Performance test method and system of circuit board for automatic driving

Abstract

The invention belongs to the technical field of testing. The invention provides a performance test method and system of a circuit board for automatic driving. The method comprises the steps of receiving multiple groups of target type driving conditions sent by a real automobile; respectively extracting data transmission characteristics and data processing characteristics based on the first data transmission information and the first data processing information, and performing reverse simulation generation of data based on the data transmission characteristics and the data processing characteristics to obtain test data; and transmitting second data transmission information to the circuit board for automatic driving according to the time stamp, controlling the circuit board for automatic driving to process second data processing information according to the time stamp, and carrying out fault monitoring on the circuit board for automatic driving based on a preset index. According to the invention, continuous high-low data intensity working condition data of a real automobile is collected, actual data change is simulated to generate test data, and test and monitoring are carried out according to time stamps, so that the data processing capability of the circuit board can be comprehensively detected, and the test accuracy and reliability are improved.

Description

Technical Field

[0001] The present invention relates to the field of testing technologies, and in particular, to a performance testing method and system for a circuit board used in autonomous driving. Background Art

[0002] In an autonomous driving system, a circuit board (composed of a substrate and various electronic components arranged thereon) undertakes the key task of processing a large amount of data collected in real time by multiple sensors such as lidar, cameras, and millimeter-wave radars. Its data processing ability directly determines the accuracy of the vehicle's perception, decision-making, and control of the surrounding environment. Facing a data traffic of up to several GB per second, the circuit board needs to have strong parallel processing, real-time analysis, and fast response capabilities. Once data processing delays, crashes, or freezes occur, the vehicle will not be able to identify obstacles and plan paths in time, seriously threatening driving safety.

[0003] Most of the existing circuit board performance testing methods focus on the detection of conventional indicators such as electrical performance and physical structure, and insufficiently evaluate the performance of the circuit board in high-load data processing scenarios. Traditional testing is often carried out in a static environment with low data traffic and a single working condition, making it difficult to simulate the complex and variable data input situations in actual autonomous driving operations, and unable to effectively monitor whether the circuit board will experience failures such as crashes and freezes under extreme data loads.

[0004] As autonomous driving technology develops towards higher levels, the requirements for the data processing stability and reliability of circuit boards are becoming increasingly stringent. There is an urgent need for a performance testing method that can truly simulate actual data processing scenarios and comprehensively detect the data processing ability and potential faults of circuit boards. Summary of the Invention

[0005] In view of this, the present invention provides a performance testing method, system, electronic device, computer storage medium, and computer program product for a circuit board used in autonomous driving to solve the above technical problems.

[0006] The present invention discloses a performance testing method for a circuit board used in autonomous driving, and the method includes the following steps:

[0007] Receiving multiple sets of target type driving working conditions sent by a real vehicle, where the target type driving working conditions include a low data intensity working condition and a high data intensity working condition with adjacent time; both the low data intensity working condition and the high data intensity working condition include first data transmission information and first data processing information related to the circuit board;

[0008] Based on the first data transmission information and the first data processing information, data transmission characteristics and data processing characteristics are respectively obtained. Based on the data transmission characteristics and the data processing characteristics, reverse simulation generation of data is performed to obtain test data, and the test data includes second data transmission information and second data processing information with time stamps;

[0009] Transmit the second data transmission information to the circuit board for autonomous driving according to the time stamp, and control the circuit board for autonomous driving to process the second data processing information according to the time stamp, and perform fault monitoring on the circuit board for autonomous driving based on preset indicators.

[0010] The present invention also discloses a performance testing system for a circuit board for autonomous driving. The system includes at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor. The processor executes the computer program to implement the following method steps:

[0011] Receive multiple sets of target type driving conditions sent by a real vehicle. The target type driving conditions include a low data intensity condition and a high data intensity condition with adjacent time. Both the low data intensity condition and the high data intensity condition include first data transmission information and first data processing information related to the circuit board;

[0012] Based on the first data transmission information and the first data processing information, data transmission characteristics and data processing characteristics are respectively obtained. Based on the data transmission characteristics and the data processing characteristics, reverse simulation generation of data is performed to obtain test data, and the test data includes second data transmission information and second data processing information with time stamps;

[0013] Transmit the second data transmission information to the circuit board for autonomous driving according to the time stamp, and control the circuit board for autonomous driving to process the second data processing information according to the time stamp, and perform fault monitoring on the circuit board for autonomous driving based on preset indicators.

