Method and device for determining parameter range of driving scene

Abstract

The invention discloses a parameter range determination method and device for a driving scene, and the method comprises the steps: obtaining a plurality of parameter pairs of the driving scene, each parameter pair comprises a value of a first motion parameter and a value of a second motion parameter, and the two motion parameters are used for describing the state of a main body in the driving scene; dividing the plurality of parameter pairs on a two-dimensional distribution plane of the plurality of parameter pairs through at least one division line, wherein a longitudinal axis of the two-dimensional distribution plane indicates a first motion parameter and a transverse axis of the two-dimensional distribution plane indicates a second motion parameter, so as to obtain a plurality of groups of data; and determining the parameter range of the first motion parameter according to the regression analysis result of two groups of data which are farthest from each other in the plurality of groups of data, wherein the two groups of data are respectively positioned on the two sides of any parting line. The parameter range of the motion parameters obtained in this way not only considers the correlation among the motion parameters, but also fits the distribution condition of actual data of the motion parameters in the driving scene, and can provide reference for testing of the driving scene.

Description

Technical Field

[0001] The present application relates to the field of intelligent driving, and in particular to a method and device for determining a parameter range of a driving scenario. Background Art

[0002] The parameter range of the motion parameters (or scene parameters) of the driving scene is very important for the safety test of the autonomous vehicle. For example, whether the value of a certain motion parameter in the test scene is within the parameter range of the motion parameter will affect the safe driving of the autonomous vehicle in the test scene. Driving scenes can be, for example, vehicle cut-in, vehicle cut-out, and vehicle following.

[0003] The parameter range of the motion parameter currently determined by the traditional regression analysis method cannot fit the distribution of the actual data of the motion parameter well, so the parameter range of the motion parameter determined is unreasonable. Summary of the invention

[0004] The present application discloses a method and device for determining a parameter range of a driving scene. The determination of the parameter range of a motion parameter not only takes into account the correlation between the motion parameters, but also fits the distribution of actual data of the motion parameters in the driving scene more closely, thus providing a reference for testing the driving scene.

[0005] In a first aspect, the present application provides a method for determining a parameter range of a driving scene, the method comprising: obtaining multiple parameter pairs of the driving scene, each parameter pair comprising a value of a first motion parameter and a value of a second motion parameter, the first motion parameter and the second motion parameter being used to describe the state of a subject in the driving scene; determining a parameter range of the first motion parameter according to a regression analysis result of two sets of data that are farthest apart from each other among multiple sets of data; the parameter range of the first motion parameter is used to constrain the value of the first motion parameter in a simulation test of the driving scene; wherein the multiple sets of data are obtained by dividing the multiple parameter pairs on a two-dimensional distribution plane of the multiple parameter pairs by at least one dividing line, each set of data in the multiple sets of data comprises at least one parameter pair in the multiple parameter pairs, the vertical axis of the two-dimensional distribution plane indicates the first motion parameter, the horizontal axis of the two-dimensional distribution plane indicates the second motion parameter, the at least one dividing line is determined based on the distribution of the multiple parameter pairs on the two-dimensional distribution plane, and the two sets of data that are farthest apart are respectively located on both sides of any one of the at least one dividing line.

[0006] Here, the driving scene may be, for example, a scene where the front vehicle cuts in, a scene where the front vehicle cuts out, a scene where the front vehicle follows, or a scene where the front vehicle brakes.

[0007] The motion parameters are used to describe the state of the subject (such as vehicles, pedestrians, and any other possible traffic participants) in the driving scene. The operating parameters include, for example, at least one of the parameters such as speed, distance, relative speed, relative distance, displacement, acceleration, angular velocity, posture, lane change time, etc. It is understood that different driving scenes may correspond to different motion parameters.

[0008] Here, the first motion parameter and the second motion parameter are any two motion parameters in a driving scenario.

[0009] It can be understood that each parameter pair represents a data point on the above two-dimensional distribution plane, wherein the horizontal coordinate of the data point is a value of the second motion parameter, and the vertical coordinate of the data point is a value of the first motion parameter.

[0010] For example, a dividing line can divide the multiple parameter pairs into two groups of data on a two-dimensional distribution plane, one group of data is located on one side of the dividing line, and the other group of data is located on the other side of the dividing line. Two dividing lines can divide the multiple parameter pairs into three groups of data on a two-dimensional distribution plane.

[0011] Exemplarily, since the first motion parameter is indicated by the longitudinal axis of the two-dimensional distribution plane, each dividing line can divide the multiple parameter pairs into upper and lower parts on the two-dimensional distribution plane. When there are multiple dividing lines, the multiple dividing lines are arranged in sequence along the longitudinal axis direction on the two-dimensional distribution plane. In this case, the two groups of data that are farthest apart from each other in the multiple groups of data obtained by dividing the multiple dividing lines include: a group of data consisting of parameter pairs located above the uppermost dividing line on the two-dimensional distribution plane, and a group of data consisting of parameter pairs located below the lowermost dividing line on the two-dimensional distribution plane.

[0012] In the above method, the multiple parameter pairs are divided into multiple groups of data on the two-dimensional distribution plane by at least one dividing line, and the value of the first motion parameter indicated by the vertical axis of the two-dimensional distribution plane is determined by the regression analysis results of the two groups of data that are farthest apart from each other. In this way, the parameter range of the obtained motion parameters not only takes into account the correlation between the motion parameters, but also is more in line with the distribution of the actual data of the motion parameters in the driving scene.

[0013] Optionally, the plurality of parameter pairs present an asymmetric distribution.

[0014] Exemplarily, an asymmetric distribution is the opposite of a symmetric distribution (e.g., Gaussian distribution, uniform distribution, etc.). From the distribution morphology, the data presents a sparse distribution in one part and a dense distribution in another part. From the probability density curve, the probability density curve obtained based on these multiple parameter pairs is asymmetric on the left and right.

[0015] Asymmetric distribution includes but is not limited to long-tail distribution, mixed distribution, etc. Based on the different positions of the long tail, the long-tail distribution can be further divided into left-skewed distribution and right-skewed distribution.

[0016] By implementing the above implementation method, for data presenting an asymmetric distribution (i.e., multiple parameter pairs), the parameter range of the first motion parameter is determined by dividing the data into multiple groups based on the distribution of the data on the two-dimensional distribution plane and then performing regression analysis on the two groups of data that are farthest apart. Compared with the parameter range of the first motion parameter determined by directly performing regression analysis on the entire data, this will be more in line with the distribution of the actual data of the motion parameter in the driving scene, and thus the parameter range of the determined first motion parameter will also be more reasonable.

[0017] Optionally, the two groups of data include a first group of data and a second group of data, the upper limit of the parameter range of the first motion parameter is determined by an upper boundary in the regression analysis results of the first group of data, and the lower limit of the parameter range of the first motion parameter is determined by a lower boundary in the regression analysis results of the second group of data; wherein, on the two-dimensional distribution plane, the first group of data is located above the second group of data, and the upper boundary is asymmetric to the lower boundary.

[0018] Exemplarily, the upper limit corresponding to the first set of data is determined based on the statistical mean line of the first set of data and the standard deviation of a preset multiple (e.g., 3 times) above the mean line. The lower limit corresponding to the second set of data is determined based on the statistical mean line of the second set of data and the standard deviation of a preset multiple (e.g., 3 times) below the mean line.

[0019] By implementing the above implementation, the upper boundary corresponding to the first set of data and the lower boundary corresponding to the second set of data are asymmetric, which is more in line with the actual distribution of motion parameters than directly performing regression analysis on multiple parameter pairs to obtain symmetrical upper and lower boundaries.

[0020] Exemplarily, the method further includes: in the driving scenario, when the first value of the first motion parameter does not belong to the parameter range of the first motion parameter, determining that the first value of the first motion parameter is abnormal.

[0021] By implementing the above implementation, the parameter range of the first motion parameter can be used to constrain the value of the first motion parameter in the driving scene. In this way, abnormal data can be identified and eliminated from the sample data of the driving scene, which is conducive to improving test efficiency.

[0022] Optionally, an absolute value of a first skewness calculated based on the multiple parameter pairs is greater than zero, and the first skewness is used to indicate a degree of distribution dispersion of the multiple parameter pairs; wherein the first skewness is the third-order central moment of the multiple parameter pairs.

[0023] By implementing the above implementation, the distribution dispersion of the multiple parameter pairs can be known through the skewness, and thus it can also be known whether the multiple parameter pairs present an asymmetric distribution.

[0024] Optionally, the plurality of parameter pairs are divided into a third group of data and a fourth group of data via a first segmentation line, and a distribution discreteness of the fourth group of data is greater than a distribution discreteness of the third group of data.

[0025] By implementing the above implementation method, the multiple parameter pairs are divided into two groups of data. The parameter range of the first motion parameter can be determined based on the regression analysis results of each of the two groups of data. Compared with directly determining the parameter range of the first motion parameter by regression analysis of the multiple parameter pairs, it will be more in line with the distribution of actual data of the motion parameters in the driving scene, and thus the parameter range of the determined first motion parameter will also be more reasonable.

[0026] Optionally, the first dividing line is obtained by linear fitting of multiple mean points based on the statistics of the multiple parameter pairs; the first skewness is greater than zero, the third group of data includes parameter pairs among the multiple parameter pairs that are below the first dividing line, and the fourth group of data includes parameter pairs among the multiple parameter pairs that are above the first dividing line; or, the first skewness is less than zero, the fourth group of data includes parameter pairs among the multiple parameter pairs that are below the first dividing line, and the third group of data includes parameter pairs among the multiple parameter pairs that are above the first dividing line.

[0027] Exemplarily, when the first segmentation line is obtained by linearly fitting a plurality of statistical mean points based on the plurality of parameters, the first segmentation line is also referred to as a mean line.

[0028] That is to say, on the two-dimensional distribution plane, when the first skewness is greater than zero, the distribution dispersion degree of the group of data located above the first dividing line among the above-mentioned multiple parameter pairs is higher than the distribution dispersion degree of the group of data located below the first dividing line. In this case, the first skewness can also indicate that the direction of the distribution dispersion degree is right-skewed; when the first skewness is less than zero, the distribution dispersion degree of the group of data located below the first dividing line among the above-mentioned multiple parameter pairs is higher than the distribution dispersion degree of the group of data located above the first dividing line. In this case, the first skewness can also indicate that the direction of the distribution dispersion degree is left-skewed.

