type of turning intention information includes the longitudinal acceleration of the vehicle and the brake pedal of the vehicle, and the first type of turning intention is the turning intention information other than the second type of turning intention information.

7. The driving intention determination method according to claim 1, characterized in that: Based on all lane change memberships, the fuzzy relationship combination is performed using the defuzzification function to determine the left lane change probability and the right lane change probability, including: Determine whether there is a change in lane sequence value; When there is a change in the lane sequence value, if the self-lane sequence value switches from the middle lane sequence value to the left lane sequence value, the left lane change probability is determined to be 1; if the self-lane sequence value switches from the middle lane sequence value to the right lane sequence value, the right lane change probability is determined to be 1; When there is no change in the lane sequence value, the step of determining the left lane change probability and the right lane change probability by combining fuzzy relations based on all lane change memberships using the defuzzification function is performed.

8. The driving intention determination method according to claim 1, characterized in that: Determining a target driving intention based on the left turn probability, the right turn probability, the left lane change probability, and the right lane change probability includes: Determine the maximum probability among the left turn probability, the right turn probability, the left lane change probability and the right lane change probability; Determining whether the maximum probability is greater than a minimum probability threshold; When the maximum probability value is greater than the minimum probability threshold, determining the intention corresponding to the maximum probability value as the target driving intention; When the maximum probability value is not greater than the minimum probability threshold value, it is determined that the target driving intention is going straight.

9. A driving intention determination device, characterized in that: include: A membership determination module, used for determining a turning membership corresponding to each turning intention information and a lane change membership corresponding to each lane change intention information based on a fuzzy membership method; wherein the turning intention information and the lane change intention information include vehicle motion state information and active intention information; the fuzzy membership method is a method for determining the membership corresponding to each intention information based on first membership information and second membership information set corresponding to each intention information; A turning probability determination module, used to determine the probability of a left turn and the probability of a right turn by combining fuzzy relations based on all turning memberships using a defuzzification function; wherein the defuzzification function is a function for determining the intention probability based on the membership and the weight coefficient; a lane change probability determination module, configured to determine a left lane change probability and a right lane change probability by combining fuzzy relations based on all lane change memberships using the defuzzification function; A driving intention determination module is used to determine a target driving intention based on the left turn probability, the right turn probability, the left lane change probability and the right lane change probability.

10. A driving intention determination device, characterized in that: include: Memory for storing computer programs; A processor, configured to execute the computer program to implement the steps of the driving intention determination method as claimed in any one of claims 1 to 8.

11. A readable storage medium, characterized in that: The readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the driving intention determination method according to any one of claims 1 to 8 are implemented.

Images