Method, device and electronic equipment for determining optimization degree of automatic driving algorithm

Abstract

The embodiment of the application provides a kind of automatic driving algorithm optimization degree determination method, device and electronic equipment, can be determined according to the driving data of test vehicle based on the first automatic driving algorithm on the road, determine multiple first sample pairs, wherein the first sample pair includes first sample standard entropy and the average takeover times corresponding to first sample standard entropy, the first linear equation can be obtained by regression processing to multiple first sample pairs, then according to the driving data of test vehicle based on the second automatic driving algorithm on the road, determine multiple second sample pairs, wherein the second sample pair includes second sample standard entropy and the average takeover times corresponding to second sample standard entropy, according to multiple second sample pairs and first linear equation, the optimization degree of second automatic driving algorithm compared with first automatic driving algorithm can be determined. By quantifying automatic driving algorithm, the optimization degree of automatic driving algorithm can be evaluated.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to a method, apparatus and electronic device for determining the optimization degree of an autonomous driving algorithm. Background Technology

[0002] With the rapid development of technology and the arrival of the era of intelligent vehicles, autonomous driving technology will occupy an extremely important position in future intelligent transportation systems. Autonomous driving technology benefits society, drivers, and pedestrians alike. Even with interference from other traffic accident rates, the use of autonomous driving technology can steadily reduce the overall traffic accident rate.

[0003] Autonomous driving algorithms are one of the most important parts of autonomous driving technology. In order to improve autonomous driving technology, it is often necessary to continuously improve autonomous driving algorithms. However, the optimization degree of autonomous driving algorithms cannot be quantitatively evaluated. Summary of the Invention

[0004] To address the problems in the prior art, this application provides a method, apparatus, and electronic device for determining the optimization degree of an autonomous driving algorithm, which can quantitatively evaluate the optimization degree of an autonomous driving algorithm.

[0005] In a first aspect, embodiments of this application provide a method for determining the optimization degree of an autonomous driving algorithm, the method comprising:

[0006] Based on the driving data of the test vehicle on the road using the first autonomous driving algorithm, multiple first sample pairs are determined; the first sample pair includes the first sample standard entropy and the average number of interventions corresponding to the first sample standard entropy.

[0007] Regression processing is performed on the multiple first sample pairs to obtain the first linear equation;

[0008] Based on the driving data of the test vehicle on the road using the second autonomous driving algorithm, multiple second sample pairs are determined; the second sample pair includes the average number of interventions corresponding to the second sample standard entropy and the second sample standard entropy.

[0009] Based on the plurality of second sample pairs and the first linear equation, the optimization degree of the second autonomous driving algorithm compared to the first autonomous driving algorithm is determined.

[0010] In one possible implementation, determining the optimization degree of the second autonomous driving algorithm relative to the first autonomous driving algorithm based on the plurality of second sample pairs and the first linear equation includes:

[0011] Determine the first driving state curve corresponding to the first linear equation in the set coordinate system;

[0012] Determine the driving state data points corresponding to the plurality of second sample pairs in the set coordinate system;

[0013] Based on the positional relationship between multiple driving state data points and the first driving state curve, the optimization degree of the second autonomous driving algorithm compared to the first autonomous driving algorithm is determined.

[0014] In one possible implementation, determining the optimization degree of the second autonomous driving algorithm compared to the first autonomous driving algorithm based on the positional relationship between multiple driving state data points and the first driving state curve includes:

[0015] Determine the ratio of the number of points above the first driving state curve to the number of points below the first driving state curve among the plurality of driving state data points.

[0016] Based on the ratio, the degree of optimization of the second autonomous driving algorithm compared to the first autonomous driving algorithm is determined.

[0017] In one possible implementation, determining the optimization degree of the second autonomous driving algorithm relative to the first autonomous driving algorithm based on the plurality of second sample pairs and the first linear equation includes:

[0018] Regression processing is performed on the multiple second sample pairs to obtain the second linear equation;

[0019] Determine the first driving state curve corresponding to the first linear equation in the set coordinate system, and determine the second driving state curve corresponding to the second linear equation in the set coordinate system;

[0020] Based on the positional relationship between the first driving state curve and the second driving state curve, the optimization degree of the second autonomous driving algorithm compared to the first autonomous driving algorithm is determined.

[0021] In one possible implementation, the following operations are performed for each test vehicle operating based on the first autonomous driving algorithm:

[0022] The test vehicle acquires driving data on any road segment, including the speed, position, and category of objects within a set range collected by the test vehicle; the speed of the object refers to the speed of the object relative to the test vehicle; the position of the object refers to the position of the object relative to the test vehicle.

