Method and device for judging the accuracy of simulated driving data

Abstract

The application provides a simulation driving data accuracy judgment method and device, the method comprises the following steps: obtaining a plurality of simulation driving data and real driving data corresponding to a target driving scene; for each simulation driving data and the corresponding real driving data: calculating the cumulative proportion corresponding to each simulation driving data and each real driving data respectively; fitting the cumulative proportion of all simulation driving data and the logarithmic value of the simulation driving data to determine the first function corresponding to each simulation driving data, and fitting the cumulative proportion of all real driving data and the logarithmic value of the real driving data to determine the second function corresponding to each real driving data; calculating the maximum perpendicular distance of the first function and the second function; if the maximum perpendicular distance corresponding to each simulation driving data is less than the corresponding preset data threshold, it is determined that the plurality of simulation driving data corresponding to the target driving scene is accurate.

Description

Technical Field

[0001] The present application relates to the field of autonomous driving technology, and in particular to a method and device for determining the accuracy of simulated driving data. Background Art

[0002] Autonomous driving simulation testing can achieve test mileage in a short period of time, which is difficult to achieve in real life. It can quickly simulate any driving scenario, helping developers to traverse various driving scenarios, significantly reducing test mileage and testing costs. Although autonomous driving simulation testing can generate simulation results for different driving scenarios, it is difficult for personnel to determine whether the simulation results are consistent with the actual results. Summary of the Invention

[0003] In view of this, the purpose of this application is to provide at least a method and device for judging the accuracy of simulated driving data. By calculating the vertical distance between a first function corresponding to the simulated driving data and a second function corresponding to the real driving data, the technical problem that the existing technology cannot judge whether the simulated data of the autonomous driving algorithm is accurate is solved, thereby achieving the technical effect of determining whether the simulated driving data is accurate.

[0004] This application mainly includes the following aspects:

[0005] In a first aspect, an embodiment of the present application provides a method for determining the accuracy of simulated driving data, the method comprising: obtaining multiple simulated driving data and real driving data corresponding to a target driving scene; the data quantity of each simulated driving data and each real driving data is a preset quantity; for each simulated driving data and the corresponding real driving data: respectively calculating the cumulative proportions corresponding to each simulated driving data and each real driving data; performing data fitting on the cumulative proportions of all simulated driving data and the logarithmic values ​​of the simulated driving data to determine a first function corresponding to each simulated driving data, and performing data fitting on the cumulative proportions of all real driving data and the logarithmic values ​​of the real driving data to determine a second function corresponding to each real driving data; calculating the maximum vertical distance between the first function and the second function; determining whether the maximum vertical distance corresponding to each simulated driving data is less than the corresponding preset data threshold; if the maximum vertical distance corresponding to each simulated driving data is less than the corresponding preset data threshold, determining that the multiple simulated driving data corresponding to the target driving scene are accurate.

[0006] Optionally, respectively calculating the cumulative ratio corresponding to each simulated driving data and each real driving data includes: for each simulated driving data, calculating the ratio of the number of remaining simulated driving data smaller than the simulated driving data to a preset number, and determining the ratio as the cumulative ratio corresponding to each simulated driving data; for each real driving data, calculating the ratio of the number of remaining real driving data smaller than the real driving data to a preset number, and determining the ratio as the cumulative ratio corresponding to each real driving data.

[0007] Optionally, data fitting is performed on the cumulative proportions of all simulated driving data and the logarithmic values ​​of the simulated driving data to determine the first function corresponding to each item of simulated driving data, and data fitting is performed on the cumulative proportions of all real driving data and the logarithmic values ​​of the real driving data to determine the second function corresponding to each item of real driving data, including: taking the logarithmic value of each simulated driving data corresponding to each item of simulated driving data as an independent variable and the cumulative proportion of each simulated driving data corresponding to each item of simulated driving data as a dependent variable to perform data fitting to obtain the first function corresponding to each item of simulated driving data; taking the logarithmic value of each real driving data corresponding to each item of real driving data as an independent variable and the cumulative proportion of each real driving data of each real driving data as a dependent variable to perform data fitting to obtain the second function corresponding to each item of real driving data.

[0008] Optionally, before calculating the cumulative ratio of each simulated driving data and each real driving data respectively, the method also includes: sorting a preset number of simulated driving data and real driving data according to data size; determining the simulated quantile corresponding to the sorted simulated driving data, and the real quantile corresponding to the sorted real driving data; the simulated quantile and the real quantile both include corresponding quantile identifiers; judging whether the difference between the simulated quantile and the real quantile with the same quantile identifier is less than the quantile threshold; calculating the cumulative ratio of each simulated driving data and each real driving data respectively, including: if the difference between the simulated quantile and the corresponding real quantile with the same quantile identifier is less than the quantile threshold, then calculating the cumulative ratio of each simulated driving data and each real driving data respectively.

