Automatic driving comprehensive decision-making evaluation method and device considering multiple driving factors

Abstract

The present application relates to the technical field of driving decision evaluation, and in particular to a method and device for comprehensive decision evaluation of automatic driving considering multiple types of driving factors, wherein the method comprises: collecting multiple parameters of the driving trajectory, and obtaining the position and speed of the surrounding vehicle; based on the pre-established single evaluation index model considering safety, single evaluation index model for timeliness, single evaluation index model for comfort, and single evaluation index model for energy consumption, the safety score, timeliness score, comfort score, and energy consumption score of the driving trajectory are obtained according to multiple parameters, the position and speed of the surrounding vehicle; the manual score of the driving trajectory is obtained, and the manual score is used as the true value of the total score, and the weights of the safety score, timeliness score, comfort score, and energy consumption score are calculated respectively to obtain the comprehensive decision evaluation result of automatic driving. Thus, the problems that the fitted future trajectory is not the optimal trajectory, the single-dimensional model is not comprehensive enough, and the user needs cannot be accurately evaluated are solved.

Description

Automatic driving comprehensive decision-making evaluation method and device considering multiple driving factors Technical Field

[0001] The present application relates to the technical field of driving decision evaluation, and in particular to a method and device for evaluating comprehensive decision making of autonomous driving that takes into account multiple types of driving factors. Background Art

[0002] With the development of autonomous driving technology, autonomous vehicles have greatly improved the safety and efficiency of transportation and attracted widespread attention. Autonomous driving technology mainly includes perception, decision-making, control, map making and other aspects. Decision-making in complex traffic scenarios mainly includes behavior selection and trajectory planning, which is also one of the important challenges of advanced autonomous driving.

[0003] In related technologies, with the breakthrough of deep learning, on the one hand, decision models are established based on neural networks trained with natural driving data sets such as HighD, Nuscenes, and Waymo, and the Euclidean length of the decision trajectory and the real future trajectory is used as the evaluation criterion. On the other hand, rule-based decision models use objective physical metrics to compare performance with benchmarks.

[0004] However, in the related art, the evaluation method that uses the Euclidean length of the decision trajectory and the real future trajectory as the evaluation standard, even the most experienced drivers will make mistakes, and the real trajectory is not the optimal trajectory, and the fitted future trajectory cannot be regarded as a suitable indicator for decision evaluation. In most cases, the rule-based decision model selects a specific single performance indicator for evaluation and decision-making. It can be understood that it only considers one aspect of driving and ignores other factors, such as time efficiency, energy consumption, safety and comfort. It is not comprehensive and needs to be improved. Summary of the invention

[0005] The present application provides an automatic driving comprehensive decision-making evaluation method and device that considers multiple types of driving factors, so as to solve the problems in related technologies, that the fitted future trajectory is not the optimal trajectory, and that the single-dimensional model in most cases only considers one aspect of driving and ignores other factors, which is not comprehensive enough and cannot accurately evaluate user needs.

[0006] The first aspect of the present application provides an automatic driving comprehensive decision-making evaluation method that takes into account multiple types of driving factors, including the following steps: collecting multiple parameters of a driving trajectory, and obtaining the position and speed of a surrounding vehicle; based on a pre-established single evaluation index model considering safety, a single evaluation index model for timeliness, a single evaluation index model for comfort, and a single evaluation index model for energy consumption, respectively obtaining the safety score, timeliness score, comfort score, and energy consumption score of the driving trajectory according to the multiple parameters and the position and speed of the surrounding vehicle; and obtaining a manual score of the driving trajectory, and taking the manual score as the true value of the total score, respectively calculating the weights of the safety score, the timeliness score, the comfort score, and the energy consumption score, so as to obtain an automatic driving comprehensive decision-making evaluation result.

[0007] Optionally, in one embodiment of the present application,

[0008] The expression of the single safety evaluation index model can be:

[0009]

[0010] Among them, K1, K2, K3 and G are the coefficients that need to be determined; M eq is the equivalent mass; r eq is the equivalent distance to reflect the sensitivity of the safety distance in different directions; r is the relative velocity; θ is the angle between the relative velocity and the distance vector.

[0011] The expression of the single evaluation index model of timeliness can be:

[0012]

[0013] Where v is the longitudinal velocity of the vehicle; v limit Speed ​​limit for the road.

[0014] The expression of the comfort single evaluation index model can be:

[0015] s comf =ω·v+k·j 2 +L turn

[0016] Where, ω is the angular velocity; k is the sensitivity coefficient; j is the rate of change of acceleration; L turn A penalty for comfort when reversing.