[0023] Based on the speed, position, and category of objects within a preset range collected by the test vehicle, the first sample standard entropy is determined when the test vehicle travels on any road segment.

[0024] In one possible implementation, the position of the object includes the distance and orientation angle of the object relative to the test vehicle; determining the first sample standard entropy of the test vehicle traveling on any road segment based on the speed, position, and category of objects within a preset range collected by the test vehicle includes:

[0025] The first independent entropy matrix is ​​determined based on the speed of objects within a preset range collected when the test vehicle travels on any road segment.

[0026] The second independent entropy matrix is ​​determined based on the orientation angles of objects within a preset range collected by the test vehicle while it is traveling on any road segment.

[0027] The third independent entropy matrix is ​​determined based on the categories of objects collected by the test vehicle within a preset range while it is traveling on any road segment.

[0028] Based on the distances of objects within a preset range collected by the test vehicle while it is traveling on any road segment, a distance coefficient matrix is ​​determined;

[0029] The K-value matrix is ​​determined based on the first independent entropy matrix, the second independent entropy matrix, the third independent entropy matrix, and the distance coefficient matrix.

[0030] Based on the K-value matrix, determine the first sample standard entropy when the test vehicle is traveling on any road segment.

[0031] In one possible implementation, determining the first sample standard entropy of the test vehicle traveling on any road segment based on the K-value matrix includes:

[0032] The K-value matrix is ​​processed using the Lagrange function to obtain an estimate of the entropy;

[0033] Based on the estimated entropy, a first sample standard entropy is determined when the test vehicle travels on any road segment.

[0034] Secondly, embodiments of this application provide an apparatus for determining the optimization degree of an autonomous driving algorithm, the apparatus comprising:

[0035] The data acquisition unit is used to determine multiple first sample pairs based on the driving data of the test vehicle driving on the road based on the first autonomous driving algorithm; the first sample pair includes the first sample standard entropy and the average number of interventions corresponding to the first sample standard entropy.

[0036] A processing unit is used to perform regression processing on the plurality of first sample pairs to obtain a first linear equation;

[0037] Based on the driving data of the test vehicle on the road using the second autonomous driving algorithm, multiple second sample pairs are determined; the second sample pair includes the average number of interventions corresponding to the second sample standard entropy and the second sample standard entropy.

[0038] Based on the plurality of second sample pairs and the first linear equation, the optimization degree of the second autonomous driving algorithm compared to the first autonomous driving algorithm is determined.

[0039] Thirdly, embodiments of this application provide an electronic device, including a memory and a processor, wherein the memory stores a computer program that can run on the processor, and when the computer program is executed by the processor, it implements the method described in any one of the methods for determining the optimization degree of an autonomous driving algorithm in the first aspect.

[0040] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the method described in any one of the methods for determining the optimization degree of an autonomous driving algorithm in the first aspect.

[0041] This application provides a method, apparatus, and electronic device for determining the optimization degree of an autonomous driving algorithm. Based on driving data of a test vehicle using a first autonomous driving algorithm on a road, multiple first sample pairs are determined. Each first sample pair includes a first sample standard entropy and the average number of interventions corresponding to the first sample standard entropy. Regression processing is performed on the multiple first sample pairs to obtain a first linear equation. Then, based on driving data of the test vehicle using a second autonomous driving algorithm on a road, multiple second sample pairs are determined. Each second sample pair includes a second sample standard entropy and the average number of interventions corresponding to the second sample standard entropy. Based on the multiple second sample pairs and the first linear equation, the optimization degree of the second autonomous driving algorithm compared to the first autonomous driving algorithm can be determined. By quantifying the autonomous driving algorithm, the optimization degree of the autonomous driving algorithm can be evaluated. Attached Figure Description

[0042] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 This application scenario diagram illustrates a method for determining the optimization degree of an autonomous driving algorithm, as provided in an embodiment of this application.

[0044] Figure 2A flowchart illustrating a method for determining the optimization degree of an autonomous driving algorithm, provided in an embodiment of this application;

[0045] Figure 3 A flowchart illustrating another method for determining the optimization degree of an autonomous driving algorithm provided in this application embodiment;

[0046] Figure 4 A schematic diagram of a road segment object provided in an embodiment of this application;

[0047] Figure 5 A schematic diagram of a coordinate system provided for an embodiment of this application;

[0048] Figure 6 A schematic diagram of another coordinate system provided for an embodiment of this application;

[0049] Figure 7 This is a schematic diagram of the structure of an autonomous driving algorithm optimization degree determination device provided in an embodiment of this application;

[0050] Figure 8 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0051] To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0052] It should be noted that the terms "comprising" and "having" and their variations used in this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or units that are not explicitly listed or that are inherent to such process, method, product, or device.