[0009] Optionally, after determining whether the difference between the simulated quantile and the true quantile with the same quantile identifier is less than the quantile threshold, the method further includes: if the difference between the simulated quantile and the true quantile with the same quantile identifier is not less than the quantile threshold, then determining the quantile identifier whose difference between the simulated quantile and the corresponding true quantile is greater than or equal to the quantile threshold as the target quantile identifier; determining whether the target quantile identifier is a preset quantile identifier; if the target quantile identifier is not the preset quantile identifier, determining that the simulated driving data corresponding to the target quantile identifier is inaccurate; if the target quantile identifier is the preset quantile identifier, then calculating the cumulative proportion of each simulated driving data and each real driving data respectively.

[0010] Optionally, calculating the maximum vertical distance between the first function and the second function includes: determining the intersection range of the independent variables of the first function and the second function; within the intersection range of the independent variables, calculating the difference between the first function and the second function, and determining the absolute value of the difference as the vertical distance function between the first function and the second function; and determining the maximum value of the vertical distance function as the maximum vertical distance between the first function and the second function.

[0011] Optionally, the first function and the second function are fitting functions or piecewise functions.

[0012] In a second aspect, an embodiment of the present application further provides a device for judging the accuracy of simulated driving data, the device comprising: an acquisition module for acquiring multiple simulated driving data and real driving data corresponding to a target driving scene; the data quantity of each simulated driving data and each real driving data is a preset quantity; for each simulated driving data and the corresponding real driving data: a first calculation module for respectively calculating the cumulative proportions corresponding to each simulated driving data and each real driving data; a first determination module for performing data fitting on the cumulative proportions of all simulated driving data and the logarithmic values ​​of the simulated driving data to determine a first function corresponding to each simulated driving data, and performing data fitting on the cumulative proportions of all real driving data and the logarithmic values ​​of the real driving data to determine a second function corresponding to each real driving data; a second calculation module for calculating the maximum vertical distance between the first function and the second function; a judgment module for judging whether the maximum vertical distance corresponding to each simulated driving data is less than the corresponding preset data threshold; and a second determination module for determining that the multiple simulated driving data corresponding to the target driving scene are accurate if the maximum vertical distance corresponding to each simulated driving data is less than the corresponding preset data threshold.

[0013] In a third aspect, an embodiment of the present application further provides an electronic device comprising: a processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor. When the electronic device is running, the processor and the memory communicate through the bus, and the machine-readable instructions are executed by the processor to execute the steps of the method for determining the accuracy of simulated driving data in the first aspect or any possible implementation of the first aspect.

[0014] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of judging the accuracy of the simulated driving data in the above-mentioned first aspect or any possible implementation of the first aspect are executed.

[0015] The embodiment of the present application provides a method and device for determining the accuracy of simulated driving data, which obtains multiple simulated driving data and real driving data corresponding to a target driving scene; the data quantity of each simulated driving data and each real driving data is a preset quantity; for each simulated driving data and the corresponding real driving data: respectively calculating the cumulative ratio of each simulated driving data and each real driving data; performing data fitting on the cumulative ratio of all simulated driving data and the logarithmic value of the simulated driving data to determine a first function corresponding to each simulated driving data, and performing data fitting on the cumulative ratio of all real driving data and the logarithmic value of the real driving data to determine a second function corresponding to each real driving data; calculating the maximum vertical distance between the first function and the second function; determining whether the maximum vertical distance corresponding to each simulated driving data is less than the corresponding preset data threshold; if the maximum vertical distance corresponding to each simulated driving data is less than the corresponding preset data threshold, then determining that the multiple simulated driving data corresponding to the target driving scene are accurate. By calculating the vertical distance between the first function corresponding to the simulated driving data and the second function corresponding to the real driving data, the present application solves the technical problem that the prior art cannot determine whether the simulation data of the autonomous driving algorithm is accurate, thereby achieving the technical effect of determining whether the simulated driving data is accurate.

[0016] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without creative work.

[0018] Figure 1 A flow chart of a method for determining the accuracy of simulated driving data provided in an embodiment of the present application is shown.

[0019] Figure 2 A flowchart of the steps of calculating the maximum vertical distance between the first function and the second function provided in an embodiment of the present application is shown.

[0020] Figure 3 A flowchart of another method for determining the accuracy of simulated driving data provided in an embodiment of the present application is shown.

[0021] Figure 4 A functional module diagram of a device for judging the accuracy of simulated driving data provided in an embodiment of the present application is shown.