[0017] The expression of the energy consumption single evaluation index model can be:

[0018]

[0019] Among them, P e is the total energy consumption of vehicle movement; Δt is the time interval; d is the total distance.

[0020] Optionally, in one embodiment of the present application, the expression of the autonomous driving comprehensive decision evaluation result may be:

[0021] s DEI =Cr(w1s risk +w2s eff +w s s commf +w4s eng +b)

[0022] Where Cr is the collision detection coefficient; w i is the corresponding weight; b is the offset.

[0023] Optionally, in one embodiment of the present application, the manual score is used as the true value of the total score, and the weights of the safety score, the timeliness score, the comfort score and the energy consumption score are calculated respectively, including: performing linear fitting based on the manual score to calculate the weights of the safety score, the timeliness score, the comfort score and the energy consumption score.

[0024] Optionally, in one embodiment of the present application, the multiple parameters may be, but are not limited to, multiple parameters including lateral and longitudinal speeds, acceleration, heading angle, and vehicle position.

[0025] The second aspect of the present application provides an automatic driving comprehensive decision-making evaluation device that takes into account multiple types of driving factors, including: a collection module, which is used to collect multiple parameters of a driving trajectory and obtain the position and speed of a surrounding vehicle; a generation module, which is used to obtain the safety score, timeliness score, comfort score and energy consumption score of the driving trajectory based on a pre-established single evaluation index model considering safety, a single evaluation index model considering timeliness, a single evaluation index model considering comfort and a single evaluation index model considering energy consumption according to the multiple parameters; and a decision module, which is used to obtain a manual score of the driving trajectory, and use the manual score as the true value of the total score to calculate the weights of the safety score, the timeliness score, the comfort score and the energy consumption score, respectively, to obtain an automatic driving comprehensive decision-making evaluation result.

[0026] Optionally, in one embodiment of the present application,

[0027] The expression of the single safety evaluation index model can be:

[0028]

[0029] Among them, K1, K2, K3 and G are the coefficients that need to be determined; M eq is the equivalent mass; r eq is the equivalent distance to reflect the sensitivity of the safety distance in different directions; r is the relative velocity; θ is the angle between the relative velocity and the distance vector.

[0030] The expression of the single evaluation index model of timeliness can be:

[0031]

[0032] Where v is the longitudinal velocity of the vehicle; v limit Speed ​​limit for the road.

[0033] The expression of the comfort single evaluation index model can be:

[0034] s comf =ω·v+k·j 2 +L turn

[0035] Where, ω is the angular velocity; k is the sensitivity coefficient; j is the rate of change of acceleration; L turn A penalty for comfort when reversing.

[0036] The expression of the energy consumption single evaluation index model can be:

[0037]

[0038] Among them, P e is the total energy consumption of vehicle movement; Δt is the time interval; d is the total distance.

[0039] Optionally, in one embodiment of the present application, the expression of the autonomous driving comprehensive decision evaluation result may be:

[0040] s DEI =Cr(w1s risk +w2s eff +w3s comf +w4s eng +b)

[0041] Where Cr is the collision detection coefficient; w i is the corresponding weight; b is the offset.

[0042] Optionally, in one embodiment of the present application, the decision module includes: a fitting unit, used to perform linear fitting based on the manual score to calculate the weights of the safety score, the timeliness score, the comfort score and the energy consumption score.

[0043] Optionally, in one embodiment of the present application, the multiple parameters may be, but are not limited to, multiple parameters including lateral and longitudinal speeds, acceleration, heading angle, and vehicle position.

[0044] The third aspect of the present application provides a vehicle, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the comprehensive decision-making evaluation method for autonomous driving that considers multiple types of driving factors as described in the above embodiments.

[0045] The fourth aspect embodiment of the present application provides a computer-readable storage medium, which stores a computer program. When the program is executed by a processor, it implements the above-mentioned automatic driving comprehensive decision-making evaluation method considering multiple types of driving factors.

[0046] The embodiment of the present application can establish independent evaluation index models for the four evaluation indicators of safety, timeliness, comfort and energy consumption, and determine the weights of the four scores by fitting with manual scoring, so as to obtain the comprehensive decision evaluation results of autonomous driving, guide the evaluation of future autonomous driving decisions, assist in the development of subsequent high-level autonomous driving algorithms, and be widely applicable to various traffic scenarios. The evaluation decision is more comprehensive, the accuracy of autonomous driving decisions is improved, and the needs of users are better met. In this way, the problems in the related technology that the fitted future trajectory is not the optimal trajectory, and the single-dimensional model only considers one aspect of driving in most cases, ignores other factors, is not comprehensive enough, and cannot accurately evaluate user needs are solved.