[0053] With the rapid development of technology, autonomous driving technology is playing an increasingly important role in people's daily travel. However, for one of the most important parts of autonomous driving technology, the improvement and optimization of autonomous driving algorithms cannot be quantitatively evaluated, and staff cannot assess whether the iterative autonomous driving algorithm has improved compared to the previous generation.

[0054] Based on this, embodiments of this application provide a method for determining the optimization degree of an autonomous driving algorithm. Multiple first sample pairs are determined based on driving data of a test vehicle traveling on a road using a first autonomous driving algorithm. Each first sample pair includes a first sample standard entropy and the average number of interventions corresponding to the first sample standard entropy. Regression processing is performed on the multiple first sample pairs to obtain a first linear equation. Then, multiple second sample pairs are determined based on driving data of the test vehicle traveling on a road using a second autonomous driving algorithm. Each second sample pair includes a second sample standard entropy and the average number of interventions corresponding to the second sample standard entropy. Based on the multiple second sample pairs and the first linear equation, the optimization degree of the second autonomous driving algorithm compared to the first autonomous driving algorithm can be determined. By quantifying the autonomous driving algorithm, it is possible to assess whether the iterative autonomous driving algorithm has improved compared to the previous generation of autonomous driving algorithms.

[0055] The following is a brief introduction to the application scenarios to which the technical solutions of the embodiments of this application are applicable. It should be noted that the application scenarios described below are only for illustrating the embodiments of this application and are not intended to limit the scope. In specific implementation, the technical solutions provided by the embodiments of this application can be flexibly applied according to actual needs.

[0056] Figure 1 This is a schematic diagram illustrating an application scenario for the autonomous driving algorithm optimization degree determination method provided in this application embodiment. For example... Figure 1 As shown, this application scenario includes a server 10 and an in-vehicle terminal 20. The in-vehicle terminal 20 is connected to the server 10 via a communication network, and the server 10 can connect to multiple in-vehicle terminals 20 respectively.

[0057] Test vehicles based on the first and second autonomous driving algorithms travel on test roads. A driver is present in each vehicle. When the autonomous driving system presents a collision risk, the system alerts the driver to take over. The system records the number of times the driver needs to take over on different road sections. An onboard terminal 20 is installed in each test vehicle. During travel, the onboard terminal 20 collects data on the types, distances, speeds, and azimuth angles of other objects entering within a set distance range on the test road. After the test, the collected travel data and the number of times the driver needed to take over on each road section are sent to server 10. Server 10 determines the entropy value based on the travel data and places the entropy value and the number of times the driver needed to take over on the same coordinate system. By comparing the first driving state curve generated by the entropy value and the number of takeovers during the test vehicle's journey using the first autonomous driving algorithm with the driving state data points generated by the entropy value and the number of takeovers during the test vehicle's journey using the second autonomous driving algorithm, the system can determine whether the second autonomous driving algorithm has been improved.

[0058] It should be noted that, Figure 1 The installation location of the vehicle terminal 20 shown in the diagram is for illustrative purposes only and does not represent a fixed location in the vehicle. The vehicle terminal 20 can also be installed in other locations within the vehicle, such as in the door or between the front seats.

[0059] To further illustrate the technical solutions provided in the embodiments of this application, a detailed description is provided below in conjunction with the accompanying drawings and specific implementation methods. Although the embodiments of this application provide method operation steps as shown in the following embodiments or drawings, the method may include more or fewer operation steps based on conventional or non-inventive methods. In steps where there is no logically necessary causal relationship, the execution order of these steps is not limited to the execution order provided in the embodiments of this application. In actual processing or when the device executes the method, it may be executed in the order shown in the embodiments or drawings, or in combination.

[0060] Figure 2 A flowchart illustrating a method for determining the optimization degree of an autonomous driving algorithm according to an embodiment of this application is shown. This method can be completed by the cooperation of an in-vehicle terminal and a server. Figure 2 As shown, the optimization method for this autonomous driving algorithm includes the following steps:

[0061] Step S201: Based on the driving data of the test vehicle driving on the road using the first autonomous driving algorithm, determine multiple first sample pairs.

[0062] The first sample pair can include the first sample standard entropy and the average number of interventions corresponding to the first sample standard entropy. The first sample standard entropy can be used to characterize the driving difficulty of the road segment where the test vehicle is located. The higher the first sample entropy, the more complex the road condition information that the test vehicle needs to process, and the higher the driving difficulty. Conversely, the lower the first sample entropy, the simpler the road condition information that the test vehicle needs to process, and the lower the driving difficulty. The average number of interventions corresponding to the first sample standard entropy refers to the average number of interventions per unit distance traveled by the test vehicle when driving on a road segment with a driving difficulty of the first sample standard entropy.