[0022] Figure 5 A schematic structural diagram of an electronic device provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0023] In order to make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. It should be understood that the drawings in the present application only serve the purpose of illustration and description and are not used to limit the scope of protection of the present application. In addition, it should be understood that the schematic drawings are not drawn to scale. The flowcharts used in this application illustrate the operations implemented according to some embodiments of the present application. It should be understood that the operations of the flowcharts can be implemented out of sequence, and steps without logical context can be reversed or implemented simultaneously. In addition, those skilled in the art, under the guidance of the contents of this application, can add one or more other operations to the flowchart, or remove one or more operations from the flowchart.

[0024] In addition, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application generally described and shown in the drawings here can be arranged and designed in various configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work are within the scope of protection of the present application.

[0025] Existing technologies do not disclose how to determine whether the simulated driving data of an autonomous driving algorithm is consistent with the actual driving data. Therefore, it is impossible to determine whether the simulated driving data is accurate, and further, it is impossible to determine whether the autonomous driving algorithm can be put into practical application.

[0026] Based on this, the embodiments of the present application provide a method and apparatus for determining the accuracy of simulated driving data. By calculating the vertical distance between a first function corresponding to the simulated driving data and a second function corresponding to the real driving data, this method solves the technical problem of the prior art being unable to determine whether the simulated data of the autonomous driving algorithm is accurate, thereby achieving the technical effect of determining whether the simulated driving data is accurate. The specific details are as follows:

[0027] See also Figure 1 , Figure 1 This is a flow chart of a method for determining the accuracy of simulated driving data provided by an embodiment of the present application. Figure 1 As shown, the method for determining the accuracy of simulated driving data provided in the embodiment of the present application includes the following steps:

[0028] S101: Acquire multiple pieces of simulated driving data and real driving data corresponding to a target driving scenario.

[0029] The multiple simulated driving data include multiple driving data types such as speed and acceleration obtained by simulating a target driving scenario using an autonomous driving algorithm; the multiple real driving data include multiple driving data types such as speed and acceleration obtained by driving a real vehicle in the target driving scenario. The simulation and real vehicle driving are repeated a preset number of times, and the resulting simulated driving data and real driving data are integrated to obtain a preset number of simulated driving data corresponding to each simulated driving data item, and a preset number of real driving data corresponding to each real driving data item. The preset number of times is the same as the preset number of data items.

[0030] Furthermore, the data quantity of each item of simulated driving data and each item of real driving data is a preset quantity.

[0031] For example, if the preset number is 100, the target driving scenario is input into the autonomous driving algorithm for 100 simulations, and each simulation can obtain multiple simulated driving data, that is, 100 sets of simulated driving data such as speed and acceleration can be obtained; multiple staff members drive the real vehicle 100 times in this target driving scenario, and each drive can obtain multiple real driving data, that is, 100 sets of real driving data such as speed and acceleration can be obtained; subsequent operations are performed on each simulated driving data and real driving data such as speed or acceleration to determine whether the simulated driving data of the autonomous driving algorithm in terms of speed or acceleration is accurate in the target driving scenario.

[0032] For example, if the simulated driving data of any item such as speed or acceleration is inaccurate, it can be directly considered that the simulated driving data obtained by the autonomous driving algorithm in the target driving scenario is inaccurate, and the staff needs to improve the autonomous driving algorithm.

[0033] For each simulated driving data and the corresponding real driving data:

[0034] S102: Calculate the cumulative ratio corresponding to each simulated driving data and each real driving data respectively.

[0035] Calculating the cumulative ratios corresponding to each simulated driving data and each real driving data separately includes:

[0036] For each simulated driving data, the ratio of the number of remaining simulated driving data smaller than the simulated driving data to a preset number is calculated, and the ratio is determined as the cumulative proportion corresponding to each simulated driving data; for each real driving data, the ratio of the number of remaining real driving data smaller than the real driving data to a preset number is calculated, and the ratio is determined as the cumulative proportion corresponding to each real driving data.

[0037] For example, for the simulated driving data and real driving data of the speed item: if there are 100 simulated driving data in the simulated driving data of the speed item, and for the first simulated driving data, 20 simulated driving data are smaller than the first simulated driving data, then the cumulative proportion of the first simulated driving data is 20 divided by 100, which is equal to 0.2, and 0.2 is the cumulative proportion of the first simulated driving data in the simulated driving data of the speed item. The above calculation is performed for the remaining 99 simulated driving data, and the cumulative proportion corresponding to each simulated driving data is obtained.

[0038] S103. Perform data fitting on the cumulative proportions of all simulated driving data and the logarithmic values ​​of the simulated driving data to determine a first function corresponding to each item of simulated driving data, and perform data fitting on the cumulative proportions of all real driving data and the logarithmic values ​​of the real driving data to determine a second function corresponding to each item of real driving data.