[0047] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0049] FIG1 is a flow chart of an automatic driving comprehensive decision-making evaluation method considering multiple types of driving factors provided according to an embodiment of the present application;

[0050] FIG2 is a schematic diagram of simulation of 12 scenarios provided according to a specific embodiment of the present application;

[0051] FIG3 is a test flow chart of an automatic driving comprehensive decision-making evaluation method considering multiple types of driving factors provided according to a specific embodiment of the present application;

[0052] FIG4 is a block diagram of an automatic driving comprehensive decision-making evaluation device considering multiple types of driving factors according to an embodiment of the present application;

[0053] FIG5 is a schematic diagram of the structure of a vehicle provided according to an embodiment of the present application. DETAILED DESCRIPTION

[0054] Embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0055] The following describes the automatic driving comprehensive decision evaluation method considering multiple types of driving factors in the embodiment of the present application with reference to the accompanying drawings. In view of the problem that the fitted future trajectory mentioned in the above background technology is not the optimal trajectory, and the single-dimensional model only considers one aspect of driving in most cases, ignores other factors, is not comprehensive enough, and cannot accurately evaluate user needs, the present application provides an automatic driving comprehensive decision evaluation method considering multiple types of driving factors. In this method, independent evaluation index models can be established for the four evaluation indicators of safety, timeliness, comfort and energy consumption, and the weights of the four scores can be determined by manual scoring fitting to obtain the automatic driving comprehensive decision evaluation results, guide the future automatic driving decision evaluation, assist the subsequent high-level automatic driving algorithm development, and can be widely applied to various traffic scenarios. Its evaluation decision is more comprehensive, can improve the accuracy of automatic driving decisions, and can better meet user needs. Therefore, the problems in the related technology that the fitted future trajectory is not the optimal trajectory, and the single-dimensional model only considers one aspect of driving in most cases, ignores other factors, is not comprehensive enough, and cannot accurately evaluate user needs are solved.

[0056] Specifically, Figure 1 is a flowchart of an automatic driving comprehensive decision-making evaluation method considering multiple types of driving factors provided according to an embodiment of the present application.

[0057] As shown in FIG1 , the automatic driving comprehensive decision evaluation method considering multiple types of driving factors includes the following steps:

[0058] In step S101, multiple parameters of the driving trajectory are collected, and the position and speed of the surrounding vehicle are obtained. The multiple parameters may include but are not limited to multiple parameters including lateral and longitudinal speeds, acceleration, heading angle, and vehicle position.

[0059] It is understandable that multiple parameters of the driving trajectory can be collected using high-precision maps, sensors, and GPS (Global Positioning System) equipment, and this application does not impose specific restrictions. Among them, the driving trajectory can be the driving trajectory of the driving vehicle, or the driving trajectory of the driver. Furthermore, the driving trajectory of the driving vehicle can be but is not limited to the vehicle's forward, backward, left turn, and right turn, etc., and the driver's driving trajectory can be but is not limited to the driver's head movements, such as not looking straight ahead, closing eyes, driving while talking on the phone, etc., trunk movements, such as bending over to pick up items during driving, tidying up items on the co-pilot sideways, etc., upper limb movements, such as smoking, etc., this application does not impose specific restrictions.

[0060] In addition, the surrounding vehicles can be all vehicles within 3 meters of the driving vehicle, or vehicles within one lane of the driving vehicle. The specific setting method can be set by technical personnel in this field according to actual conditions, and this application does not make any specific restrictions.

[0061] As a possible implementation method, the embodiment of the present application can collect the driving trajectory of the driving vehicle and the driver in real time, and obtain the position and speed of the surrounding vehicles.

[0062] For example, the embodiments of the present application can use a camera installed in front of the driver's seat to collect the driver's driving trajectory, such as whether his eyes are facing forward, whether he has his eyes closed, whether he is bending over, etc., use the on-board GPS to collect the driving trajectory of the vehicle, such as whether the road ahead is straight, etc., and use millimeter-wave radars and cameras installed around the vehicle to collect the environment around the vehicle, such as the position and speed of surrounding vehicles.

[0063] In step S102, based on the pre-established single evaluation index model considering safety, timeliness, comfort and energy consumption, the safety score, timeliness score, comfort score and energy consumption score of the driving trajectory are obtained according to multiple parameters, the position and speed of the surrounding vehicle.

[0064] The expression of the single safety evaluation index model can be:

[0065]

[0066] Among them, K1, K2, K3 and G are the coefficients that need to be determined; M eq is the equivalent mass; r eq is the equivalent distance to reflect the sensitivity of the safety distance in different directions; r is the relative velocity; θ is the angle between the relative velocity and the distance vector.