[0063] In some embodiments, the driving data of any test vehicle based on the first autonomous driving algorithm on any road segment can be processed using methods such as... Figure 3 The steps shown determine the first sample pair:

[0064] Step S2011: Obtain driving data of the test vehicle on any road segment.

[0065] The driving data includes the speed, position, and category of objects within a defined range collected by the test vehicle. Objects within the defined range can include those that enter the test vehicle's designated distance range on the road during its operation. The test vehicle can detect objects entering its designated distance range using an image acquisition device or radar. The object's speed refers to its speed relative to the test vehicle; the object's position refers to its position relative to the test vehicle.

[0066] For example, multiple test vehicles based on the first autonomous driving algorithm can travel on a test road segment at different times. For instance, 50 test vehicles based on the first autonomous driving algorithm can travel on the test road segment. The test road segment can be divided into multiple road segments, and for each test vehicle, driving data of that test vehicle on each road segment can be obtained.

[0067] Taking test vehicle A as an example, assume the test road segment includes five road segments: L1, L2, L3, L4, and L5. While test vehicle A is traveling on road segment L1, its onboard terminal can collect the speed, position, and type of objects entering within a set distance range. For example... Figure 4 As shown, the object categories can include different types of vehicles, pedestrians, etc. For example, the object categories can be pedestrians, tricycles, autonomous vehicles, non-autonomous small cars, and large vehicles, etc.; the object's position is the position of each object relative to the test vehicle. Based on the object's position, the direction angle and distance of each object relative to the test vehicle can be determined; the object's speed is the speed of each object relative to the test vehicle. During the test vehicle's operation, the onboard terminal will report the number of times the driver needs to take over the vehicle. After each test vehicle finishes its journey, it can send the collected driving data and the number of times the driver needed to take over the vehicle for each road segment during the journey to the server.

[0068] Step S2012: Determine the first independent entropy matrix based on the speed of objects within a preset range collected when the test vehicle is driving on any road segment.

[0069] Taking test vehicle a traveling on road segment L1 as an example, the server can determine the first independent entropy matrix of the test vehicle on that road segment based on the speeds of objects within a set range collected by test vehicle a while traveling on road segment L1. The first independent entropy matrix is ​​the independent entropy matrix in the velocity dimension. Specifically, the independent entropy matrix in the velocity dimension includes the conditional entropy corresponding to multiple speeds of objects within the set range collected during the test vehicle's travel, and the entropy value of the conditional entropy corresponding to each speed. The conditional entropy in the velocity dimension characterizes the probability of a collision occurring with the test vehicle at the current speed of an object. The formula for calculating the conditional entropy is H(P)=-∑x,yP~(x)P(y|x)logP(y|x), where P(y|x) represents the probability of y occurring given x, and P~(x) is an empirical distribution. The independent entropy matrix in the velocity dimension is shown in the table below, where H… sn Let E represent the conditional entropy corresponding to velocity Sn, and let E represent the entropy value of the conditional entropy corresponding to velocity Sn.

[0070]

[0071] Step S2013: Determine the second independent entropy matrix based on the orientation angles of objects within a preset range collected when the test vehicle is driving on any road segment.

[0072] Taking test vehicle a traveling on road segment L1 as an example, the server can determine the second independent entropy matrix of the test vehicle on that road segment based on the azimuth angles of objects within a set range collected by the test vehicle a while traveling on road segment L1. The second independent entropy matrix is ​​the independent entropy matrix in the azimuth angle dimension. Specifically, the independent entropy matrix in the azimuth angle dimension includes the conditional entropy corresponding to multiple azimuth angles of objects within the set range collected during the test vehicle's travel, and the entropy value of the conditional entropy corresponding to each azimuth angle. The conditional entropy in the azimuth angle dimension characterizes the probability of a collision between the test vehicle and an object at the current azimuth angle. The formula for calculating the conditional entropy is H(P)=-∑x,yP~(x)P(y|x)logP(y|x), where P(y|x) represents the probability of y occurring given x, and P~(x) is an empirical distribution. The independent entropy matrix in the azimuth angle dimension is shown in the table below, where H… dn Let represent the conditional entropy corresponding to the direction angle dn, and E represent the entropy value of the conditional entropy corresponding to the direction angle dn.