[0039] Performing data fitting on the cumulative proportions of all simulated driving data and the logarithmic values ​​of the simulated driving data to determine a first function corresponding to each simulated driving data, and performing data fitting on the cumulative proportions of all real driving data and the logarithmic values ​​of the real driving data to determine a second function corresponding to each real driving data includes:

[0040] The logarithmic value of each simulated driving data corresponding to each simulated driving data is used as the independent variable, and the cumulative proportion of each simulated driving data corresponding to each simulated driving data is used as the dependent variable for data fitting to obtain the first function corresponding to each simulated driving data; the logarithmic value of each real driving data corresponding to each real driving data is used as the independent variable, and the cumulative proportion of each real driving data of each real driving data is used as the dependent variable for data fitting to obtain the second function corresponding to each real driving data.

[0041] Exemplarily, for the simulated driving data of speed, the logarithmic value of each simulated driving data in the simulated driving data of speed is used as the independent variable, and the cumulative proportion of each simulated driving data in the simulated driving data of speed is used as the dependent variable for data fitting, and the fitted function is determined as the first function corresponding to the simulated driving data of speed.

[0042] Exemplarily, a first function corresponding to the simulated driving data item of speed is determined by performing data fitting on the cumulative ratio of the 100 simulated driving data items of speed and the logarithmic values ​​of the 100 simulated driving data items. A second function corresponding to the real driving data item of speed is determined by performing data fitting on the cumulative ratio of the 100 real driving data items of speed and the logarithmic values ​​of the 100 real driving data items.

[0043] The first function and the second function may be fitting functions or piecewise functions.

[0044] S104: Calculate the maximum vertical distance between the first function and the second function.

[0045] See also Figure 2 , Figure 2 A flowchart showing the steps of calculating the maximum vertical distance between the first function and the second function provided in an embodiment of the present application is shown. Calculating the maximum vertical distance between the first function and the second function includes:

[0046] S1041. Determine the intersection range of the independent variables of the first function and the second function.

[0047] That is to say, for each item of simulated driving data, the logarithmic values ​​of all the simulated driving data in the simulated driving data are obtained, and the logarithmic value of the smallest simulated driving data in the simulated driving data is determined as the minimum value of the independent variable value range of the first function corresponding to the simulated driving data; the logarithmic value of the largest simulated driving data in the simulated driving data is determined as the maximum value of the independent variable value range of the first function corresponding to the simulated driving data.

[0048] For each item of real driving data, the logarithmic values ​​of all the real driving data in the item of real driving data are obtained, and the minimum logarithmic value of the real driving data in the item of real driving data is determined as the minimum value of the independent variable value range of the second function corresponding to the real driving data; and the maximum logarithmic value of the real driving data in the item of real driving data is determined as the maximum value of the independent variable value range of the second function corresponding to the real driving data.

[0049] The intersection of the value range of the independent variable of the first function and the value range of the independent variable of the second function is determined as the intersection range of the independent variables of the first function and the second function.

[0050] S1042. Calculate the difference between the first function and the second function within the intersection range of the independent variables, and determine the absolute value of the difference as the vertical distance function between the first function and the second function.

[0051] Within the intersection range of the independent variables, the first function and the second function are subtracted, and the absolute value of the difference is determined as the vertical distance function between the first and second functions. In other words, the independent variable value range of the vertical distance function is the intersection range of the independent variables of the first and second functions.

[0052] S1043. Determine the maximum value of the vertical distance function as the maximum vertical distance between the first function and the second function.

[0053] That is to say, each item of simulated driving data corresponds to a first function, each item of real driving data corresponds to a second function, and each item of simulated driving data and the corresponding real driving data correspond to a vertical distance function.

[0054] return Figure 1 S105: Determine whether the maximum vertical distance corresponding to each item of simulated driving data is less than the corresponding preset data threshold.

[0055] Determine whether the maximum vertical distance corresponding to each simulated driving data item is less than the corresponding preset data threshold. Each simulated driving data item corresponds to a preset data threshold, that is, the simulated driving data item speed corresponds to a preset speed threshold, and the simulated driving data item acceleration corresponds to a preset acceleration threshold.

[0056] S106: Determine whether the multiple simulated driving data corresponding to the target driving scenario are accurate.

[0057] If the maximum vertical distance corresponding to each item of simulated driving data is less than the corresponding preset data threshold, it is determined that the multiple items of simulated driving data corresponding to the target driving scene are accurate.

[0058] If the maximum vertical distance corresponding to each item of simulated driving data is not less than the corresponding preset data threshold, it is determined that the simulated driving data for which the maximum vertical distance corresponding to the simulated driving data is greater than or equal to the corresponding preset data threshold is inaccurate.

[0059] See also Figure 3 A flowchart of another method for determining the accuracy of simulated driving data provided in an embodiment of the present application is shown.

[0060] Before respectively calculating the cumulative ratio of each simulated driving data and each real driving data, the method further includes:

[0061] S201 , sorting a preset amount of simulated driving data and real driving data according to data size.