[0029] Optionally, when the multiple parameter pairs present a long-tailed distribution, the absolute value of the first skewness is greater than a first threshold, and the fourth group of data is divided into a fifth group of data and a sixth group of data by a second dividing line, wherein the distribution discreteness of the sixth group of data is greater than the distribution discreteness of the fifth group of data.

[0030] For example, from the distribution morphology, the long-tail distribution generally has: when the data points are projected on the longitudinal axis of the two-dimensional distribution plane, the number of data points shows an increasing or decreasing trend along the longitudinal axis. From the probability density curve, the probability density curve of the long-tail distribution is similar to a bell curve but has a long tail on the left or right side of the curve. When the tail is on the left, the long-tail distribution is also called a left-skewed distribution or a negatively skewed distribution; when the tail is on the right, the long-tail distribution is also called a right-skewed distribution or a positively skewed distribution.

[0031] In implementing the above implementation method, for multiple parameter pairs that present a long-tail distribution, these multiple parameter pairs are first divided into two groups of data, such as the third group of data and the fourth group of data mentioned above, and then the group of data with a higher degree of distribution dispersion among the two groups of data (i.e., the fourth group of data) is divided at least once. This is conducive to accurately estimating the parameter range of the first motion parameter so that it fits the actual distribution of these multiple parameter pairs.

[0032] Optionally, the second dividing line is obtained by linear fitting of multiple mean points based on the statistics of the fourth group of data; the first skewness is greater than zero, the fifth group of data includes parameter pairs in the fourth group of data that are below the second dividing line, and the sixth group of data includes parameter pairs in the fourth group of data that are above the second dividing line; or, the first skewness is less than zero, the sixth group of data includes parameter pairs in the fourth group of data that are below the second dividing line, and the fifth group of data includes parameter pairs in the fourth group of data that are above the second dividing line.

[0033] Exemplarily, when the second segmentation line is obtained by performing linear fitting on a plurality of mean value points based on statistics of the fourth set of data, the second segmentation line is also referred to as a mean value line.

[0034] That is to say, on the two-dimensional distribution plane, when the first skewness is greater than zero, the distribution dispersion degree of the group of data located above the second dividing line in the above-mentioned fourth group of data is higher than the distribution dispersion degree of the group of data located below the second dividing line. In this case, the first skewness can also indicate that the direction of the distribution dispersion degree is right-skewed (i.e., corresponding to the above-mentioned right-skewed distribution); when the first skewness is less than zero, the distribution dispersion degree of the group of data located below the second dividing line in the above-mentioned fourth group of data is higher than the distribution dispersion degree of the group of data located above the second dividing line. In this case, the first skewness can also indicate that the direction of the distribution dispersion degree is left-skewed (i.e., corresponding to the above-mentioned left-skewed distribution).

[0035] Optionally, when the multiple parameter pairs present a mixed distribution, the first dividing line is a dividing line determined based on a target quantile, and the target quantile is one of multiple quantiles based on statistics of the multiple parameter pairs; wherein the sum of the absolute value of the skewness of the third group of data and the absolute value of the skewness of the fourth group of data is less than a second threshold, or, the target quantile is the quantile among the multiple quantiles at which the sum of the absolute values ​​of the skewnesses of the two groups of data divided by the multiple parameter pairs corresponding to the multiple quantiles is the smallest.

[0036] For example, from the distribution morphology, the mixed distribution generally has: when the data points are projected on the vertical axis of the two-dimensional distribution plane, the number of data points fluctuates along the vertical axis, neither increasing nor decreasing continuously. From the probability density curve, unlike the probability density curve of the long-tailed distribution, which has only one peak, the probability density curve of the mixed distribution generally has multiple peaks.

[0037] Exemplarily, the target quantile is a quantile with an optimal partitioning result among multiple quantiles based on statistics of the multiple parameter pairs, thereby achieving optimal partitioning for multiple parameter pairs of mixed distribution.

[0038] In the implementation method described above, for multiple parameter pairs presenting a mixed distribution, the multiple parameter pairs are optimally divided and grouped to determine a division method that makes the sum of the absolute values ​​of the skewnesses of the two groups of data after division smaller. In this way, the distributions of the two groups of data after division of the multiple parameter pairs are more symmetrical, so that the parameter range of the determined motion parameters is also more reasonable.

[0039] In a second aspect, the present application provides a device for determining a parameter range of a driving scene, the device comprising: an acquisition unit, used to acquire multiple parameter pairs of the driving scene, each parameter pair comprising a value of a first motion parameter and a value of a second motion parameter, the first motion parameter and the second motion parameter being used to describe the state of a subject in the driving scene; a processing unit, used to determine a parameter range of the first motion parameter based on a regression analysis result of two sets of data that are farthest apart from each other among multiple sets of data; the parameter range of the first motion parameter is used to constrain the value of the first motion parameter in a simulation test of the driving scene; wherein the multiple sets of data are obtained by dividing the multiple parameter pairs on a two-dimensional distribution plane of the multiple parameter pairs by at least one dividing line, each set of data in the multiple sets of data comprises at least one parameter pair in the multiple parameter pairs, the longitudinal axis of the two-dimensional distribution plane indicates the first motion parameter, the horizontal axis of the two-dimensional distribution plane indicates the second motion parameter, the at least one dividing line is determined based on the distribution of the multiple parameter pairs on the two-dimensional distribution plane, and the two sets of data that are farthest apart are respectively located on both sides of any one of the at least one dividing line.

[0040] Optionally, the plurality of parameter pairs present an asymmetric distribution.

[0041] Optionally, the two groups of data include a first group of data and a second group of data, the upper limit of the parameter range of the first motion parameter is determined by an upper boundary in the regression analysis results of the first group of data, and the lower limit of the parameter range of the first motion parameter is determined by a lower boundary in the regression analysis results of the second group of data; wherein, on the two-dimensional distribution plane, the first group of data is located above the second group of data, and the upper boundary is asymmetric to the lower boundary.

[0042] Optionally, an absolute value of a first skewness calculated based on the multiple parameter pairs is greater than zero, and the first skewness is used to indicate a degree of distribution dispersion of the multiple parameter pairs; wherein the first skewness is the third-order central moment of the multiple parameter pairs.

[0043] Optionally, the plurality of parameter pairs are divided into a third group of data and a fourth group of data via a first segmentation line, and a distribution discreteness of the fourth group of data is greater than a distribution discreteness of the third group of data.

[0044] Optionally, the first dividing line is obtained by linear fitting of multiple mean points based on the statistics of the multiple parameter pairs; the first skewness is greater than zero, the third group of data includes parameter pairs among the multiple parameter pairs that are below the first dividing line, and the fourth group of data includes parameter pairs among the multiple parameter pairs that are above the first dividing line; or, the first skewness is less than zero, the fourth group of data includes parameter pairs among the multiple parameter pairs that are below the first dividing line, and the third group of data includes parameter pairs among the multiple parameter pairs that are above the first dividing line.

[0045] Optionally, when the multiple parameter pairs present a long-tailed distribution, the absolute value of the first skewness is greater than a first threshold, and the fourth group of data is divided into a fifth group of data and a sixth group of data by a second dividing line, wherein the distribution discreteness of the sixth group of data is greater than the distribution discreteness of the fifth group of data.

[0046] Optionally, the second dividing line is obtained by linear fitting of multiple mean points based on the statistics of the fourth group of data; the first skewness is greater than zero, the fifth group of data includes parameter pairs in the fourth group of data that are below the second dividing line, and the sixth group of data includes parameter pairs in the fourth group of data that are above the second dividing line; or, the first skewness is less than zero, the sixth group of data includes parameter pairs in the fourth group of data that are below the second dividing line, and the fifth group of data includes parameter pairs in the fourth group of data that are above the second dividing line.

[0047] Optionally, when the multiple parameter pairs present a mixed distribution, the first dividing line is a dividing line determined based on a target quantile, and the target quantile is one of multiple quantiles based on statistics of the multiple parameter pairs; wherein the sum of the absolute value of the skewness of the third group of data and the absolute value of the skewness of the fourth group of data is less than a second threshold, or, the target quantile is the quantile among the multiple quantiles at which the sum of the absolute values ​​of the skewnesses of the two groups of data divided by the multiple parameter pairs corresponding to the multiple quantiles is the smallest.

[0048] In a third aspect, the present application provides a computing device comprising a processor and a memory, wherein the memory is used to store program instructions; the processor calls the program instructions in the memory so that the device executes the method in the first aspect or any possible implementation of the first aspect.

[0049] In a fourth aspect, the present application provides a computer-readable storage medium, comprising computer instructions, which, when executed by a processor, implement the method in the above-mentioned first aspect or any possible implementation manner of the first aspect.

[0050] In a fifth aspect, the present application provides a computer program product, which, when executed by a processor, implements the method in the first aspect or any possible embodiment of the first aspect. The computer program product, for example, may be a software installation package, and when the method provided by any possible design of the first aspect is needed, the computer program product may be downloaded and executed on a processor to implement the method in the first aspect or any possible embodiment of the first aspect.

[0051] The technical effects of the second to fifth aspects are the same as those of the first aspect, and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 It is a schematic diagram of the architecture of a communication system provided by an embodiment of the present application;

[0053] Figure 2 It is a schematic diagram of data distribution after the combination of two motion parameters in a vehicle cut-in scenario provided by an embodiment of the present application;

[0054] Figure 3 is a flow chart of a method for determining a parameter range of a driving scenario provided in an embodiment of the present application;

[0055] Figure 4 It is a schematic diagram of dividing a plurality of parameter pairs on a two-dimensional distribution plane provided in an embodiment of the present application;

[0056] Figure 5AIt is a schematic diagram of dividing a plurality of parameter pairs on a two-dimensional distribution plane provided in an embodiment of the present application;

[0057] Figure 5B It is another schematic diagram of dividing multiple parameter pairs on a two-dimensional distribution plane provided in an embodiment of the present application;

[0058] Figure 5C is a schematic diagram of a parameter range of a first motion parameter provided in an embodiment of the present application;

[0059] Figure 6 It is a schematic diagram of dividing multiple parameter pairs of some deceleration scenarios provided in an embodiment of the present application by a dividing line determined based on quantiles;

[0060] Figure 7 is a schematic diagram of a parameter range of a first motion parameter provided in an embodiment of the present application;

[0061] Figure 8 It is a structural schematic diagram of a parameter range analysis device provided in an embodiment of the present application;

[0062] Fig. 9 It is a structural diagram of a computing device provided in this embodiment of the present application. DETAILED DESCRIPTION

[0063] It should be noted that the prefixes such as "first" and "second" used in this application are only used to distinguish different description objects, and have no limiting effect on the position, order, priority, quantity or content of the described objects. For example, if the described object is a "field", the ordinal number before the "field" in the "first field" and the "second field" does not limit the position or order between the "fields". "First" and "second" do not limit whether the "fields" they modify are in the same message, nor do they limit the order of the "first field" and the "second field". For another example, if the described object is a "level", the ordinal number before the "level" in the "first level" and the "second level" does not limit the priority between the "levels". For another example, the number of described objects is not limited by the prefix and can be one or more. Taking "first device" as an example, the number of "devices" can be one or more. In addition, the objects modified by different prefixes can be the same or different. For example, if the object being described is a "device", then the "first device" and the "second device" can be the same device, the same type of device, or different types of devices; for another example, if the object being described is "information", then the "first information" and the "second information" can be information of the same content or information of different contents. In short, the use of prefixes used to distinguish the described objects in the embodiments of the present application does not constitute a limitation on the described objects. For the statement of the described objects, refer to the description in the context of the claims or embodiments, and no unnecessary limitation should be constituted due to the use of such prefixes.