[0073]

[0074]

[0075] Step S2014: Determine the third independent entropy matrix based on the categories of objects collected by the test vehicle within a preset range while it is driving on any road segment.

[0076] Taking test vehicle a traveling on road segment L1 as an example, the server can determine the third independent entropy matrix of the test vehicle on that road segment based on the categories of objects collected by test vehicle a within a set range while traveling on road segment L1. The third independent entropy matrix is ​​the independent entropy matrix of the category dimension. The independent entropy matrix of the test vehicle in the category dimension includes the conditional entropy corresponding to multiple categories of objects collected by the test vehicle during its travel, and the entropy value of the conditional entropy corresponding to each category. The conditional entropy of the category dimension is used to characterize the probability of a collision when the test vehicle encounters a certain object category. The formula for calculating the conditional entropy is H(P)=-∑x,yP~(x)P(y|x)logP(y|x), where P(y|x) represents the probability of y occurring given x, and P~(x) is the empirical distribution. The independent entropy matrix of the category dimension is shown in the table below, where H... cn Let E represent the conditional entropy corresponding to category Cn, and let E represent the entropy value of the conditional entropy corresponding to category Cn.

[0077]

[0078] Step S2015: Determine the distance coefficient matrix based on the distances of objects within a preset range collected when the test vehicle is driving on any road segment.

[0079] Taking test vehicle a traveling on road segment L1 as an example, the server can determine the distance coefficient matrix of the test vehicle on that road segment based on the distances of objects within a set range collected by test vehicle a while traveling on road segment L1.

[0080] k11 k12 k13……k1n

[0081] k21 k22 k23……k2n

[0082] k31 k32 k33……k3n

[0083] It should be noted that the execution order of steps S2012, S2013, S2014 and S2015 is not fixed. For example, step S2014 can be executed first, followed by steps S2013, S2012 and S2015 respectively. Alternatively, steps S2012, S2013, S2014 and S2015 can be executed simultaneously. This application does not limit this.

[0084] Step S2016: Determine the K-value matrix based on the first independent entropy matrix, the second independent entropy matrix, the third independent entropy matrix, and the distance coefficient matrix.

[0085] The first independent entropy matrix, the second independent entropy matrix, the third independent entropy matrix, and the distance coefficient matrix can be input into the following formula:

[0086] H(P)1=K 11 H s1 +K 12 H s2 +K 13 H s3 +………+K 1n H sn +β1

[0087] H(P)2=K 21 H d1 +K 22 H d2 +K 23 H d3 +………+K 2n H dn +β2

[0088] H(P)3=K 31 H c1 +K 32 H c2 +K 33 H c3 +………+K 3n H cn +β3

[0089] Output | K in H in +β i |, where i represents the i-th latitude, 1≤i≤3, and n represents the n-th time of data collection. Since objects on the opposite side of the median strip do not affect driving, we can let K on the opposite side of the median strip... in =0, where β is the compensation factor and H(P) is the entropy at different dimensions. Then, according to the least squares method, for |K in H in +β i By performing regression analysis, we can obtain the K-value matrix.

[0090] Step S2017: Determine the first sample standard entropy of the test vehicle when it is driving on any road segment based on the K-value matrix.

[0091] Taking the example of test vehicle a traveling on road segment L1, for instance, the K-value matrix can be processed using the Lagrange function to obtain an estimate of the entropy of test vehicle a traveling on road segment L1. Based on the estimated entropy of test vehicle a traveling on road segment L1 The first sample standard entropy of test vehicle a traveling on road segment L1 can be determined.

[0092] Specifically, after obtaining the estimated entropy of test vehicle a traveling on road segment L1, the estimated entropy of test vehicle a traveling on road segment L1 can be labeled to obtain the estimated value of test vehicle a traveling at each point on road segment L1. (p) can be used as a marker. Add weather and time disturbance compensation to (p), resulting in H(p) = F1 H(p) + δ(weather and time disturbance compensation), where F1 is the coefficient matrix. Input H(p) into... It can output the first sample standard entropy of test vehicle a traveling on road segment L1.

[0093] Step S2018: Determine the first sample pair based on the first sample standard entropy when the test vehicle is driving on any road segment and the average number of takeovers by the test vehicle on the corresponding road segment.

[0094] Using the above method, for each test vehicle, the server can determine the first sample standard entropy when the test vehicle is driving on each road segment, and obtain the average number of interventions when the test vehicle is driving on each road segment. In this way, multiple first sample standard entropies and the average number of interventions corresponding to each first sample standard entropy can be obtained. Each first sample standard entropy and its corresponding average number of interventions can be used as a first sample pair to obtain multiple first sample pairs.