[0062] The preset amount of simulated driving data and real driving data are sorted in ascending order or descending order respectively. The sorting method of the simulated driving data and the real driving data needs to be unified, that is, the simulated driving data and the real driving data are sorted in ascending order respectively, or the simulated driving data and the real driving data are sorted in descending order respectively.

[0063] S202: Determine the simulated quantile corresponding to the sorted simulated driving data and the true quantile corresponding to the sorted real driving data; both the simulated quantile and the true quantile include corresponding quantile identifiers.

[0064] The quantile marker refers to a preset position in the sorted simulated driving data or real driving data.

[0065] For example, if the simulated driving data and the real driving data are sorted in ascending order respectively, and the quantiles are set to quartiles, the quantiles are identified as the lower quartile, the median, and the upper quartile (i.e., the 25% data after sorting, the 50% data after sorting, and the 75% data after sorting). The simulated quantiles include the simulated lower quartile, the simulated median, and the simulated upper quartile; the true quantiles include the true lower quartile, the true median, and the true upper quartile. That is, the simulated lower quartile, the simulated median, and the simulated upper quartile corresponding to the sorted simulated driving data, and the true lower quartile, the true median, and the true upper quartile corresponding to the sorted real driving data are determined.

[0066] S203: Determine whether the differences between the simulated quantile and the true quantile with the same quantile identifier are both smaller than the quantile threshold.

[0067] Determine whether the difference between the simulated quantile and the true quantile with the same quantile identifier is less than the quantile threshold, that is, determine whether there is a simulated quantile with the same quantile identifier and the true quantile whose difference is greater than or equal to the quantile threshold.

[0068] Calculate the cumulative ratio of each simulated driving data and each real driving data separately, including:

[0069] If the difference between the simulated quantiles with the same quantile identifier and the corresponding true quantiles is less than the quantile threshold, that is, if there is no simulated quantile with the same quantile identifier and the true quantile whose difference is greater than or equal to the quantile threshold, then execute step S207 and calculate the cumulative proportion of each simulated driving data and each real driving data respectively.

[0070] That is to say, before calculating the cumulative ratio of each simulated driving data and each real driving data respectively, a preliminary judgment is made on the accuracy of each simulated driving data by judging whether the simulated quantile of each simulated driving data is similar to the real quantile of each real driving data corresponding to each simulated driving data.

[0071] For example, for each item of simulated driving data that is sorted in ascending order, determine whether the difference between the simulated lower quartile and the true lower quartile is less than the quantile threshold, determine whether the difference between the simulated median and the true median is less than the quantile threshold, and determine whether the difference between the simulated upper quartile and the true upper quartile is less than the quantile threshold. If the difference between the simulated lower quartile and the true lower quartile is less than the quantile threshold, and the difference between the simulated median and the true median is less than the quantile threshold, and the difference between the simulated upper quartile and the true upper quartile is also less than the quantile threshold, then the simulated driving data is considered to be similar to the real driving data during preliminary verification. That is, it is preliminarily determined that the simulated driving data is accurate, and subsequent operations can be performed, and the accuracy of the simulated driving data can be further determined.

[0072] The quantile threshold can be set for each simulated driving data item. The quantile threshold can also be set based on the quantile identifier of the simulated quantile.

[0073] Exemplarily, the simulation quantile threshold corresponding to the simulated driving data for speed is different from the simulation quantile threshold corresponding to the simulated driving data for acceleration.

[0074] For example, for the simulated driving data corresponding to the speed item, the simulated quantile threshold corresponding to the difference between the simulated lower quartile and the true lower quartile is set as the first simulated quantile threshold; the simulated quantile threshold corresponding to the difference between the simulated median and the true median is set as the second simulated quantile threshold; and the simulated quantile threshold corresponding to the difference between the simulated upper quartile and the true upper quartile is set as the third simulated quantile threshold. If the difference between the simulated lower quartile and the true lower quartile is less than the first simulated quantile threshold, and the difference between the simulated median and the true median is less than the second simulated quantile threshold, and the difference between the simulated upper quartile and the true upper quartile is also less than the third simulated quantile threshold, then the simulated driving data is considered to be similar to the real driving data during preliminary verification.

[0075] After determining whether the differences between the simulated quantiles and the true quantiles with the same quantile identifier are both less than the quantile threshold, the method further includes:

[0076] S204 : Determine the quantile identifier whose difference between the simulated quantile and the corresponding true quantile is greater than or equal to the quantile threshold as the target quantile identifier.

[0077] If the difference between the simulated quantile and the true quantile with the same quantile identifier is not all less than the quantile threshold, that is, there is a simulated quantile with the same quantile identifier and the true quantile whose difference is greater than or equal to the quantile threshold, then the quantile identifier whose difference between the simulated quantile and the corresponding true quantile is greater than or equal to the quantile threshold is determined as the target quantile identifier.