[0067] The expression of the single evaluation index model of timeliness can be:

[0068]

[0069] Where v is the longitudinal velocity of the vehicle; v limit Speed ​​limit for the road.

[0070] The expression of the comfort single evaluation index model can be:

[0071] s comf =ω·v+k·j 2 +L turn

[0072] Where, ω is the angular velocity; k is the sensitivity coefficient; j is the rate of change of acceleration; L turn A penalty for comfort when reversing.

[0073] The expression of the energy consumption single evaluation index model can be:

[0074]

[0075] Among them, P e is the total energy consumption of vehicle movement; Δt is the time interval; d is the total distance.

[0076] As a possible implementation method, the embodiment of the present application can establish single evaluation index models based on safety, timeliness, comfort and energy consumption respectively, and combine multiple parameters, the position and speed of the surrounding vehicle to obtain the safety score, timeliness score, comfort score and energy consumption score of the driving trajectory.

[0077] Specifically, a single safety evaluation index model is established. The expression of the single safety evaluation index model can be:

[0078]

[0079] Among them, K1, K2, K3 and G are the coefficients that need to be determined; M eq is the equivalent mass; r eq is the equivalent distance to reflect the sensitivity of the safety distance in different directions; r is the relative velocity; θ is the angle between the relative velocity and the distance vector.

[0080] Among them, the first term represents the potential field and the second term represents the kinetic energy field.

[0081] Furthermore, eq The expression can be:

[0082] r eq =(r n 2 +k e ·r t 2 ) 1 / 2

[0083] Among them, r n is the longitudinal distance between the other vehicle and the vehicle; r t is the lateral distance; k e It is a proportional coefficient that reflects the different sensitivities of risk in the horizontal and vertical directions.

[0084] M eq The expression can be:

[0085] M eq =M·(av e b +c)

[0086] Where M is the mass of the other vehicle, reflecting the consequences of a collision with it; a, b, c are constant coefficients; v e is the longitudinal speed.

[0087] In the embodiment of the present application, in order to improve the efficiency of the evaluation algorithm, only traffic participants within the effective operating range may be considered, and other vehicles are considered only when the equivalent distance between the vehicle and the vehicle is lower than a certain threshold. The setting of the certain threshold can be set by a person skilled in the art according to actual conditions, and this application does not impose any specific restrictions.

[0088] A single evaluation index model of timeliness is established. Timeliness can be measured by the ratio of current speed to expected speed, and the expected speed can be taken as the speed limit of the current road. In particular, a vehicle will be punished when it exceeds the speed limit. Generally, speeding less than 10% is tolerable, and more speeding will be punished incrementally. Therefore, the expression of the single evaluation index model of timeliness can be:

[0089]

[0090] Where v is the longitudinal velocity of the vehicle; v limit Speed ​​limit for the road.

[0091] A single evaluation index model for comfort is established. Factors that affect comfort may include vehicle bumps, rapid acceleration and deceleration, turns at intersections, etc., and penalties for reversing. Therefore, the expression of the single evaluation index model for comfort can be:

[0092] s comf =ω·v+k·j 2 +L turn

[0093] Where, ω is the angular velocity; k is the sensitivity coefficient; j is the rate of change of acceleration; L turn A penalty for comfort when reversing.

[0094] Among them, the first item uses centrifugal force to represent the impact of vehicle turning on comfort, the second item reflects the impact of sudden acceleration and deceleration on comfort, and the last item is the penalty for reversing on comfort.

[0095] Establish a single evaluation index model for energy consumption. Assume that the power transmission efficiency remains constant under different conditions, that is, the influence of vehicle type and driving environment is eliminated. At this time, the total energy consumption of vehicle movement can be calculated by the energy consumed by the vehicle during driving. Among them, the total energy consumption can be composed of acceleration, wind resistance, slope resistance and road resistance. In the embodiment of the present application, the slope resistance can be ignored, and it is assumed that the vehicle is rolling friction. The energy consumption of road resistance is proportional to the speed, the air resistance is proportional to the square of the speed, the energy consumption is proportional to the cube of the speed, and the acceleration resistance is proportional to the speed and acceleration. Therefore, the expression of the energy consumption model can be:

[0096]

[0097]

[0098]

[0099]

[0100] Among them, P f is the road friction term; P w is the wind resistance term; P j is the acceleration term; P e is the total energy consumption of vehicle movement; η T is the mechanical transmission efficiency; G is the gravity of the vehicle; f is the rolling resistance coefficient; u a is the vehicle speed in km / h; C D is the air resistance coefficient, A is the frontal area of ​​the vehicle. δ is the conversion coefficient of the rotational mass; u is the vehicle speed in m / s.