[0064] It should be noted that the description methods such as "at least one of a1, a2, ... and an" used in the embodiments of the present application include the situation where any one of a1, a2, ... and an exists alone, and also include any combination of any multiple of a1, a2, ... and an, and each situation can exist alone. For example, the description method of "at least one of a, b and c" includes the situation where a is alone, b is alone, c is alone, a combination of a and b, a combination of a and c, b and c, or a combination of abc.

[0065] The technical solution in this application will be described below in conjunction with the accompanying drawings.

[0066] See also Figure 1 , Figure 1 is a schematic diagram of the architecture of a communication system provided in an embodiment of the present application. The communication system is used to determine the parameter range of the motion parameter of the driving scene. Figure 1 As shown, the communication system includes a data source device and a network side device, wherein the data source device and the network side device can be connected and communicated wirelessly or wiredly. In a possible implementation, the data collected by the data source device can also be transferred to the network side device via a removable storage medium such as a mobile hard disk.

[0067] Here, the data source device may be a drone equipped with a collection device, a data collection vehicle, a roadside device, surrounding vehicles, etc. The roadside device may be, for example, a road side unit (RSU), multi-access edge computing (MEC), or a sensor, or a component or chip inside these devices, or a system-level device composed of an RSU and a MEC, or a system-level device composed of an RSU and a sensor, or a system-level device composed of an RSU, a MEC, and a sensor.

[0068] Here, the network side device can be, for example, a server deployed on the network side (such as a parameter range analysis server, a simulation test server, etc.), or a component or chip in the server, or a system-level device composed of multiple servers. The network side device can be deployed in a cloud environment or an edge environment, which is not specifically limited here.

[0069] Exemplarily, the data source device is used to collect the trajectory data of the vehicle traveling on the road, and provide the collected trajectory data to the network side device. The network side device processes the trajectory data collected by the data source device to obtain the actual data of each motion parameter of the driving scene. In some possible embodiments, the vehicle trajectory data may also come from the accident database, traffic supervision database, etc. of the traffic supervision department, which is not specifically limited here.

[0070] Here, the driving scene may be, for example, a scene where the front vehicle cuts in, a scene where the front vehicle cuts out, a scene where the front vehicle follows, or a scene where the front vehicle brakes.

[0071] Here, motion parameters can also be called scene parameters or vehicle behavior parameters. Motion parameters are used to describe the state of the subject (such as vehicles, pedestrians, and any other possible traffic participants) in the driving scene. Operation parameters include, for example, at least one of the parameters such as speed, distance, relative speed, relative distance, displacement, acceleration, angular velocity, posture, lane change time, etc.

[0072] Each driving scenario has its own corresponding motion parameters. Taking the scene of the vehicle cutting in as an example, assuming that vehicle 1 is the main vehicle and vehicle 2 is the cutting vehicle that will cut in front of vehicle 1, the motion parameters corresponding to this driving scenario include: the speed of vehicle 1, the lateral speed of vehicle 2, the longitudinal distance of vehicle 2 relative to vehicle 1, and the longitudinal speed of vehicle 2 relative to vehicle 1.

[0073] Exemplarily, the process of determining the parameter range of the motion parameter by the network side device may be: obtaining multiple parameter pairs of the driving scene, each parameter pair including the value of the first motion parameter and the value of the second motion parameter, further, the network side device divides the multiple parameter pairs into multiple groups of data on the two-dimensional distribution plane of the multiple parameter pairs by at least one dividing line pair, the vertical axis of the two-dimensional distribution plane indicates the first motion parameter and the horizontal axis indicates the second motion parameter, based on the regression analysis results of the two groups of data that are farthest apart in the multiple groups of data, the parameter range of the first motion parameter is determined, and the parameter range of the first motion parameter is used to generate the data of the first motion parameter used in the simulation test of the driving scene. The process of obtaining the parameter range of the first motion parameter can be specifically referred to the description of the following method embodiment, which will not be repeated here.

[0074] It can be understood that multiple parameter pairs of the driving scene can be obtained from the actual data of each motion parameter of the above-mentioned driving scene.

[0075] here, Figure 1 The system architecture shown is only an example and is not intended to limit Figure 1 In addition, the method provided in the embodiment of the present application can be applied not only to Figure 1The system shown may also be applicable to other systems. For example, the communication system further includes an intermediate processing device, which performs an operation of processing the trajectory data collected by the data source device to obtain the actual data of each motion parameter of the driving scene, without being performed by the network side device. In this case, the network side device can obtain multiple parameter pairs of the driving scene from the intermediate processing device.

[0076] Here, the parameter range determination method provided in the embodiment of the present application is not only applicable to autonomous driving simulation systems, but also to other simulation test systems that require the parameter range of motion parameters, such as model-in-the-loop test systems, hardware-in-the-loop test systems, etc.

[0077] Here, "autonomous driving" is not limited to fully autonomous driving systems, highly autonomous driving systems, conditional autonomous driving systems, or partially autonomous driving systems. Those skilled in the art can understand that non-fully manual driving systems that provide intelligent driving can be included in this concept for simulation processing.

[0078] In an embodiment of the present application, there may be multiple motion parameters for a driving scene. Considering the correlation between the motion parameters, the motion parameters of the driving scene may be combined in pairs, with one motion parameter being used as an independent variable and the other motion parameter being used as a dependent variable. For the data distribution of each of the two combined motion parameters on a two-dimensional plane, a parameter range of the motion parameter used as the dependent variable may be determined, and the parameter range of the motion parameter may be used to constrain the value of the motion parameter in the simulation test of the driving scene. In some possible embodiments, the motion parameters are not limited to being combined in pairs. For example, the correlation between three motion parameters may also be considered. In this case, one of the motion parameters is used as the dependent variable and the other two motion parameters are used as independent variables.

[0079] It can be understood that when the number of motion parameters of a driving scene is greater than two, there are multiple situations in which two motion parameters are combined from the multiple motion parameters. Taking the driving scene as a vehicle cut-in scene as an example, the motion parameters of the vehicle cut-in scene include four motion parameters: the main vehicle speed, the lateral speed of the cut-in vehicle, the longitudinal distance of the cut-in vehicle relative to the main vehicle, and the longitudinal speed of the cut-in vehicle relative to the main vehicle. These four motion parameters are combined in pairs. In each combination, one motion parameter is used as an independent variable (i.e., indicated by the horizontal axis) and the other motion parameter is used as a dependent variable (i.e., indicated by the vertical axis). There are a total of 12 combinations.

[0080] See also Figure 2 , Figure 2 is a schematic diagram of data distribution after the combination of two motion parameters in a vehicle cut-in scenario provided by an embodiment of the present application. Figure 2It can be seen that in the vehicle cut-in scenario, the motion parameters include the main vehicle speed and the lateral speed of the cut-in vehicle ( Figure 2 The longitudinal distance of the cutting vehicle relative to the main vehicle ( Figure 2 The relative distance in the figure) and the longitudinal speed of the cutting vehicle relative to the main vehicle ( Figure 2 These four motion parameters are combined in pairs to draw data distribution diagrams. There are 12 combinations in total. For details, see Figure 2 There are 12 pictures listed in it.

[0081] Below Figure 2 Take (1) in the example for illustration. Figure 2 In (1), the horizontal axis of the two-dimensional distribution plane represents the “host vehicle speed” and the vertical axis represents the “relative speed”. Figure 2 (1) shows the data distribution when "main vehicle speed" is the independent variable and "relative speed" is the dependent variable. Figure 2 From the data points in (1), we can see that in the vertical direction, the data points in the upper half are sparsely distributed while the data points in the lower half are densely distributed. It can be understood that Figure 2 The coordinates of each data point in (1) are composed of a value of two motion parameters, "main vehicle speed" and "relative speed", wherein the horizontal coordinate of the data point is a value of the "main vehicle speed" and the vertical coordinate of the data point is a value of the "relative speed".

[0082] Understandably, Figure 2 The dense data points in each picture are only an example of the distribution form of the corresponding data in the vehicle cut-in scenario, and do not limit the distribution form of the data to only Figure 2 shown.

[0083] In the embodiment of the present application, the above Figure 2 The parameter range of the motion parameter as the dependent variable in each figure can be obtained by Figure 3 The method provided in the embodiment is used to determine. It can be understood that Figure 2 The parameter range of the motion parameter determined in each figure can be used to constrain the value of the corresponding motion parameter in the simulation test of the driving scene. In this way, a reasonable parameter range is provided for the simulation test of the driving scene, which is conducive to improving the test efficiency.

[0084] See also Figure 3 , Figure 3 is a flow chart of a method for determining a parameter range of a driving scenario provided by an embodiment of the present application. The method can be applied, for example, to Figure 1 The network side device shown. The method includes but is not limited to the following steps:

[0085] S301: Acquire multiple parameter pairs of a driving scene, each parameter pair including a value of a first motion parameter and a value of a second motion parameter.

[0086] Here, the driving scene includes but is not limited to a vehicle cut-in scene, a vehicle cut-out scene, a vehicle following scene, a front vehicle braking scene, a deceleration scene, etc. Exemplarily, the vehicle cut-in scene can also be divided into a small vehicle cut-in scene and a large vehicle cut-in scene.

[0087] The motion parameters include but are not limited to at least one of speed, distance, relative speed, relative distance, displacement, acceleration, angular velocity, posture, lane change time and the like.