[0078] That is to say, if the difference between the simulated quantile and the true quantile with the same quantile identifier is not all less than the quantile threshold, the quantile identifier whose difference between the simulated quantile and the corresponding true quantile is greater than or equal to the quantile threshold is determined as the target quantile identifier.

[0079] Exemplarily, for each item of simulated driving data after being arranged in ascending order, if the difference between the simulated lower quartile and the true lower quartile is greater than or equal to the quantile threshold, the difference between the simulated median and the true median is less than the quantile threshold, and the difference between the simulated upper quartile and the true upper quartile is less than the quantile threshold, then the lower quartile corresponding to the simulated lower quartile is determined as the target quantile identifier.

[0080] S205: Determine whether the target quantile identifier is a preset quantile identifier.

[0081] If the target quantile identifier is a preset quantile identifier, step S207 is executed to respectively calculate the cumulative ratio of each simulated driving data and each real driving data.

[0082] That is to say, if the preset quantile identifier is set to the upper quartile or the lower quartile, the difference between the simulated lower quartile and the true lower quartile is greater than or equal to the quantile threshold, or the difference between the simulated upper quartile and the true upper quartile is greater than or equal to the quantile threshold, it is still preliminarily judged that this simulated driving data is accurate, and step S207 is continued to be executed to make further judgment on the accuracy of the simulated driving data.

[0083] The highest digit or the lowest digit in the quantile identifier may be set as the preset quantile identifier.

[0084] S206: Determine whether the simulated driving data corresponding to the target quantile identifier is inaccurate.

[0085] If the target quantile identifier is not a preset quantile identifier, it is determined that the simulated driving data corresponding to the target quantile identifier is inaccurate.

[0086] That is, if the preset quantile identifier is set as the upper quartile or the lower quartile, if the difference between the simulation median and the real median is greater than or equal to the quantile threshold, it is determined that the simulation driving data corresponding to the target quantile identifier is inaccurate.

[0087] In the case that the simulation driving data corresponding to the target quantile identifier is inaccurate, it can also be considered that the simulation driving data is inaccurate, and then the staff needs to adjust the automatic driving algorithm to obtain new simulation driving data and then make a judgment.

[0088] S207, respectively calculate the cumulative proportion of each simulation driving data and each real driving data.

[0089] Step S207 is the same processing method as step S102 described above, and will not be repeated here.

[0090] Based on the same application concept, the embodiment of the present application also provides a simulation driving data accuracy judgment device corresponding to the simulation driving data accuracy judgment method provided by the above-mentioned embodiment. Since the principle of solving the problem in the device of the embodiment of the present application is similar to the simulation driving data accuracy judgment method of the above-mentioned embodiment of the present application, the implementation of the device can be referred to the implementation of the method, and the repeated parts will not be repeated.

[0091] As shown in Figure 3 , the simulation driving data accuracy judgment device comprises a simulation driving data obtaining unit 1, a simulation driving data accuracy judgment unit 2 and a simulation driving data accuracy judgment result output unit 3. Figure 3A function module diagram of the simulation driving data accuracy judgment device 300 provided by the embodiment is provided. The simulation driving data accuracy judgment device 10 comprises an acquisition module 101, a first calculation module 102, a first determination module 103, a second calculation module 104, a judgment module 105 and a second determination module 106. The acquisition module 101 is configured to acquire a plurality of simulation driving data corresponding to a target driving scene and real driving data. The data quantity of each simulation driving data and each real driving data is a preset quantity. For each simulation driving data and corresponding real driving data: the first calculation module 102 is configured to calculate the cumulative proportion of each simulation driving data and each real driving data, respectively. The first determination module 103 is configured to perform data fitting on the cumulative proportion of all simulation driving data and the logarithmic value of the simulation driving data to determine a first function corresponding to each simulation driving data, and perform data fitting on the cumulative proportion of all real driving data and the logarithmic value of the real driving data to determine a second function corresponding to each real driving data. The second calculation module 104 is configured to calculate the maximum perpendicular distance of the first function and the second function. The judgment module 105 is configured to judge whether the maximum perpendicular distance corresponding to each simulation driving data is less than a corresponding preset data threshold. The second determination module 106 is configured to determine that the plurality of simulation driving data corresponding to the target driving scene is accurate if the maximum perpendicular distance corresponding to each simulation driving data is less than the corresponding preset data threshold.

[0092] Based on the same application concept, see Figure 4 As shown in the structural schematic diagram of the electronic device provided by the embodiment, the electronic device 20 comprises a processor 201, a memory 202 and a bus 203. The memory 202 stores machine readable instructions executable by the processor 201. When the electronic device 20 is running, the processor 201 and the memory 202 communicate through the bus 203. The machine readable instructions are executed by the processor 201 to perform the steps of the simulation driving data accuracy judgment method in any one of the above embodiments.