[0101] However, in the actual operation of the vehicle, the energy consumption considers the fuel consumption of the entire trajectory. The time average will lead to a reduction in the fuel consumption penalty when the vehicle speed is low. Therefore, the embodiment of the present application can be measured by distance average, and the expression of the energy consumption single evaluation index model can be:

[0102]

[0103] Among them, P e is the total energy consumption of vehicle movement; Δt is the time interval; d is the total distance.

[0104] In step S103, the manual score of the driving trajectory is obtained, and the manual score is used as the total score truth value, and the weights of the safety score, timeliness score, comfort score and energy consumption score are calculated respectively to obtain the automatic driving comprehensive decision evaluation result. Among them, the expression of the automatic driving comprehensive decision evaluation result can be:

[0105] S DEI =Cr(w1s risk +w2s eff +w3s comf +w4S eng +b)

[0106] Where Cr is the collision detection coefficient; w i is the corresponding weight; b is the offset.

[0107] It is understandable that the embodiments of the present application can use linear fitting to calculate the weights of the safety score, timeliness score, comfort score and energy consumption score, or other methods, and the present application does not make specific limitations.

[0108] As a possible implementation method, the embodiment of the present application can obtain a manual score of the driving trajectory through a third-party expert scoring method, and calculate the weights of the safety score, timeliness score, comfort score and energy consumption score to obtain an autonomous driving comprehensive decision-making evaluation result.

[0109] Specifically, the embodiment of the present application can record a driving video from the driver's first-person perspective, and invite multiple experts to score the driving video. The average score of each driving trajectory is taken as the manual score of the trajectory. The manual score is taken as the true value of the total score, and the weights of the safety score, timeliness score, comfort score and energy consumption score are calculated to obtain the comprehensive decision-making evaluation result of the autonomous driving. The expression of the comprehensive decision-making evaluation result of the autonomous driving can be:

[0110] S DEI =Cr(w1s risk +w2S eff +w3s comf +w4s eng +b)

[0111] Where Cr is the collision detection coefficient; w i is the corresponding weight; b is the offset.

[0112] Optionally, in one embodiment of the present application, the manual score is used as the true value of the total score, and the weights of the safety score, timeliness score, comfort score and energy consumption score are calculated respectively, including: performing linear fitting based on the manual score to calculate the weights of the safety score, timeliness score, comfort score and energy consumption score.

[0113] In the actual implementation process, the embodiment of the present application can perform linear fitting based on the results of manual scoring to calculate the weights of the safety score, timeliness score, comfort score and energy consumption score.

[0114] Specifically, in the embodiment of the present application, the values ​​of the four indicator scores of safety score, timeliness score, comfort score and energy consumption score are calculated directly according to the physical meaning, and there may be large differences in their magnitudes. Therefore, the embodiment of the present application can standardize the four individual evaluation scores of safety, timeliness, comfort and energy consumption, and express the standardized safety risk, efficiency, comfort and energy consumption scores as s risk 、s eff 、s comf 、s eng , and bring it into the expression of the comprehensive decision-making evaluation result of autonomous driving written above.

[0115] Furthermore, in the embodiment of the present application, if a safety accident such as a collision occurs, other performances including high efficiency and comfort become meaningless. Therefore, Cr can be used to correct the overall model, and its expression can be:

[0116]

[0117] In addition, the embodiment of the present application also performs linear fitting with the manual scores obtained by the expert scoring method to obtain the value of each weight, wherein the loss function L r is the mean square error, which can be expressed as:

[0118]

[0119] Finally, the present embodiment substitutes the obtained weights and the four standardized items into s DEI In this way, the calculation of the comprehensive decision-making evaluation results of autonomous driving is completed.

[0120] In addition, the comprehensive decision-making evaluation method for autonomous driving considering multiple types of driving factors proposed in this application is introduced in detail in combination with a specific embodiment.

[0121] Among them, Figure 3 is a test flowchart of an automatic driving comprehensive decision-making evaluation method considering multiple types of driving factors provided according to a specific embodiment of the present application.

[0122] Specifically, the embodiment of the present application tests the proposed autonomous driving comprehensive decision evaluation method and compares it with the existing single evaluation model. After comparison, it is found that the embodiment of the present application has an advantage in MAE (Mean Absolute Error), where the expression of MAE can be:

[0123]

[0124] The specific test process may be shown in FIG3 , and the specific steps may be:

[0125] Step S1: Set up a simulation scenario and collect driving data.