[0095] After determining multiple first sample pairs through step S201, step S202 can be executed.

[0096] Step S202: Perform regression processing on multiple first sample pairs to obtain the first linear equation.

[0097] After obtaining multiple sample pairs, regression processing can be performed on multiple first sample pairs to obtain the first linear equation.

[0098] Step S203: Based on the driving data of the test vehicle driving on the road using the second autonomous driving algorithm, determine multiple second sample pairs.

[0099] The process of determining multiple second sample pairs is the same as the process of determining multiple first sample pairs in step S201. The only difference is that the test vehicle drives on the road and collects data based on the second autonomous driving algorithm, which will not be described in detail here.

[0100] Step S204: Based on multiple second sample pairs and the first linear equation, determine the optimization degree of the second autonomous driving algorithm compared to the first autonomous driving algorithm.

[0101] In one optional implementation, a first driving state curve corresponding to the first linear equation in a set coordinate system can be determined, and driving state data points corresponding to multiple second sample pairs in the set coordinate system can be determined respectively. Based on the positional relationship between the multiple driving state data points and the first driving state curve, the optimization degree of the second autonomous driving algorithm compared to the first autonomous driving algorithm can be determined.

[0102] The positional relationship between the first driving state curve corresponding to the first linear equation in the set coordinate system and the driving state data points corresponding to multiple second sample pairs in the set coordinate system, such as... Figure 5 As shown in the figure, the hollow circles represent the driving state data points corresponding to multiple second sample pairs in a set coordinate system, and the curve formed by the "+" signs represents the first driving state curve, which can be a straight line. For example, the ratio of the number of points above the first driving state curve to the number of points below the first driving state curve can be determined, and the optimization degree of the second autonomous driving algorithm compared to the first autonomous driving algorithm can be determined based on this ratio.

[0103] For example, if the ratio of the number of points above the first driving state curve to the number of points below the first driving state curve in the driving state data points corresponding to multiple second sample pairs in the set coordinate system is 9:1, it indicates that under the same sample standard entropy, i.e., the same driving difficulty, the second autonomous driving algorithm requires more driver intervention than the first autonomous driving algorithm, indicating that the second autonomous driving algorithm is less efficient than the first autonomous driving algorithm. Conversely, if the ratio of the number of points above the first driving state curve to the number of points below the first driving state curve in the driving state data points corresponding to multiple second sample pairs in the set coordinate system is 1:9, it indicates that under the same sample standard entropy, i.e., the same driving difficulty, the second autonomous driving algorithm requires fewer driver interventions than the first autonomous driving algorithm, indicating that the second autonomous driving algorithm is an improvement over the first autonomous driving algorithm.

[0104] In some embodiments, the ratio of the number of points located in a defined area to the total number of points among multiple driving state data points can be determined. Based on this ratio, the optimization degree of the second autonomous driving algorithm compared to the first autonomous driving algorithm can be determined, such as... Figure 6 As shown, the dashed line represents the first driving state curve corresponding to the first linear equation in the set coordinate system. The set area can include the severely declining area, the slightly declining area, the area with no obvious improvement, the area with good improvement, and the area with excellent improvement.

[0105] For example, if the ratio of the number of points in the improved region to the total number of points in the driving state data points corresponding to multiple second sample pairs in the set coordinate system is 9:1, it means that under the same standard entropy (i.e., the same driving difficulty), the second autonomous driving algorithm requires fewer driver interventions than the first autonomous driving algorithm, indicating a significant improvement. Conversely, if the ratio of the number of points in the severely deteriorated region to the total number of points in the driving state data points corresponding to multiple second sample pairs in the set coordinate system is 9:1, it means that under the same standard entropy (i.e., the same driving difficulty), the second autonomous driving algorithm requires more driver interventions than the first autonomous driving algorithm, indicating a significant deterioration.

[0106] In another optional implementation, if the acquired driving data of the second autonomous driving algorithm is large enough to obtain a sufficient number of second sample pairs, regression processing can be performed on multiple second sample pairs to obtain a second linear equation. The first driving state curve corresponding to the first linear equation in the set coordinate system is determined, as well as the second driving state curve corresponding to the second linear equation in the set coordinate system is determined. Based on the positional relationship between the first driving state curve and the second driving state curve, the optimization degree of the second autonomous driving algorithm compared to the first autonomous driving algorithm is determined.