[0093] Specifically, when the machine-readable instructions are executed by the processor 201, the following processing can be performed: obtaining multiple simulated driving data and real driving data corresponding to the target driving scene; the data quantity of each simulated driving data and each real driving data is a preset quantity; for each simulated driving data and the corresponding real driving data: respectively calculating the cumulative proportion of each simulated driving data and each real driving data; performing data fitting on the cumulative proportions of all simulated driving data and the logarithmic values ​​of the simulated driving data to determine the first function corresponding to each simulated driving data, and performing data fitting on the cumulative proportions of all real driving data and the logarithmic values ​​of the real driving data to determine the second function corresponding to each real driving data; calculating the maximum vertical distance between the first function and the second function; judging whether the maximum vertical distance corresponding to each simulated driving data is less than the corresponding preset data threshold; if the maximum vertical distance corresponding to each simulated driving data is less than the corresponding preset data threshold, it is determined that the multiple simulated driving data corresponding to the target driving scene are accurate.

[0094] Based on the same application concept, an embodiment of the present application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method for judging the accuracy of simulated driving data provided in the above embodiment are executed.

[0095] Specifically, the storage medium can be a general storage medium, such as a mobile disk, a hard disk, etc. When the computer program on the storage medium is run, it can execute the above-mentioned method for judging the accuracy of simulated driving data, by obtaining multiple simulated driving data and real driving data corresponding to the target driving scene; the data quantity of each simulated driving data and each real driving data is a preset quantity; for each simulated driving data and the corresponding real driving data: respectively calculating the cumulative proportion corresponding to each simulated driving data and each real driving data; performing data fitting on the cumulative proportion of all simulated driving data and the logarithmic value of the simulated driving data to determine the first function corresponding to each simulated driving data, and performing data fitting on the cumulative proportion of all real driving data and the logarithmic value of the real driving data to determine the second function corresponding to each real driving data; calculating the maximum vertical distance between the first function and the second function; judging whether the maximum vertical distance corresponding to each simulated driving data is less than the corresponding preset data threshold; if the maximum vertical distance corresponding to each simulated driving data is less than the corresponding preset data threshold, then determining that the multiple simulated driving data corresponding to the target driving scene are accurate.

[0096] Those skilled in the art can clearly understand that, for the convenience and brevity of the description, the specific working process of the system and the device described above can refer to the corresponding process in the foregoing method embodiment, and will not be repeated here. In several embodiments provided in the present application, it should be understood that the disclosed system, device and method can be implemented in other ways. The device embodiments described above are only schematic, for example, the division of the units is only a logical function division, and there can be another division manner in actual implementation, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some communication interfaces, devices or units, and can be electrical, mechanical or other forms.

[0097] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, i.e., they can be located in one place or distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.

[0098] In addition, the functional units in each embodiment of the present application can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit.

[0099] If the functions are realized in the form of software function units and sold or used as independent products, they can be stored in a non-volatile computer readable storage medium executable by a processor. Based on this understanding, the technical solutions of the present application essentially or the parts that contribute to the prior art or the parts of the technical solutions can be embodied in the form of software products, and the computer software products are stored in a storage medium, including a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the embodiments of the present application. The foregoing storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), magnetic disk or optical disk, and various program code storage media.

[0100] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical scope disclosed in the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for determining the accuracy of simulated driving data, characterized in that: The method for determining the accuracy of the simulated driving data includes: Acquire multiple simulated driving data and real driving data corresponding to the target driving scenario; the number of data for each simulated driving data and each real driving data is a preset number; For each simulated driving data and the corresponding real driving data: Calculate the cumulative ratio corresponding to each simulated driving data and each real driving data respectively; Performing data fitting on the cumulative proportions of all simulated driving data and the logarithmic values ​​of the simulated driving data to determine a first function corresponding to each item of simulated driving data, and performing data fitting on the cumulative proportions of all real driving data and the logarithmic values ​​of the real driving data to determine a second function corresponding to each item of real driving data; calculating a maximum vertical distance between the first function and the second function; Determining whether the maximum vertical distance corresponding to each item of simulated driving data is less than a corresponding preset data threshold; If the maximum vertical distance corresponding to each item of simulated driving data is less than the corresponding preset data threshold, it is determined that the multiple items of simulated driving data corresponding to the target driving scene are accurate; Among them, for each simulated driving data, the ratio of the number of remaining simulated driving data smaller than the simulated driving data to the preset number is calculated, and the ratio is determined as the cumulative proportion corresponding to each simulated driving data; for each real driving data, the ratio of the number of remaining real driving data smaller than the real driving data to the preset number is calculated, and the ratio is determined as the cumulative proportion corresponding to each real driving data.