[0126] Specifically, the embodiment of the present application can be based on the driving simulator of the SimPRIME model produced by the Korean INNOSIMULATION company, which is a three-degree-of-freedom simulation platform that can simulate the three motion states of the vehicle, namely vertical, roll and pitch, and collect and test 180 driving data of 15 drivers of different ages, driving experience and different driving styles in 12 different scenarios. Among them, the schematic diagram of the 12 scenarios can be shown in Figure 2. Among them, the white box is the car, other vehicles, pedestrians and bicycles are other participants, the curve represents the driver's driving task, and the straight line represents the default trajectory of other participants, with speed annotations. It can be understood that in the simulation software of the embodiment of the present application, other participants have simple intelligence and can react according to the surrounding behavior.

[0127] Step S2: Recording a driving video.

[0128] Specifically, in terms of recording driving, the embodiment of the present application uses the Tobii Pro Glasses3 eye tracker to obtain the driver's first-person perspective. Furthermore, after recording the video, the embodiment of the present application invited 12 volunteers to score the trajectory, and the scoring was based on a percentage system with 5 points for each level. Among them, a score of 90 represents excellent driving and a score of 60 represents acceptable driving. In addition, volunteers can also continuously evaluate activities in the same type of scenes to produce intuitive comparisons, ensure the relative rationality of the scores, and use the average of the 12 scores as the final manual score for a scene.

[0129] Step S3: Test and verify.

[0130] Specifically, in the embodiment of the present application, 90% of the data may be used for training, and 10% may be randomly selected for verification to avoid overfitting. The fitting result may be shown in Table 1.

[0131] Table 1 Model fitting results

[0132] Variable fitting results w10.21w20.29w30.32w40.05b10.00 surface

[0133] Furthermore, the embodiment of the present application performs an error test on the validation set, and the results are shown in Table 2. Compared with the single factor evaluation model, the comprehensive evaluation model proposed in the embodiment of the present application has a more accurate score, and the MAE is reduced by 55.82%, 45.74%, 51.52% and 42.81% respectively.

[0134] Table 2 Evaluation model comparison and verification results

[0135] Evaluation method MAE Max Error Safety 11.52 22.40 Timeliness 9.38 33.85 Comfort 10.50 19.47 Energy consumption 8.90 23.33 Application 5.09 14.95 surface

[0136] According to the automatic driving comprehensive decision evaluation method considering multiple driving factors proposed in the embodiment of the present application, an independent evaluation index model can be established for the four evaluation indicators of safety, timeliness, comfort and energy consumption, and the weights of the four scores can be determined by fitting with manual scoring to obtain the automatic driving comprehensive decision evaluation result, and verified by experiments. The automatic driving comprehensive decision evaluation method considering multiple driving factors proposed in the embodiment of the present application is more accurate in scoring, which can better guide the future automatic driving decision evaluation, more effectively assist the subsequent high-level automatic driving algorithm development, and is widely applicable to various traffic scenarios. The evaluation decision is more comprehensive, which can improve the accuracy of automatic driving decision and better meet the diverse needs of users. Therefore, the problems in the related technology that the fitted future trajectory is not the optimal trajectory, and the single-dimensional model only considers one aspect of driving in most cases, ignores other factors, is not comprehensive enough, and cannot accurately evaluate user needs are solved.

[0137] Next, an automatic driving comprehensive decision-making evaluation device considering multiple types of driving factors proposed in an embodiment of the present application is described with reference to the accompanying drawings.

[0138] FIG4 is a block diagram of an automatic driving comprehensive decision-making evaluation device that takes into account multiple types of driving factors according to an embodiment of the present application.

[0139] As shown in FIG. 4 , the automatic driving comprehensive decision-making evaluation device 10 considering multiple types of driving factors includes: a collection module 100 , a generation module 200 and a decision module 300 .

[0140] Specifically, the acquisition module 100 is used to collect multiple parameters of the driving trajectory and obtain the position and speed of the surrounding vehicle.

[0141] The generation module 200 is used to obtain the safety score, timeliness score, comfort score and energy consumption score of the driving trajectory according to multiple parameters, the position and speed of the surrounding vehicle based on the pre-established single evaluation index model considering safety, timeliness score, comfort score and energy consumption score.

[0142] The decision module 300 is used to obtain the manual score of the driving trajectory, and use the manual score as the true value of the total score to calculate the weights of the safety score, timeliness score, comfort score and energy consumption score respectively, so as to obtain the comprehensive decision evaluation result of autonomous driving.