[0088] For example, different driving scenarios may have different corresponding motion parameters. For example, the motion parameters in the vehicle cut-in scenario include the main vehicle speed, the lateral speed of the cut-in vehicle, the longitudinal distance of the cut-in vehicle relative to the main vehicle, and the longitudinal speed of the cut-in vehicle relative to the main vehicle. For another example, the motion parameters in the deceleration scenario include the main vehicle speed, the relative speed of the two vehicles, the deceleration, the jerk (or acceleration change rate), and the relative distance between the two vehicles.

[0089] Here, the first motion parameter and the second motion parameter may be any two motion parameters in the driving scenario.

[0090] It can be seen that a parameter pair includes two values, namely, a value of the first motion parameter and a value of the second motion parameter. The above-mentioned multiple parameter pairs can be obtained by combining the values ​​of the two motion parameters (i.e., the first motion parameter and the second motion parameter) in the driving scene. It can be understood that each parameter pair represents a data point on the two-dimensional distribution plane.

[0091] For example, Figure 2 In (1), the first motion parameter is the “main vehicle speed” and the second motion parameter is the “relative speed”. Since the first reference data includes multiple parameter pairs, each parameter pair includes a value of the “main vehicle speed” and a value of the “relative speed”. Each parameter pair represents Figure 2 A data point in (1) of Figure 2 The number of data points in (1) is the number of parameter pairs.

[0092] In one implementation, the plurality of parameter pairs are distributed asymmetrically.

[0093] It can be understood that an asymmetric distribution is the opposite of a symmetric distribution (e.g., Gaussian distribution, uniform distribution, etc.). Exemplarily, from the perspective of distribution morphology, multiple parameter pairs presenting asymmetric parts can be understood as the multiple parameter pairs showing a distribution morphology in which one part is sparse and the other part is dense on a two-dimensional distribution plane. From the perspective of probability density curves, the multiple parameter pairs presenting asymmetric distribution can be understood as the probability density curves obtained based on the multiple parameter pairs being asymmetric on the left and right.

[0094] Here, asymmetric distribution includes but is not limited to long-tail distribution, mixed distribution, etc.

[0095] Exemplarily, from the distribution morphology, the long-tail distribution generally has: when the data points are projected on the longitudinal axis of the two-dimensional distribution plane, for example, the number of data points shows an increasing or decreasing trend along the longitudinal axis. From the probability density curve, the probability density curve of the long-tail distribution is similar to a bell curve but has a long tail on the left or right side of the curve. Based on the different locations of the long tail, the long-tail distribution can be further divided into a left-skewed distribution and a right-skewed distribution. Among them, the left-skewed distribution can also be called a negatively skewed distribution, and the right-skewed distribution can also be called a positively skewed distribution.

[0096] For example, from the distribution morphology, the mixed distribution generally has: when the data points are projected on the vertical axis of the two-dimensional distribution plane, the number of data points fluctuates along the vertical axis, neither increasing nor decreasing continuously. From the probability density curve, unlike the probability density curve of the long-tailed distribution, which has only one peak, the probability density curve of the mixed distribution generally has multiple peaks.

[0097] In one implementation, when the plurality of parameter pairs present an asymmetric distribution, the absolute value of the first skewness calculated based on the plurality of parameter pairs is greater than zero. The first skewness is used to indicate the degree of distribution dispersion of the plurality of parameter pairs.

[0098] Exemplarily, the first skewness is the third-order center distance of the plurality of parameter pairs. The first skewness can be expressed, for example, by the following formula (1).

[0099]

[0100] Wherein, X represents the above-mentioned multiple parameter pairs, S(X) represents the skewness corresponding to the multiple parameter pairs, that is, the first skewness, E() represents the expectation, μ represents the mean of the multiple parameter pairs, and σ represents the standard deviation of the multiple parameter pairs. It can be understood that formula (1) is only an example of the representation of skewness, and the representation of skewness is not limited to that shown in formula (1), and can also be some deformation based on formula (1).

[0101] Furthermore, the first skewness can also indicate the direction of the distribution discreteness of the multiple parameter pairs. For example, when the multiple parameter pairs present a long-tailed distribution, if S(X)>0, it indicates that the direction of the distribution discreteness of the multiple parameter pairs is right-skewed, indicating that the right side of the probability density curve obtained based on the multiple parameter pairs has a long tail, and also indicates that the multiple parameter pairs specifically present a right-skewed distribution; if S(X)<0, it indicates that the direction of the distribution discreteness of the multiple parameter pairs is left-skewed, indicating that the left side of the probability density curve obtained based on the multiple parameter pairs has a long tail, and also indicates that the multiple parameter pairs specifically present a left-skewed distribution.

[0102] S302: Determine the parameter range of the first motion parameter based on the regression analysis results of two groups of data that are farthest apart from each other among the multiple groups of data, wherein the multiple groups of data are obtained by dividing the multiple parameter pairs on the two-dimensional distribution plane of the multiple parameter pairs by at least one dividing line, each group of data in the above multiple groups of data includes at least one parameter pair among the multiple parameter pairs, the vertical axis of the two-dimensional distribution plane indicates the first motion parameter and the horizontal axis indicates the second motion parameter, and the above two groups of data that are farthest apart from each other are respectively located on both sides of any of the above dividing lines.

[0103] Here, the at least one segmentation line is determined based on the distribution of the multiple parameter pairs on the two-dimensional distribution plane.

[0104] For example, a dividing line can divide the multiple parameter pairs into two groups of data on a two-dimensional distribution plane, one group of data is located on one side of the dividing line, and the other group of data is located on the other side of the dividing line. Two dividing lines can divide the multiple parameter pairs into three groups of data on a two-dimensional distribution plane.

[0105] Exemplarily, since the first motion parameter is indicated by the longitudinal axis of the two-dimensional distribution plane, each segmentation line can divide the multiple parameter pairs into upper and lower parts on the two-dimensional distribution plane. When there are multiple segmentation lines, the multiple segmentation lines are arranged in sequence along the longitudinal axis direction on the two-dimensional distribution plane. In this case, the two groups of data that are farthest apart from each other in the multiple groups of data obtained by dividing the multiple segmentation lines include: a group of data consisting of parameter pairs located above the uppermost segmentation line on the two-dimensional distribution plane, and a group of data consisting of parameter pairs located below the lowermost segmentation line on the two-dimensional distribution plane.

[0106] Here, the dividing line may be, for example, a mean line corresponding to the parameter pair to be analyzed, or a dividing line determined based on a quantile, wherein the quantile is determined based on statistics of the parameter pair to be analyzed.

[0107] In one implementation, the two groups of data that are farthest apart from each other in the above-mentioned multiple groups of data include the first group of data and the second group of data, then the upper limit of the parameter range of the first motion parameter is determined by the upper limit in the regression analysis result of the first group of data, and the lower limit of the parameter range of the first motion parameter is determined by the lower limit in the regression analysis result of the second group of data, wherein the upper limit corresponding to the first group of data is asymmetric to the lower limit corresponding to the second group of data. From the two-dimensional distribution plane, the first group of data is located above the second group of data, and for any dividing line, the first group of data is located on the upper side of the dividing line, and the second group of data is located on the lower side of the dividing line.

[0108] Exemplarily, the upper limit corresponding to the first set of data is determined based on the statistical mean line of the first set of data and the standard deviation of a preset multiple (e.g., 3 times) above the mean line. The lower limit corresponding to the second set of data is determined based on the statistical mean line of the second set of data and the standard deviation of a preset multiple (e.g., 3 times) below the mean line.

[0109] In one implementation, the plurality of parameter pairs include only two groups of data, for example, a first group of data and a second group of data, and the value range of the first motion parameter is determined according to the regression analysis results of the first group of data and the regression analysis results of the second group of data, wherein the distribution discreteness of the second group of data is greater than the distribution discreteness of the first group of data.

[0110] With the above Figure 2 The multiple parameter pairs shown in (1) are used as an example for division. The division results can be referred to Figure 4 . Figure 4 Schematic diagram of the division of multiple parameter pairs on a two-dimensional distribution plane provided in an embodiment of the present application. Figure 4 It can be seen that the mean line divides these multiple parameter pairs into upper and lower parts, where the parameter pairs on one side of the mean line constitute the first group of data, and the parameter pairs on the other side of the mean line constitute the second group of data. Here, the mean line is based on Figure 2 The linear fitting method of the mean line can be, for example, as follows: the data points (i.e., parameter pairs) are divided into multiple intervals of equal length according to the main vehicle speed on the horizontal axis, and the mean of the horizontal coordinate and the mean of the vertical coordinate of the data points in each interval are calculated respectively to obtain multiple horizontal coordinate means and multiple vertical coordinate means, and the linear fitting is performed with the multiple horizontal coordinate means as independent variables and the multiple vertical coordinate means as dependent variables to obtain the mean line.

[0111] exist Figure 4In the case where multiple parameter pairs are divided into upper and lower groups of data by a mean line, the upper limit of the parameter range of "relative speed" is determined by the upper limit of the regression analysis results of the group of data above the mean line, and the lower limit of the parameter range of "relative speed" is determined by the lower limit of the regression analysis results of the group of data below the mean line. Here, the generation of the upper and lower limits can refer to the description of the corresponding content above. It can be understood that Figure 4 The upper and lower bounds in the parameter range are determined only based on Figure 4 The multiple parameter pairs shown determine the parameter range of "relative speed".

[0112] Furthermore, since the distribution dispersion of the second group of data is greater than that of the first group of data, if the first skewness calculated based on multiple parameter pairs is greater than zero, it means that the distribution dispersion of the group of data above the mean line is higher, so the second group of data includes the parameter pairs above the mean line among the multiple parameter pairs, and the first group of data includes the parameter pairs below the mean line among the multiple parameter pairs; if the first skewness is less than zero, it means that the distribution dispersion of the group of data below the mean line is higher, so the second group of data includes the parameter pairs below the mean line among the multiple parameter pairs, and the first group of data includes the parameter pairs above the mean line among the multiple parameter pairs, wherein the mean line is obtained by linear fitting based on the multiple parameter pairs.

[0113] Understandably, Figure 4 This is just an example of dividing the multiple parameter pairs on the two-dimensional distribution plane, and does not limit the division form of the multiple parameter pairs to only Figure 4 In some possible embodiments, Figure 4 The dividing line in may not be the mean line, but other dividing lines determined based on the quantiles. Alternatively, the plurality of parameter pairs are not limited to being divided into only two groups, but may be divided into three groups, four groups, or other numbers of groups.