2. The method for determining the accuracy of simulated driving data according to claim 1, wherein: The step of performing data fitting on the cumulative proportions of all simulated driving data and the logarithmic values ​​of the simulated driving data to determine a first function corresponding to each item of simulated driving data, and performing data fitting on the cumulative proportions of all real driving data and the logarithmic values ​​of the real driving data to determine a second function corresponding to each item of real driving data comprises: The logarithm of each simulated driving data corresponding to each simulated driving data is used as an independent variable, and the cumulative proportion of each simulated driving data corresponding to each simulated driving data is used as a dependent variable to perform data fitting to obtain a first function corresponding to each simulated driving data; The logarithm of each real driving data corresponding to each real driving data is used as an independent variable, and the cumulative proportion of each real driving data of each real driving data is used as a dependent variable for data fitting to obtain a second function corresponding to each real driving data.

3. The method for determining the accuracy of simulated driving data according to claim 1, wherein: Before respectively calculating the cumulative ratio of each simulated driving data and each real driving data, the method further includes: Sort the preset amount of simulated driving data and real driving data by data size; Determining a simulated quantile corresponding to the sorted simulated driving data and a true quantile corresponding to the sorted real driving data; the simulated quantile and the true quantile both including corresponding quantile identifiers; Determine whether the differences between the simulated quantile and the true quantile with the same quantile identifier are both less than a quantile threshold; The step of respectively calculating the cumulative ratio of each simulated driving data to each real driving data comprises: If the differences between the simulation quantiles and the corresponding real quantiles with the same quantile identifier are all smaller than the quantile threshold, the cumulative proportion of each simulation driving data and each real driving data is calculated respectively.

4. The method for determining the accuracy of simulated driving data according to claim 3, wherein: After determining whether the differences between the simulated quantiles and the true quantiles with the same quantile identifiers are both less than the quantile threshold, the method further includes: If the differences between the simulated quantiles and the true quantiles with the same quantile identifier are not all less than the quantile threshold, the quantile identifier with the difference between the simulated quantile and the corresponding true quantile being greater than or equal to the quantile threshold is determined as the target quantile identifier; Determining whether the target quantile identifier is a preset quantile identifier; If the target quantile identifier is not a preset quantile identifier, determining that the simulated driving data corresponding to the target quantile identifier is inaccurate; If the target quantile identifier is a preset quantile identifier, the cumulative ratio of each simulated driving data and each real driving data is calculated respectively.

5. The method for determining the accuracy of simulated driving data according to claim 1, wherein: Calculating the maximum vertical distance between the first function and the second function includes: Determining an intersection range of independent variables of the first function and the second function; Within the intersection range of the independent variables, calculating the difference between the first function and the second function, and determining the absolute value of the difference as the vertical distance function between the first function and the second function; The maximum value of the vertical distance function is determined as the maximum vertical distance between the first function and the second function.

6. The method for determining the accuracy of simulated driving data according to claim 1, wherein: The first function and the second function are fitting functions or piecewise functions.

7. A device for judging the accuracy of simulated driving data, characterized in that: The device comprises: An acquisition module is used to acquire multiple simulated driving data and real driving data corresponding to a target driving scenario; the number of data for each simulated driving data and each real driving data is a preset number; For each simulated driving data and the corresponding real driving data: A first calculation module is used to calculate the cumulative ratio corresponding to each simulated driving data and each real driving data respectively; a first determination module, configured to perform data fitting on the cumulative proportions of all simulated driving data and the logarithmic values ​​of the simulated driving data to determine a first function corresponding to each item of simulated driving data, and to perform data fitting on the cumulative proportions of all real driving data and the logarithmic values ​​of the real driving data to determine a second function corresponding to each item of real driving data; a second calculation module, configured to calculate a maximum vertical distance between the first function and the second function; a judgment module, configured to judge whether the maximum vertical distance corresponding to each item of simulated driving data is less than a corresponding preset data threshold; a second determining module, configured to determine that the plurality of simulated driving data corresponding to the target driving scene are accurate if the maximum vertical distance corresponding to each item of simulated driving data is less than a corresponding preset data threshold; Among them, the first calculation module is also used to: for each simulated driving data, calculate the ratio of the number of remaining simulated driving data that is smaller than the simulated driving data to a preset number, and determine the ratio as the cumulative proportion corresponding to each simulated driving data; for each real driving data, calculate the ratio of the number of remaining real driving data that is smaller than the real driving data to a preset number, and determine the ratio as the cumulative proportion corresponding to each real driving data.

8. An electronic device, characterized in that: include: A processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor. When the electronic device is running, the processor and the memory communicate through the bus. When the processor is running, the machine-readable instructions execute the steps of the method for determining the accuracy of simulated driving data as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method for determining the accuracy of simulated driving data according to any one of claims 1 to 6 are executed.

Images