[0143] Optionally, in one embodiment of the present application,

[0144] The expression of the single safety evaluation index model can be:

[0145]

[0146] Among them, K1, K2, K3 and G are the coefficients that need to be determined; M eq is the equivalent mass; r eq is the equivalent distance to reflect the sensitivity of the safety distance in different directions; r is the relative velocity; θ is the angle between the relative velocity and the distance vector.

[0147] The expression of the single evaluation index model of timeliness can be:

[0148]

[0149] Where v is the longitudinal velocity of the vehicle; v limit Speed ​​limit for the road.

[0150] The expression of the comfort single evaluation index model can be:

[0151] s comf =ω·v+k·j 2 +L turn

[0152] Where, ω is the angular velocity; k is the sensitivity coefficient; j is the rate of change of acceleration; L turn A penalty for comfort when reversing.

[0153] The expression of the energy consumption single evaluation index model can be:

[0154]

[0155] Among them, P e is the total energy consumption of vehicle movement; Δt is the time interval; d is the total distance.

[0156] Optionally, in one embodiment of the present application, the expression of the autonomous driving comprehensive decision evaluation result may be:

[0157] s DEI =Cr(w1s risk +w2s eff +w3s comf +w4s eng +b)

[0158] Where Cr is the collision detection coefficient; w i is the corresponding weight; b is the offset.

[0159] Optionally, in one embodiment of the present application, the decision module 300 includes: a fitting unit.

[0160] Among them, the fitting unit is used to perform linear fitting based on the manual score and calculate the weights of the safety score, timeliness score, comfort score and energy consumption score.

[0161] Optionally, in one embodiment of the present application, the multiple parameters may be, but are not limited to, multiple parameters including lateral and longitudinal speeds, acceleration, heading angle, and vehicle position.

[0162] It should be noted that the aforementioned explanation of the embodiment of the comprehensive decision-making evaluation method for autonomous driving considering multiple types of driving factors is also applicable to the comprehensive decision-making evaluation device for autonomous driving considering multiple types of driving factors of this embodiment, and will not be repeated here.

[0163] According to the automatic driving comprehensive decision evaluation device considering multiple types of driving factors proposed in the embodiment of the present application, it is possible to establish independent evaluation index models for the four evaluation indicators of safety, timeliness, comfort and energy consumption, and determine the weights of the four scores by fitting with manual scoring, to obtain the automatic driving comprehensive decision evaluation results, and verify through experiments. The automatic driving comprehensive decision evaluation method considering multiple types of driving factors proposed in the embodiment of the present application is more accurate in scoring, which can better guide the future automatic driving decision evaluation, more effectively assist the subsequent high-level automatic driving algorithm development, and is widely applicable to various traffic scenarios. The evaluation decision is more comprehensive, which can improve the accuracy of automatic driving decision and better meet the diverse needs of users. Therefore, it solves the problems in the related technology that the fitted future trajectory is not the optimal trajectory, and the single-dimensional model only considers one aspect of driving in most cases, ignores other factors, is not comprehensive enough, and cannot accurately evaluate user needs.

[0164] FIG5 is a schematic diagram of the structure of a vehicle provided according to an embodiment of the present application. The vehicle may include:

[0165] A memory 501 , a processor 502 , and a computer program stored in the memory 501 and executable on the processor 502 .

[0166] When the processor 502 executes the program, the automatic driving comprehensive decision-making evaluation method considering multiple types of driving factors provided in the above embodiment is implemented.

[0167] Furthermore, the vehicle also includes:

[0168] The communication interface 503 is used for communication between the memory 501 and the processor 502 .

[0169] The memory 501 is used to store computer programs that can be executed on the processor 502 .

[0170] The memory 501 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0171] If the memory 501, the processor 502 and the communication interface 503 are implemented independently, the communication interface 503, the memory 501 and the processor 502 can be connected to each other through a bus and communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one thick line is used in FIG5, but it does not mean that there is only one bus or one type of bus.

[0172] Optionally, in a specific implementation, if the memory 501, the processor 502 and the communication interface 503 are integrated on a chip, the memory 501, the processor 502 and the communication interface 503 can communicate with each other through an internal interface.

[0173] The processor 502 may be a central processing unit (CPU), or an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application.

[0174] An embodiment of the present application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-mentioned comprehensive decision-making evaluation method for autonomous driving that takes into account multiple types of driving factors.

[0175] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or N embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0176] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the features. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0177] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code comprising one or N executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present application belong.

[0178] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purpose of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or N wirings (electronic devices), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically by optically scanning the paper or other medium and then editing, interpreting or processing in other suitable ways as necessary and then storing it in a computer memory.