[0107] For example, both the first and second driving state curves can be straight lines. If the slope of the second driving state curve is greater than the slope of the first driving state curve, it indicates that the second autonomous driving algorithm has deteriorated compared to the first. The larger the difference between the slopes of the second and first driving state curves, the more severe the deterioration of the second autonomous driving algorithm compared to the first. If the slope of the second driving state curve is less than the slope of the first driving state curve, it indicates that the second autonomous driving algorithm has improved compared to the first. The larger the difference between the slopes of the first and second driving state curves, the better the improvement of the second autonomous driving algorithm compared to the first. If the slope of the second driving state curve is equal to the slope of the first driving state curve, it indicates that the second autonomous driving algorithm has neither improved nor deteriorated compared to the first.

[0108] Based on the same inventive concept, this embodiment of the invention also provides a structural schematic diagram of an autonomous driving algorithm optimization degree determination device, as shown below. Figure 7 As shown, the autonomous driving algorithm optimization determination device includes:

[0109] The data acquisition unit 701 is used to determine multiple first sample pairs based on the driving data of the test vehicle driving on the road based on the first autonomous driving algorithm. The first sample pair includes the first sample standard entropy and the average number of interventions corresponding to the first sample standard entropy. The first sample standard entropy is used to characterize the driving difficulty of the road segment where the test vehicle is located. The average number of interventions corresponding to the first sample standard entropy refers to the average number of interventions per unit driving distance when the test vehicle is driving on a road segment with a driving difficulty of the first sample standard entropy.

[0110] The processing unit 702 is used to perform regression processing on multiple first sample pairs to obtain a first linear equation;

[0111] Based on the driving data of the test vehicle on the road using the second autonomous driving algorithm, multiple second sample pairs are determined; the second sample pairs include the average number of interventions corresponding to the standard entropy of the second sample and the standard entropy of the second sample.

[0112] Based on multiple second sample pairs and the first linear equation, the optimization degree of the second autonomous driving algorithm compared to the first autonomous driving algorithm is determined.

[0113] Based on the same inventive concept, this application also provides an electronic device, which can be the server mentioned above. This electronic device includes at least a memory for storing data and a processor. The processor for data processing can be implemented using a microprocessor, CPU, GPU (Graphics Processing Unit), DSP, or FPGA. The memory stores operation instructions, which can be computer-executable code, to implement the various steps in the autonomous driving algorithm optimization determination method described in this application.

[0114] Figure 8 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 8 As shown, the electronic device 800 includes a memory 801, a processor 802, a data acquisition module 803, and a bus 804. The memory 801, processor 802, and data acquisition module 803 are all connected via the bus 804, which is used for data transmission between the memory 801, processor 802, and data acquisition module 803.

[0115] The memory 801 can be used to store software programs and modules. The processor 802 executes various functional applications and data processing of the electronic device 800 by running the software programs and modules stored in the memory 801, such as the autonomous driving algorithm optimization determination method provided in this application embodiment. The memory 801 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, application programs of at least one application, etc.; the data storage area may store data created according to the use of the electronic device 800, etc. In addition, the memory 801 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0116] The processor 802 is the control center of the electronic device 800. It connects various parts of the electronic device 800 via the bus 804 and various interfaces and lines. It executes various functions and processes data of the electronic device 800 by running or executing software programs and / or modules stored in the memory 801, and by calling data stored in the memory 801. Optionally, the processor 802 may include one or more processing units, such as a CPU, GPU (Graphics Processing Unit), or digital processing unit.

[0117] This application also provides a computer-readable storage medium storing computer-executable instructions. When the computer program is executed by a processor, it can be used to implement the autonomous driving algorithm optimization determination method described in any embodiment of this application.

[0118] In some possible implementations, various aspects of the autonomous driving algorithm optimization determination method provided in this application can also be implemented in the form of a program product, which includes program code. When the program product is run on a computer device, the program code is used to cause the computer device to perform the steps of the autonomous driving algorithm optimization determination method according to the various exemplary embodiments of this application described above. For example, the computer device can perform actions such as... Figure 2 The flowchart illustrates the method for determining the optimization degree of autonomous driving algorithms.

[0119] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0120] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0121] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0122] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0123] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for determining the optimization degree of an autonomous driving algorithm, characterized in that, The method includes: Based on the driving data of the test vehicle on the road using the first autonomous driving algorithm, multiple first sample pairs are determined; the first sample pair includes the first sample standard entropy and the average number of interventions corresponding to the first sample standard entropy. Regression processing is performed on the multiple first sample pairs to obtain the first linear equation; Based on the driving data of the test vehicle on the road using the second autonomous driving algorithm, multiple second sample pairs are determined; the second sample pair includes the average number of interventions corresponding to the second sample standard entropy and the second sample standard entropy. Based on the plurality of second sample pairs and the first linear equation, the optimization degree of the second autonomous driving algorithm compared to the first autonomous driving algorithm is determined.