[0114] In some possible embodiments, when the multiple parameter pairs present an asymmetric distribution, the asymmetric distribution includes a long-tailed distribution, a mixed distribution, etc. Based on the different distribution types of the multiple parameter pairs, the division methods of the multiple parameter pairs may also be different. The embodiments of the present application mainly provide the division methods of the multiple parameter pairs under two distribution types.

[0115] Type 1: The above parameters show a long tail distribution

[0116] For example, based on the position of the long tail on the probability density curve, the long tail distribution can be divided into positive skewed distribution and negative skewed distribution. Here, the long tail distribution can refer to the description of the corresponding content in S301, which will not be repeated here.

[0117] In this case, if the absolute value of the first skewness calculated based on the multiple parameter pairs is greater than the first threshold, the multiple parameter pairs can be divided based on the first skewness to obtain the above multiple groups of data.

[0118] Here, the first threshold is greater than zero, and the first threshold may be a default setting or a user-defined setting.

[0119] This means that when the absolute value of the first skewness is less than or equal to the first threshold value, the asymmetry of the multiple parameter pairs is within an acceptable range or can be ignored. When the absolute value of the first skewness is greater than the threshold value, the multiple parameter pairs are divided.

[0120] In one implementation, when the above-mentioned multiple parameter pairs present a long-tail distribution, if the absolute value of the above-mentioned first skewness is greater than the first threshold, the multiple parameter pairs can be divided into three groups of data on the two-dimensional distribution plane of the multiple parameter pairs through two dividing lines based on the first skewness; and the parameter range of the first motion parameter is determined according to the regression analysis results of the two groups of data with the farthest distance among the three groups of data.

[0121] Exemplarily, the process of dividing the multiple parameter pairs into three groups of data through two dividing lines based on the first skewness can be: firstly, the multiple parameter pairs are divided into two groups of data through dividing line 1, namely, group 1 data and group 2 data, wherein the distribution discreteness of group 2 data is greater than the distribution discreteness of group 1 data. Then, the second group of data with a higher distribution discreteness is further divided into two groups of data through dividing line 2, namely, group 3 data and group 4 data, wherein the distribution discreteness of group 4 data is greater than the distribution discreteness of group 3 data. It can be seen that the multiple parameter pairs are divided into the above-mentioned group 1 data, group 3 data and group 4 data through two dividing lines. That is to say, for multiple parameter pairs presenting a long tail distribution, the multiple parameter pairs are first divided into two parts, and then the reference data of the part with a higher distribution discreteness of the two parts is divided at least once, so that it is conducive to accurately estimating the parameter range of the first motion parameter so that it fits the actual distribution of the multiple parameter pairs.

[0122] The first, third and fourth groups of data are explained below in combination with the positive and negative first skewness calculated based on the multiple parameter pairs, wherein the dividing line 1 is obtained by linear fitting of multiple mean points based on the multiple parameter pairs, and the dividing line 2 is obtained by linear fitting of multiple mean points of the second group of data with a higher degree of distribution dispersion. In this case, the dividing line 1 can be called the mean line 1, and the dividing line 2 can be called the mean line 2:

[0123] (1) If the first skewness is greater than zero (i.e., it indicates a positively skewed distribution), since the distribution dispersion of the second group of data is greater than that of the first group of data, the first group of data includes the parameter pairs among the multiple parameter pairs that are below the mean line 1, and the second group of data includes the parameter pairs among the multiple parameter pairs that are above the mean line 1. Correspondingly, since the distribution dispersion of the fourth group of data is greater than that of the third group of data, the third group of data includes the parameter pairs among the second group of data that are below the mean line 2, and the fourth group of data includes the parameter pairs among the second group of data that are above the mean line 2.

[0124] In this case, the two groups of data that are farthest apart among the three groups of data are the first group of data and the fourth group of data, among which the fourth group of data is located above the first group of data, so the upper limit of the parameter range of the first motion parameter is determined by the upper limit in the regression analysis results of the fourth group of data, and the lower limit of the parameter range of the first motion parameter is determined by the lower limit in the regression analysis results of the first group of data.

[0125] (2) If the first skewness is less than zero (i.e., it indicates a negatively skewed distribution), since the distribution dispersion of the second group of data is greater than that of the first group of data, the first group of data includes the parameter pairs among the multiple parameter pairs that are above the mean line 1, and the second group of data includes the parameter pairs among the multiple parameter pairs that are below the mean line 1. Correspondingly, since the distribution dispersion of the fourth group of data is greater than that of the third group of data, the third group of data includes the parameter pairs among the second group of data that are above the mean line 2, and the fourth group of data includes the parameter pairs among the second group of data that are below the mean line 2.

[0126] In this case, the two groups of data that are farthest apart among the three groups of data are the first group of data and the fourth group of data, among which the first group of data is located above the fourth group of data, so the upper limit of the parameter range of the first motion parameter is determined by the upper limit in the regression analysis results of the first group of data, and the lower limit of the parameter range of the first motion parameter is determined by the lower limit in the regression analysis results of the fourth group of data.

[0127] Here, when multiple parameter pairs present a long tail distribution, it is not limited that the multiple parameter pairs can only be divided into three groups. In some possible embodiments, the multiple parameter pairs can also be divided into four groups, five groups or other more groups, which are not specifically limited here.

[0128] Assumptions Figure 2 (1) shows a long-tailed distribution. Figure 2 Take (1) as an example Figure 2 The method for determining the parameter range of "relative speed" in (1) can refer to the following steps A11-A14 for the specific process:

[0129] A11: Based on Figure 2The first skewness S is calculated from the multiple parameter pairs shown in (1), assuming that the absolute value of S is greater than the first threshold and S is greater than 0.

[0130] A12: Based on the dividing line 1, the plurality of parameter pairs are divided into two groups, namely, the first group of data and the second group of data, wherein the distribution discreteness of the second group of data is greater than the distribution discreteness of the first group of data.

[0131] For example, the results of dividing these multiple parameter pairs by the dividing line 1 can be seen in Figure 5A shown. Figure 5A Schematic diagram of the division of multiple parameter pairs on a two-dimensional distribution plane provided in an embodiment of the present application. Figure 5A In the embodiment, the first group of data includes the parameter pairs located below the dividing line 1 among the multiple parameter pairs, and the second group of data includes the parameter pairs located above the dividing line 1 among the multiple parameter pairs.

[0132] exist Figure 5A , dividing line 1 is the mean line y 1 , mean line y 1 Please refer to the description of the above mean line 1. Figure 5A The mean line y shown 1 The calculation method of y can be, for example: divide the multiple parameter pairs into 8 equal-length intervals according to the main vehicle speed shown on the horizontal axis on the second distribution plane, calculate the horizontal and vertical coordinate means of each data point in each interval (each data point corresponds to a parameter pair), so as to obtain 8 horizontal coordinate means and 8 vertical coordinate means, and perform linear fitting with the 8 horizontal coordinate means as independent variables and the 8 vertical coordinate means as dependent variables to obtain the mean line y 1 .

[0133] A13: Based on the dividing line 2, the second group of data is divided into two groups of data, namely the third group of data and the fourth group of data, wherein the distribution dispersion degree of the fourth group of data is greater than the distribution dispersion degree of the third group of data.

[0134] For example, Figure 5A The result of dividing the second group of data by dividing line 2 can be seen in Figure 5B shown. Figure 5B FIG. 1 is another schematic diagram of dividing multiple parameter pairs on a two-dimensional distribution plane provided in an embodiment of the present application. Figure 5B It can be seen that the multiple parameter pairs are divided into the first group of data and the second group of data by the dividing line 1, wherein the parameter pairs in the first group of data are Figure 5BThe two-dimensional distribution plane shown is represented by a five-pointed star, and the parameter pairs in the second group of data are represented by dots on the two-dimensional distribution plane. Further, the second group of data is further divided into a third group of data and a fourth group of data by the dividing line 2, wherein the third group of data includes the parameter pairs in the second group of data located below the dividing line 2, or the third group of data includes Figure 5B The fourth group of data includes the parameter pairs located above the dividing line 2 in the second group of data. In order to distinguish the third group of data from the fourth group of data, the parameter pairs in the third group of data are represented by a plus sign on the two-dimensional distribution plane, and the parameter pairs in the fourth group of data are represented by a five-pointed star on the two-dimensional distribution plane. For details, please refer to the following Figure 5C .

[0135] exist Figure 5B , the dividing line 2 can be the mean line y 2 , mean line y 2 Please refer to the description of the above mean line 2, which will not be repeated here.

[0136] A14: Based on the regression analysis results of the first group of data and the regression analysis results of the fourth group of data, determine the parameter range of "relative speed".

[0137] Depend on Figure 5B It can be seen that the first and fourth groups of data are Figure 5B The two data sets that are the farthest apart from each other among the three data sets shown. The upper boundary in the regression analysis results of the first data set is used to determine the upper limit of the parameter range of "relative speed", and the lower boundary in the regression analysis results of the fourth data set is used to determine the lower limit of the parameter range of "relative speed".

[0138] Exemplarily, the fourth group of data is subjected to regression analysis, including: for the fourth group of data, the fourth group of data is divided into 8 intervals of equal length according to the main vehicle speed shown on the horizontal axis on the two-dimensional distribution plane, the horizontal and vertical coordinate means and standard deviations of each data point (each data point corresponds to a parameter pair) in each interval are calculated, and linear fitting is performed to obtain the mean line y μ4 and standard deviation line y σ4 , then the upper limit corresponding to the fourth group of data can be expressed as y 上 =y μ4 +ny σ4 , n is 3, for example.

[0139] Accordingly, the first group of data is subjected to regression analysis, including: for the first group of data, the first group of data is divided into 8 intervals of equal length according to the main vehicle speed shown on the horizontal axis on the two-dimensional distribution plane, the horizontal and vertical coordinate means and standard deviations of each data point (each data point corresponds to a parameter pair) in each interval are calculated, and linear fitting is performed to obtain the mean line y μ1and standard deviation line y σ1 , then the lower limit corresponding to the first group of data can be expressed as y 下 =y μ1 -ny σ1 , n is 3, for example.