[0179] It should be understood that the various parts of the present application can be implemented by hardware, software, firmware or a combination thereof. In the above embodiment, the N steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. If implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0180] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0181] In addition, each functional unit in each embodiment of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0182] The storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, etc. Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present application. A person of ordinary skill in the art may change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A comprehensive decision-making evaluation method for autonomous driving considering multiple types of driving factors, characterized in that: The method comprises the following steps: collecting multiple parameters of the driving trajectory and obtaining the position and speed of the surrounding vehicle; based on the pre-established single evaluation index model considering safety, single evaluation index model considering timeliness, single evaluation index model considering comfort and single evaluation index model considering energy consumption, respectively obtaining the safety score, timeliness score, comfort score and energy consumption score of the driving trajectory according to the multiple parameters, the position and speed of the surrounding vehicle; and obtaining the manual score of the driving trajectory, and taking the manual score as the true value of the total score, respectively calculating the weights of the safety score, the timeliness score, the comfort score and the energy consumption score, so as to obtain the comprehensive decision-making evaluation result of the automatic driving; wherein, the expression of the single evaluation index model considering safety is: Among them, K1, K2, K3 and G are the coefficients that need to be determined; M eq is the equivalent mass; r eq is the equivalent distance to reflect the sensitivity of the safety distance in different directions; r is the relative speed; θ is the angle between the relative speed and the distance vector; the expression of the single evaluation index model of timeliness is: Where v is the longitudinal velocity of the vehicle; v limit is the road speed limit; the expression of the comfort single evaluation index model is: s comf =ω·v+k·j 2 +L turn Where, ω is the angular velocity; k is the sensitivity coefficient; j is the rate of change of acceleration; L turn is the penalty for reversing on comfort; the expression of the energy consumption single evaluation index model is: Among them, P e is the total energy consumption of vehicle movement; Δt is the time interval; d is the total distance; the calculation formula of the comprehensive decision evaluation result of autonomous driving is: s DEI =Cr(w1s risk +w2s eff +w3s comf +w4s eng +b) Where Cr is the collision detection coefficient; w i is the corresponding weight; b is the offset.

2. The method according to claim 1, characterized in that: The method of taking the manual score as the true value of the total score and respectively calculating the weights of the safety score, the timeliness score, the comfort score and the energy consumption score includes: performing linear fitting based on the manual score to calculate the weights of the safety score, the timeliness score, the comfort score and the energy consumption score.

3. The method according to claim 1, characterized in that The multiple parameters include multiple parameters of lateral and longitudinal speed, acceleration, heading angle and vehicle position.

4. An automatic driving comprehensive decision-making evaluation device considering multiple types of driving factors, characterized in that: include: The acquisition module is used to collect multiple parameters of the driving trajectory and obtain the position and speed of the surrounding vehicle; A generation module, for obtaining the safety score, timeliness score, comfort score and energy consumption score of the driving trajectory respectively according to the plurality of parameters, the position and speed of the surrounding vehicle based on the pre-established single evaluation index model considering safety, timeliness score, comfort score and energy consumption score; and a decision-making module, which is used to obtain the manual score of the driving trajectory, and use the manual score as the true value of the total score to calculate the weights of the safety score, the timeliness score, the comfort score and the energy consumption score, so as to obtain the comprehensive decision-making evaluation result of the autonomous driving; wherein the expression of the single evaluation index model considering safety is: Among them, K1, K2, K3 and G are the coefficients that need to be determined; M eq is the equivalent mass; r eq is the equivalent distance to reflect the sensitivity of the safety distance in different directions; r is the relative speed; θ is the angle between the relative speed and the distance vector; the expression of the single evaluation index model of timeliness is: Where v is the longitudinal velocity of the vehicle; v limit is the road speed limit; the expression of the comfort single evaluation index model is: s comf =ω·v+k·j 2 +L turn Among them, ← is the angular velocity; k is the sensitivity coefficient; j is the rate of change of acceleration; L turn is the penalty for reversing on comfort; the expression of the energy consumption single evaluation index model is: Among them, P e is the total energy consumption of vehicle movement; Δt is the time interval; d is the total distance; the calculation formula of the comprehensive decision evaluation result of autonomous driving is: s DEI =Cr(w1s risk +w2s eff +w3s comf +w4s eng +b) Where Cr is the collision detection coefficient; w i is the corresponding weight; b is the offset.

5. A vehicle, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the automatic driving comprehensive decision-making evaluation method considering multiple types of driving factors as described in any one of claims 1 to 3.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that: The program is executed by a processor to implement an automatic driving comprehensive decision-making evaluation method taking into account multiple types of driving factors as described in any one of claims 1 to 3.

Images