2. The method according to claim 1, characterized in that, The step of determining the optimization degree of the second autonomous driving algorithm relative to the first autonomous driving algorithm based on the plurality of second sample pairs and the first linear equation includes: Determine the first driving state curve corresponding to the first linear equation in the set coordinate system; Determine the driving state data points corresponding to the plurality of second sample pairs in the set coordinate system; Based on the positional relationship between multiple driving state data points and the first driving state curve, the optimization degree of the second autonomous driving algorithm compared to the first autonomous driving algorithm is determined.

3. The method according to claim 2, characterized in that, The step of determining the optimization degree of the second autonomous driving algorithm compared to the first autonomous driving algorithm based on the positional relationship between multiple driving state data points and the first driving state curve includes: Determine the ratio of the number of points above the first driving state curve to the number of points below the first driving state curve among the plurality of driving state data points. Based on the ratio, the degree of optimization of the second autonomous driving algorithm compared to the first autonomous driving algorithm is determined.

4. The method according to claim 1, characterized in that, The step of determining the optimization degree of the second autonomous driving algorithm relative to the first autonomous driving algorithm based on the plurality of second sample pairs and the first linear equation includes: Regression processing is performed on the multiple second sample pairs to obtain the second linear equation; Determine the first driving state curve corresponding to the first linear equation in the set coordinate system, and determine the second driving state curve corresponding to the second linear equation in the set coordinate system; Based on the positional relationship between the first driving state curve and the second driving state curve, the optimization degree of the second autonomous driving algorithm compared to the first autonomous driving algorithm is determined.

5. The method according to claim 1, characterized in that, The method for determining the standard entropy of the first sample includes: For each test vehicle operating based on the first autonomous driving algorithm, the following operations are performed: The test vehicle acquires driving data on any road segment, including the speed, position, and category of objects within a set range collected by the test vehicle; the speed of the object refers to the speed of the object relative to the test vehicle; the position of the object refers to the position of the object relative to the test vehicle. Based on the speed, position, and category of objects within a preset range collected by the test vehicle, the first sample standard entropy is determined when the test vehicle travels on any road segment.

6. The method according to claim 5, characterized in that, The position of the object includes the distance and orientation angle of the object relative to the test vehicle; determining the first sample standard entropy of the test vehicle when traveling on any road segment based on the speed, position, and category of objects collected by the test vehicle within a preset range includes: The first independent entropy matrix is ​​determined based on the speed of objects within a preset range collected when the test vehicle travels on any road segment. The second independent entropy matrix is ​​determined based on the orientation angles of objects within a preset range collected by the test vehicle while it is traveling on any road segment. The third independent entropy matrix is ​​determined based on the categories of objects collected by the test vehicle within a preset range while it is traveling on any road segment. Based on the distances of objects within a preset range collected by the test vehicle while it is traveling on any road segment, a distance coefficient matrix is ​​determined; The K-value matrix is ​​determined based on the first independent entropy matrix, the second independent entropy matrix, the third independent entropy matrix, and the distance coefficient matrix. Based on the K-value matrix, determine the first sample standard entropy when the test vehicle is traveling on any road segment.

7. The method according to claim 6, characterized in that, The step of determining the first sample standard entropy of the test vehicle when driving on any road segment based on the K-value matrix includes: The K-value matrix is ​​processed using the Lagrange function to obtain an estimate of the entropy; Based on the estimated entropy, a first sample standard entropy is determined when the test vehicle travels on any road segment.

8. A device for determining the optimization degree of an autonomous driving algorithm, characterized in that, The device includes: The data acquisition unit is used to determine multiple first sample pairs based on the driving data of the test vehicle driving on the road based on the first autonomous driving algorithm; the first sample pair includes the first sample standard entropy and the average number of interventions corresponding to the first sample standard entropy. A processing unit is used to perform regression processing on the plurality of first sample pairs to obtain a first linear equation; Based on the driving data of the test vehicle on the road using the second autonomous driving algorithm, multiple second sample pairs are determined; the second sample pair includes the average number of interventions corresponding to the second sample standard entropy and the second sample standard entropy. Based on the plurality of second sample pairs and the first linear equation, the optimization degree of the second autonomous driving algorithm compared to the first autonomous driving algorithm is determined.

9. An electronic device, characterized in that, It includes a memory and a processor, wherein the memory stores a computer program that can run on the processor, and when the computer program is executed by the processor, it implements the method of any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program therein, characterized in that: When the computer program is executed by a processor, it implements the method according to any one of claims 1 to 7.

Images