[0140] Finally, the parameter range of "relative speed" determined by the upper limit corresponding to the fourth group of data and the lower limit corresponding to the first group of data is: μ1 -ny σ1 ≤Relative speed≤y μ4 +ny σ4 The parameter range of "relative speed" determined by the above method can be referred to Figure 5C As shown by Figure 5C It can be seen that the upper limit of the parameter range of "relative speed" is determined by the upper limit corresponding to the fourth group of data (i.e., the above y 上 ) is determined by the lower limit of the parameter range of “relative speed”, and the lower limit of the parameter range of “relative speed” is determined by the lower limit corresponding to the first group of data (i.e. the above y 下 ) is determined by Figure 5C It can also be seen that the upper boundary corresponding to the fourth group of data is asymmetric to the lower boundary corresponding to the first group of data, but the parameter range of "relative speed" determined in this way preserves the correlation between motion parameters while also preserving the actual possible value space of the motion parameters as much as possible, which is more in line with the distribution of actual data of motion parameters in driving scenarios.

[0141] It can be understood that the above steps A11-A14 are only an example of processing multiple parameter pairs presenting positive skewed distribution (a kind of long-tailed distribution), and the processing of multiple parameter pairs presenting positive skewed distribution is not limited to only steps A11-A1. In some possible embodiments, the multiple parameter pairs presenting positive skewed distribution can also be divided into four groups or more groups before determining the parameter range of the first motion parameter. In addition, the processing of multiple parameter pairs presenting negative skewed distribution can also be performed similarly with reference to steps A11-A14.

[0142] Type 2: Multiple parameter pairs present mixed distribution

[0143] Here, the mixed distribution can refer to the description of the corresponding content in S301, which will not be repeated here.

[0144] In this case, the above-mentioned multiple parameter pairs are divided into the first group of data and the second group of data by a segmentation line determined based on the target quantile, wherein the target quantile is one of the multiple quantiles based on the statistics of the multiple parameter pairs; wherein the sum of the absolute value of the skewness of the first group of data and the absolute value of the skewness of the second group of data is less than the second threshold, or the target quantile is the quantile with the smallest sum of the absolute values ​​of the skewness of the two groups of data after the multiple parameter pairs are divided. In this case, the parameter range of the first motion parameter can be determined based on the regression analysis results of the first group of data and the regression analysis results of the second group of data. It can be seen that for multiple parameter pairs presenting a mixed distribution, the multiple parameter pairs are optimally divided and grouped to determine a division method that makes the sum of the absolute values ​​of the skewness of the two groups of data after the division smaller, so that the distribution of the two groups of data after the multiple parameter pairs are divided is more symmetrical, so that the parameter range of the determined motion parameter is more reasonable.

[0145] Exemplarily, the multiple quantiles based on the statistics of the multiple parameter pairs include but are not limited to the 0.25 quantile, 0.5 quantile and 0.75 quantile of the multiple parameter pairs. The multiple quantiles can be counted, for example, by a box plot (or box graph), and the 0.25 quantile is the Q1 quantile of the box plot, the 0.5 quantile is the Q2 quantile of the box plot, and the 0.75 quantile is the Q3 quantile of the box plot.

[0146] In one implementation, each quantile of the above statistics is traversed, and quantile 1 is taken as an example. The multiple parameter pairs are divided into two groups of data by a dividing line determined based on quantile 1, and the sum of the absolute values ​​of the skewnesses of the two groups of data is calculated. If the sum of the absolute values ​​of the skewnesses of the two groups of data is less than the second threshold, quantile 1 is used as the above target quantile and the traversal is terminated; if the sum of the absolute values ​​of the skewnesses of the two groups of data is greater than or equal to the second threshold, the remaining quantiles are judged in turn in the same way until the sum of the absolute values ​​of the skewnesses of the two groups of data after the multiple parameter pairs are divided by a dividing line determined based on a certain quantile is less than the second threshold.

[0147] In some possible embodiments, for each quantile of the above statistics, the sum of the absolute values ​​of the skewness of the two groups of data after the above-mentioned multiple parameter pairs are divided by the dividing line determined based on the corresponding quantile can also be calculated in sequence, and the quantile with the smallest sum of the absolute values ​​of the skewness of the two groups of data after the division is taken as the above-mentioned target quantile. It can be understood that the above-mentioned two groups of data after the multiple parameter pairs are divided by the dividing line determined based on the quantile are taken as the above-mentioned first group of data and second group of data, that is, the target quantile is the quantile with the best division result among the multiple quantiles, thereby realizing the optimal division of multiple parameter pairs for the mixed distribution.

[0148] In some possible embodiments, it is also possible to traverse each value of the first motion parameter in multiple parameter pairs. The dividing line can, for example, be a dividing line whose ordinate is the value of the first motion parameter. Taking the value of the first motion parameter as the first value, if the dividing line whose ordinate is the first value divides the multiple parameter pairs into two groups of data, the sum of the absolute values ​​of the skewnesses of the two groups of data is calculated. If the sum of the absolute values ​​of the skewnesses of the two groups of data is less than the second threshold, the division result corresponding to the dividing line whose ordinate is the first value is used as the above-mentioned first group of data and second group of data, and the traversal is terminated; if the sum of the absolute values ​​of the skewnesses of the two groups of data is less than the second threshold or greater than or equal to the second threshold, the remaining values ​​of the first motion parameter are judged in turn in the same way until the sum of the absolute values ​​of the skewnesses of the two groups of data after division corresponding to a certain value is less than the second threshold. In some possible embodiments, the sum of the absolute values ​​of the skewness of the two divided groups of data corresponding to each value of the first motion parameter in the multiple parameter pairs can be calculated, and the two divided groups of data corresponding to the minimum value of the sum of the absolute values ​​can be used as the above-mentioned first group of data and second group of data.

[0149] For example, when the driving scene is a deceleration scene, taking the two motion parameters "jerkiness" and "relative distance between two vehicles" in the deceleration scene as an example, each parameter pair in the multiple parameter pairs of the deceleration scene includes a value of the two motion parameters "jerkiness" and "relative distance between two vehicles", and it is assumed that the multiple parameter pairs of the deceleration scene present a mixed distribution. Figure 6 , Figure 6 The two-dimensional distribution plane of multiple parameter pairs of the deceleration scenario is shown by Figure 6 It can be seen that the vertical axis of the two-dimensional distribution plane indicates the "relative distance" (i.e., the relative distance between the two vehicles, equivalent to the first motion parameter mentioned above), and the horizontal axis of the two-dimensional distribution plane indicates the "jerk" (equivalent to the second motion parameter mentioned above). In this case, the process of determining the parameter range of the "relative distance between the two vehicles" can refer to the following steps B11-B13:

[0150] B11: Determine multiple quantiles of statistics based on multiple parameters of the deceleration scenario.

[0151] Exemplarily, the plurality of quantiles include the 0.25 quantile (ie, the Q1 quantile of the box plot), the 0.5 quantile (ie, the Q2 quantile of the box plot), and the 0.75 quantile (ie, the Q3 quantile of the box plot) of the plurality of parameter pairs.

[0152] exist Figure 6 In the two-dimensional distribution plane shown, the 0.25 quantile indicates that the value of the “relative distance” is 30, the 0.5 quantile indicates that the value of the “relative distance” is 38, and the 0.75 quantile indicates that the value of the “relative distance” is 47.

[0153] B12: Calculate the sum of the absolute values ​​of the skewness of the two groups of data after division for the multiple parameters corresponding to each quantile, and take the two groups of data corresponding to the minimum value of the sum of the absolute values ​​as the first group of data and the second group of data.

[0154] See also Figure 6 , Figure 6 Schematic diagram of the division of multiple parameter pairs of some deceleration scenarios provided in the embodiments of the present application by the dividing lines determined based on the quantiles. Figure 6 (1) shows that these multiple parameter pairs are divided into two groups of data on the two-dimensional distribution plane by the dividing line 1 determined based on the 0.25 quantile. Figure 6 (2) shows that these multiple parameter pairs are divided into two groups of data on the two-dimensional distribution plane by the dividing line 2 determined based on the 0.5 quantile. Figure 6 (3) shows that these multiple parameter pairs are divided into two groups of data on the two-dimensional distribution plane by the dividing line 3 determined based on the 0.75 quantile. Based on the above skewness calculation formula (such as the above formula (1)), respectively calculate Figure 6 The sum of the absolute values ​​of the skewness of the two sets of data shown in each figure, assuming that the calculation obtains Figure 6 The sum of the absolute values ​​of the skewness of the two sets of data in (1) is Sum 1 , Figure 6 The sum of the absolute values ​​of the skewness of the two sets of data in (2) is Sum 2 as well as Figure 6 The sum of the absolute values ​​of the skewness of the two sets of data in (3) is Sum 3 , where Sum 2 >Sum 1 >Sum 3 , so it is determined Figure 6 The two groups of data obtained by dividing by the dividing line 3 shown in (3) are used as the first group of data and the second group of data mentioned above.

[0155] Here, by Figure 6 It can be seen that the slopes of dividing line 1, dividing line 2, and dividing line 3 are all 0. It can be understood that the embodiment of the present application does not limit the slopes of dividing line 1, dividing line 2, and dividing line 3 to only be 0, that is, the slopes of the dividing lines determined based on the quantiles are not limited, nor are the slopes of the dividing lines determined based on different quantiles limited to be the same. The dividing lines determined based on the quantiles can be used as long as they can divide the multiple parameter pairs into upper and lower groups of data.

[0156] B13: Determine the parameter range of the first motion parameter “relative distance between two vehicles” according to the regression analysis results of the first group of data and the regression analysis results of the second group of data.

[0157] In step B12, Figure 6 The two groups of data obtained after being divided by the dividing line 3 as shown in (3) are the first group of data and the second group of data, wherein the first group of data includes the parameter pairs located above the dividing line 3 among the multiple parameter pairs, and the second group of data includes the parameter pairs located below the dividing line 3 among the multiple parameter pairs. It can be seen that the first group of data is located above the second group of data on the two-dimensional distribution plane. In order to distinguish and display the two groups of data, the parameter pairs in the first group of data are represented by "five-pointed stars" and the parameter pairs in the second group of data are represented by "dots".

[0158] Here, since the first group of data is located above the second group of data, the upper limit of the parameter range of the first motion parameter "relative distance between two vehicles" is determined by the upper limit of the regression analysis results of the first group of data, and the lower limit of the parameter range of the first motion parameter "relative distance between two vehicles" is determined by the lower limit of the regression analysis results of the second group of data. The upper limit corresponding to the first group of data and the lower limit corresponding to the second group of data can be found in Figure 7 .

[0159] For example, in Figure 7 In the example, assume that the upper limit of the first set of data is y 1上 , the lower limit corresponding to the second group of data is represented by y 2下 , then based on Figure 6 The parameter range of the first motion parameter "relative distance between two vehicles" determined by the multiple parameters shown in (3) is: 1上 ≤Relative distance between two vehicles≤y 2下 Here, the calculation of the upper boundary corresponding to the first group of data and the lower boundary corresponding to the second group of data can be similar to the description of the corresponding contents in the above embodiment, and will not be repeated here.

[0160] In an embodiment of the present application, the parameter range of the first motion parameter determined in the above manner can be used to constrain the value of the first motion parameter in a simulation test of a driving scene. In one implementation, in the driving scene, when the first value of the first motion parameter does not belong to the parameter range of the first motion parameter, it is determined that the first value of the first motion parameter is abnormal. In this way, abnormal data of the motion parameter in the driving scene can be eliminated.

[0161] It can be seen that, in the implementation of the embodiment of the present application, a plurality of parameter pairs are formed by combining the values ​​of two motion parameters in a driving scenario. When the plurality of parameter pairs present an asymmetric distribution, the plurality of parameter pairs are divided into a plurality of groups of data on a two-dimensional distribution plane of the plurality of parameter pairs by at least one dividing line, and regression analysis is performed on the two groups of data that are farthest apart from each other in the plurality of groups of data to determine the parameter range of the motion parameter. In this way, the parameter range of the motion parameter obtained not only takes into account the correlation between the motion parameters, but also better fits the distribution of the actual data of the motion parameters in the driving scenario.

[0162] In one implementation, using Figure 3 The method described in the embodiment can determine Figure 2 The vertical axis of each figure shows the parameter range of the motion parameter on the two-dimensional distribution plane. Figure 2 The scene parameters of the vehicle cut-in scene shown include the main vehicle speed, the lateral speed of the cut-in vehicle ( Figure 2 The longitudinal distance of the cutting vehicle relative to the main vehicle ( Figure 2 The relative distance in the figure) and the longitudinal speed of the cutting vehicle relative to the main vehicle ( Figure 2 When the “relative speed” is used as the first motion parameter, the second motion parameter may be “main vehicle speed”, “cut-in vehicle speed” and “relative distance”. Specifically, based on Figure 2 (1) can determine the first parameter range of "relative speed" under the constraint that "relative speed" is the dependent variable and "main vehicle speed" is the independent variable, based on Figure 2 (8) can determine the second parameter range of "relative speed" under the constraint of "relative speed" as the dependent variable and "relative distance" as the independent variable, based on Figure 2 (11) can determine the third parameter range of "relative speed" under the constraint that "relative speed" is the dependent variable and "entry vehicle speed" is the independent variable. Here, the first parameter range, the second parameter range and the third parameter range of the above-mentioned "relative speed" are all used to constrain the value of the motion parameter "relative speed" in the simulation test of the vehicle entry scene.

[0163] In other words, in the simulation test of the vehicle cut-in scenario, the value of the motion parameter "relative speed" needs to satisfy: it is within the first parameter range of "relative speed", within the second parameter range of "relative speed", and within the third parameter range of "relative speed".

[0164] In another implementation, after obtaining the first parameter range, the second parameter range and the third parameter range of the above-mentioned "relative speed", the target parameter range of the "relative speed" can be obtained by taking the intersection of the first parameter range, the second parameter range and the third parameter range of the "relative speed". In this case, in the simulation test of the vehicle cut-in scene, the value of the motion parameter "relative speed" only needs to be within the target parameter range of the "relative speed". It can be understood that Figure 2 The process of determining the target parameter ranges of other motion parameters can also refer to the relevant description of the target parameter range of the above-mentioned "relative speed". For the sake of brevity of the specification, it will not be repeated here.

[0165] See also Figure 8 , Figure 83 is a schematic diagram of a parameter range analysis device provided in an embodiment of the present application. The parameter range analysis device 30 includes an acquisition unit 310 and a processing unit 312. The parameter range analysis device 30 can be implemented by hardware, software, or a combination of hardware and software.

[0166] Among them, the acquisition unit 310 is used to acquire multiple parameter pairs of the driving scene, each parameter pair includes the value of the first motion parameter and the value of the second motion parameter, and the first motion parameter and the second motion parameter are used to describe the state of the subject in the driving scene; the processing unit 312 is used to determine the parameter range of the first motion parameter according to the regression analysis results of the two groups of data that are farthest apart in the multiple groups of data, and the parameter range of the first motion parameter is used to constrain the value of the first motion parameter in the simulation test of the driving scene; wherein, the multiple groups of data are obtained by dividing the multiple parameter pairs on the two-dimensional distribution plane of the multiple parameter pairs by at least one dividing line, each group of data in the above multiple groups of data includes at least one parameter pair in the multiple parameter pairs, the vertical axis of the two-dimensional distribution plane indicates the first motion parameter and the horizontal axis indicates the second motion parameter, the above at least one dividing line is determined based on the distribution of the multiple parameter pairs on the two-dimensional distribution plane, and the above two groups of data that are farthest apart are respectively located on both sides of any dividing line of the above at least one dividing line.

[0167] The parameter range analysis device 30 can be used to implement Figure 3 The method described in the embodiment. Figure 3 In an embodiment, the acquisition unit 310 may be used to execute S301 , and the processing unit 312 may be used to execute S302 .

[0168] It should be understood that the division of the various units in the above parameter range analysis device 30 is only a division of logical functions, and in actual implementation, they can be fully or partially integrated into one physical entity, or they can be physically separated. In addition, the units in the device can be implemented in the form of a processor calling software; for example, the device includes a processor, the processor is connected to a memory, and instructions are stored in the memory. The processor calls the instructions stored in the memory to implement any of the above methods or realize the functions of the various units of the device, wherein the processor is, for example, a general-purpose processor, such as a central processing unit (CPU) or a microprocessor, and the memory is a memory inside the device or a memory outside the device. Alternatively, the units in the device can be implemented in the form of hardware circuits, and the functions of some or all of the units can be realized by designing the hardware circuits, and the hardware circuits can be understood as one or more processors; for example, in one implementation, the hardware circuit is an application-specific integrated circuit (ASIC), and the functions of some or all of the above units are realized by designing the logical relationship of the components in the circuit; for another example, in another implementation, the hardware circuit can be realized by a programmable logic device (PLD), taking a field programmable gate array (FPGA) as an example, which can include a large number of logic gate circuits, and the connection relationship between the logic gate circuits is configured by a configuration file, so as to realize the functions of some or all of the above units. All units of the above devices can be realized in the form of a processor calling software, or in the form of hardware circuits, or in part by a processor calling software, and the rest by hardware circuits.

[0169] In the embodiment of the present application, the processor is a circuit with signal processing capability. In one implementation, the processor can be a circuit with instruction reading and running capability, such as a central processing unit (CPU), a microprocessor, a graphics processing unit (GPU) (which can be understood as a microprocessor), or a digital signal processor (DSP), etc.; in another implementation, the processor can realize certain functions through the logical relationship of the hardware circuit, and the logical relationship of the hardware circuit is fixed or reconfigurable, such as a hardware circuit implemented by a processor as an application-specific integrated circuit (ASIC) or a programmable logic device (PLD), such as an FPGA. In a reconfigurable hardware circuit, the processor loads a configuration document to implement the process of hardware circuit configuration, which can be understood as the process of the processor loading instructions to implement the functions of some or all of the above units. In addition, it can also be a hardware circuit designed for artificial intelligence, which can be understood as an ASIC, such as a neural network processing unit (NPU), a tensor processing unit (TPU), a deep learning processing unit (DPU), etc.

[0170] It can be seen that each unit in the above device can be one or more processors (or processing circuits) configured to implement the above method, such as: CPU, GPU, NPU, TPU, DPU, microprocessor, DSP, ASIC, FPGA, or a combination of at least two of these processor forms.

[0171] In addition, all or part of the units in the above device can be integrated together, or can be implemented independently. In one implementation, these units are integrated together and implemented in the form of a system-on-a-chip (SOC). The SOC may include at least one processor for implementing any of the above methods or implementing the functions of each unit of the device. The type of the at least one processor may be different, for example, including a CPU and an FPGA, a CPU and an artificial intelligence processor, a CPU and a GPU, etc.

[0172] See also Fig. 9 , Fig. 9 Schematic diagram of a computing device provided in an embodiment of the present application. Fig. 9As shown, the computing device 40 includes: a processor 401, a communication interface 402, a memory 403 and a bus 404. The processor 401, the memory 403 and the communication interface 402 communicate with each other via the bus 404. It should be understood that the present application does not limit the number of processors and memories in the computing device 40.

[0173] In one implementation, the computing device 40 may be, for example, a network-side device. The network-side device may be, for example, a server deployed on the network side (e.g., a parameter range analysis server, a simulation test server, etc.), or a component or chip in the server, or a system-level device composed of multiple servers. The network-side device may be deployed in a cloud environment or an edge environment, which is not specifically limited herein.

[0174] The bus 404 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Fig. 9 The bus 404 is represented by only one line, but it does not mean that there is only one bus or one type of bus. The bus 404 may include a path for transmitting information between various components of the computing device 40 (eg, the memory 403, the processor 401, and the communication interface 402).

[0175] The processor 401 can refer to the related description of the processor in the above embodiment, which will not be repeated here.

[0176] The memory 403 is used to provide a storage space, and the storage space can store data such as an operating system and a computer program. The memory 403 can be a random access memory (RAM), an erasable programmable read only memory (EPROM), a read-only memory (ROM), or a portable read-only memory (CD-ROM), etc., or a combination of multiple types. The memory 403 can exist alone or be integrated into the processor 401.

[0177] The communication interface 402 may be used to provide information input or output for the processor 401. Alternatively, the communication interface 402 may be used to receive data sent externally and / or send data externally, and may be a wired link interface such as an Ethernet cable, or a wireless link interface (such as Wi-Fi, Bluetooth, general wireless transmission, etc.). Alternatively, the communication interface 402 may also include a transmitter (such as a radio frequency transmitter, an antenna, etc.) coupled to the interface, or a receiver, etc.

[0178] The processor 401 in the computing device 40 is used to read the computer program stored in the memory 403 to execute the aforementioned method, for example Figure 3 The method described.

[0179] In a possible design, the computing device 40 may be a processor for executing Figure 3 The processor 401 may be used to read one or more computer programs stored in the memory to perform the following operations:

[0180] Acquire multiple parameter pairs of the driving scene through the acquisition unit 310, each parameter pair includes a value of a first motion parameter and a value of a second motion parameter, and the first motion parameter and the second motion parameter are used to describe the state of the subject in the driving scene;

[0181] According to the regression analysis results of the two groups of data that are farthest apart from each other among the multiple groups of data, the parameter range of the first motion parameter is determined, and the parameter range of the first motion parameter is used to constrain the value of the first motion parameter in the simulation test of the driving scene; wherein, the multiple groups of data are obtained by dividing the multiple parameter pairs on the two-dimensional distribution plane of the multiple parameter pairs by at least one dividing line, each group of data in the above multiple groups of data includes at least one parameter pair among the multiple parameter pairs, the vertical axis of the two-dimensional distribution plane indicates the first motion parameter and the horizontal axis indicates the second motion parameter, the above at least one dividing line is determined based on the distribution of the multiple parameter pairs on the two-dimensional distribution plane, and the above two groups of data that are farthest apart from each other are respectively located on both sides of any dividing line among the above at least one dividing line.

[0182] In the above-mentioned embodiments of this invention, the description of each embodiment has its own emphasis. For the part that is not described in detail in a certain embodiment, please refer to the relevant description of other embodiments. In addition, in the various embodiments of this application, if there is no special explanation and logical conflict, the terms and / or descriptions between the various embodiments are consistent and can be referenced to each other. The technical features in different embodiments can be combined to form a new embodiment according to their internal logical relationship.

[0183] It should be noted that, a person skilled in the art can see that all or part of the steps in the various methods of the above embodiments can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electrically-erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0184] The essence of the technical solution of the present application or the part that makes the contribution or all or part of the technical solution can be embodied in the form of a software product. The computer program product is stored in a storage medium and includes a number of instructions for enabling a device (which can be a personal computer, a server, or a network device, a robot, a single-chip microcomputer, a chip, a robot, etc.) to execute all or part of the steps of the method described in each embodiment of the present application.

Claims

1. A method for determining a parameter range of a driving scenario, characterized in that: The method comprises: Acquire multiple parameter pairs of a driving scene, each parameter pair comprising a value of a first motion parameter and a value of a second motion parameter, wherein the first motion parameter and the second motion parameter are used to describe a state of a subject in the driving scene; Determining a parameter range of the first motion parameter according to regression analysis results of two sets of data that are farthest apart from each other among the multiple sets of data; the parameter range of the first motion parameter is used to constrain the value of the first motion parameter in the simulation test of the driving scene; Wherein, the multiple groups of data are obtained by dividing the multiple parameter pairs on the two-dimensional distribution plane of the multiple parameter pairs through at least one dividing line, each group of data in the multiple groups of data includes at least one parameter pair among the multiple parameter pairs, the vertical axis of the two-dimensional distribution plane indicates the first motion parameter, and the horizontal axis of the two-dimensional distribution plane indicates the second motion parameter. The at least one dividing line is determined based on the distribution of the multiple parameter pairs on the two-dimensional distribution plane, and the two groups of data that are farthest apart are respectively located on both sides of any dividing line of the at least one dividing line.

2. The method according to claim 1, characterized in that The plurality of parameter pairs present an asymmetric distribution.

3. The method according to claim 1 or 2, characterized in that: The two groups of data include a first group of data and a second group of data, an upper limit of a parameter range of the first motion parameter is determined by an upper limit in a regression analysis result of the first group of data, and a lower limit of the parameter range of the first motion parameter is determined by a lower limit in a regression analysis result of the second group of data; Wherein, on the two-dimensional distribution plane, the first group of data is located above the second group of data, and the upper boundary is asymmetric to the lower boundary.

4. The method according to any one of claims 1 to 3, characterized in that: An absolute value of a first skewness calculated based on the multiple parameter pairs is greater than zero, and the first skewness is used to indicate a distribution dispersion degree of the multiple parameter pairs; The first skewness is the third-order central moment of the multiple parameter pairs.

5. The method according to claim 4, characterized in that The plurality of parameter pairs are divided into a third group of data and a fourth group of data via a first segmentation line, and a distribution discreteness of the fourth group of data is greater than a distribution discreteness of the third group of data.

6. The method according to claim 5, characterized in that The first segmentation line is obtained by linearly fitting a plurality of statistical mean points based on the plurality of parameters; The first skewness is greater than zero, the third set of data includes parameter pairs among the multiple parameter pairs that are located below the first dividing line, and the fourth set of data includes parameter pairs among the multiple parameter pairs that are located above the first dividing line; or The first skewness is less than zero, the fourth group of data includes parameter pairs among the multiple parameter pairs that are located below the first dividing line, and the third group of data includes parameter pairs among the multiple parameter pairs that are located above the first dividing line.

7. The method according to claim 5 or 6, characterized in that: When the multiple parameters present a long-tailed distribution, The absolute value of the first skewness is greater than a first threshold, and the fourth group of data is divided into a fifth group of data and a sixth group of data via a second dividing line, wherein the distribution dispersion of the sixth group of data is greater than the distribution dispersion of the fifth group of data.

8. The method according to claim 7, characterized in that The second segmentation line is obtained by linear fitting of a plurality of mean value points based on the statistics of the fourth set of data; The first skewness is greater than zero, the fifth set of data includes parameter pairs in the fourth set of data that are below the second dividing line, and the sixth set of data includes parameter pairs in the fourth set of data that are above the second dividing line; or, The first skewness is less than zero, the sixth group of data includes parameter pairs in the fourth group of data that are below the second dividing line, and the fifth group of data includes parameter pairs in the fourth group of data that are above the second dividing line.

9. The method according to claim 5, characterized in that When the plurality of parameter pairs present a mixed distribution, the first segmentation line is a segmentation line determined based on a target quantile, and the target quantile is one of a plurality of quantiles based on statistics of the plurality of parameter pairs; The sum of the absolute value of the skewness of the third group of data and the absolute value of the skewness of the fourth group of data is less than a second threshold, or the target quantile is the quantile at which the sum of the absolute values ​​of the skewness of the two groups of data divided by the multiple parameter pairs corresponding to the multiple quantiles is the smallest.

10. A device for determining a parameter range of a driving scene, characterized in that: The device comprises: an acquisition unit, configured to acquire a plurality of parameter pairs of a driving scene, each parameter pair comprising a value of a first motion parameter and a value of a second motion parameter, wherein the first motion parameter and the second motion parameter are used to describe a state of a subject in the driving scene; A processing unit, configured to determine a parameter range of the first motion parameter according to regression analysis results of two sets of data that are farthest apart from each other among the multiple sets of data; the parameter range of the first motion parameter is used to constrain the value of the first motion parameter in the simulation test of the driving scene; Wherein, the multiple groups of data are obtained by dividing the multiple parameter pairs on the two-dimensional distribution plane of the multiple parameter pairs through at least one dividing line, each group of data in the multiple groups of data includes at least one parameter pair among the multiple parameter pairs, the vertical axis of the two-dimensional distribution plane indicates the first motion parameter, and the horizontal axis of the two-dimensional distribution plane indicates the second motion parameter. The at least one dividing line is determined based on the distribution of the multiple parameter pairs on the two-dimensional distribution plane, and the two groups of data that are farthest apart are respectively located on both sides of any dividing line of the at least one dividing line.

11. The device according to claim 10, characterized in that The plurality of parameter pairs present an asymmetric distribution.

12. The device according to claim 10 or 11, characterized in that The two groups of data include a first group of data and a second group of data, an upper limit of a parameter range of the first motion parameter is determined by an upper limit in a regression analysis result of the first group of data, and a lower limit of the parameter range of the first motion parameter is determined by a lower limit in a regression analysis result of the second group of data; Wherein, on the two-dimensional distribution plane, the first group of data is located above the second group of data, and the upper boundary is asymmetric to the lower boundary.

13. The device according to any one of claims 10 to 12, characterized in that: An absolute value of a first skewness calculated based on the multiple parameter pairs is greater than zero, and the first skewness is used to indicate a distribution dispersion degree of the multiple parameter pairs; The first skewness is the third-order central moment of the multiple parameter pairs.

14. The device according to claim 13, characterized in that The plurality of parameter pairs are divided into a third group of data and a fourth group of data via a first segmentation line, and a distribution discreteness of the fourth group of data is greater than a distribution discreteness of the third group of data.

15. The device according to claim 14, characterized in that The first segmentation line is obtained by linearly fitting a plurality of statistical mean points based on the plurality of parameters; The first skewness is greater than zero, the third set of data includes parameter pairs among the multiple parameter pairs that are located below the first dividing line, and the fourth set of data includes parameter pairs among the multiple parameter pairs that are located above the first dividing line; or The first skewness is less than zero, the fourth group of data includes parameter pairs among the multiple parameter pairs that are located below the first dividing line, and the third group of data includes parameter pairs among the multiple parameter pairs that are located above the first dividing line.

16. The device according to claim 14 or 15, characterized in that When the multiple parameters present a long-tailed distribution, The absolute value of the first skewness is greater than a first threshold, and the fourth group of data is divided into a fifth group of data and a sixth group of data via a second dividing line, wherein the distribution dispersion of the sixth group of data is greater than the distribution dispersion of the fifth group of data.

17. The device according to claim 16, characterized in that The second segmentation line is obtained by linear fitting of a plurality of mean value points based on the statistics of the fourth set of data; The first skewness is greater than zero, the fifth set of data includes parameter pairs in the fourth set of data that are below the second dividing line, and the sixth set of data includes parameter pairs in the fourth set of data that are above the second dividing line; or, The first skewness is less than zero, the sixth group of data includes parameter pairs in the fourth group of data that are below the second dividing line, and the fifth group of data includes parameter pairs in the fourth group of data that are above the second dividing line.

18. The device according to claim 14, characterized in that When the plurality of parameter pairs present a mixed distribution, the first segmentation line is a segmentation line determined based on a target quantile, and the target quantile is one of a plurality of quantiles based on statistics of the plurality of parameter pairs; The sum of the absolute value of the skewness of the third group of data and the absolute value of the skewness of the fourth group of data is less than a second threshold, or the target quantile is the quantile at which the sum of the absolute values ​​of the skewness of the two groups of data divided by the multiple parameter pairs corresponding to the multiple quantiles is the smallest.

19. A computing device, characterized in that: The device comprises at least one processor and an interface circuit, wherein the processor is used to execute instructions and / or data interaction through the interface circuit, so that the device executes the method according to any one of claims 1 to 9.

20. A computer-readable storage medium, characterized in that: The method comprises computer instructions, which, when executed by a processor, implement the method according to any one of claims 1 to 9.

Images