[0014] The present invention also discloses an electronic device, including: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor. The processor executes the computer program to implement the method as described in any one of the previous items.

[0015] The present invention also discloses a computer storage medium. The computer-readable storage medium stores a computer program, and the computer program is executed by a processor to implement the method as described in any one of the previous items.

[0016] The present invention also discloses a computer program product, which, when running on a terminal, enables the terminal to execute to implement the method described in any of the previous paragraphs.

[0017] In the present invention, by collecting continuous high and low data intensity working condition data of a real vehicle, simulating actual data changes to generate test data, and marking the test and monitoring by time, the data processing ability of the circuit board can be comprehensively detected, the test accuracy and reliability can be improved, and a strong guarantee can be provided for the safe operation of the autonomous driving system. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required to be used in the embodiments. It should be understood that the following drawings only show some embodiments of the present invention, and therefore should not be regarded as limiting the scope. For those of ordinary skill in the art, without creative efforts, other related drawings can also be obtained based on these drawings.

[0019] Figure 1 is a schematic flowchart of a method for testing the performance of a circuit board for autonomous driving disclosed in an embodiment of the present invention;

[0020] Figure 2 is a schematic structural diagram of a system for testing the performance of a circuit board for autonomous driving disclosed in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0021] The following specific embodiments illustrate the implementation manners of the present application. Those skilled in the art can easily understand other advantages and effects of the present application from the content disclosed in this specification. Obviously, the described embodiments are part of the embodiments of the present application, rather than all of them. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope protected by the present application.

[0022] In addition, the technical features involved in different implementation manners of the present application described below can be combined with each other as long as they do not conflict with each other.

[0023] As Figure 1 shown, an embodiment of the present invention discloses a method for testing the performance of a circuit board for autonomous driving, and the method includes the following steps:

[0024] S01, receiving multiple groups of target type driving working conditions sent by a real vehicle, where the target type driving working conditions include a low data intensity working condition and a high data intensity working condition that are adjacent in time; both the low data intensity working condition and the high data intensity working condition include first data transmission information and first data processing information related to the circuit board.

[0025] Traditional circuit board performance testing methods are carried out in a static environment with low data traffic and a single working condition, making it difficult to simulate the complex and variable data input situations in the actual operation of autonomous vehicles and unable to comprehensively evaluate the performance of circuit boards under different data loads. In response to this, the present invention receives multiple sets of target type driving conditions sent by real vehicles, which cover low data intensity conditions and high data intensity conditions that are adjacent in time, for performance testing of circuit boards for autonomous driving that is closer to the actual usage conditions.

[0026] In the low data intensity condition, for example, when the vehicle is in a state of good road conditions and a conventional driving mode, the amount of data transmitted from sensors such as ultrasonic radars and millimeter-wave radars to the circuit board is small, and the data transmission between other in-vehicle functional devices and the circuit board is also at a low level. These data transmission situations constitute the first data transmission information. Recording the data processing amount, data processing rate, etc. generated after the circuit board substantially processes these small amounts of data (excluding data only for conventional storage) constitutes a part of the first data processing information.

[0027] In the high data intensity condition, for example, when the high-level driving assistance mode is turned on or the vehicle is in complex road conditions, multiple sensors work simultaneously, and a large amount of detection data and data of other in-vehicle functional devices are transmitted to the circuit board at high speed, forming a high-intensity first data transmission information; real-time recording of the relevant situations generated when the circuit board receives and processes these massive amounts of data constitutes another part of the first data processing information. It can be understood that since some data does not require substantial processing, the first data processing information is a part of the data in the first data transmission information.

[0028] By collecting the data of these two adjacent-in-time working conditions, it is possible to comprehensively show the data interaction state of the circuit board under different data load levels and different working intensities. Compared with traditional single-condition data, it more truly and completely restores the complex working scenarios of the circuit board in actual operation, providing a rich and practical data basis for subsequent testing.

[0029] In addition, the above low data intensity condition and high data intensity condition are adjacent in time, which is conducive to subsequent simulation of sudden changes in working conditions, and further performance testing of circuit boards for autonomous driving that is closer to the actual usage conditions can be carried out.

[0030] S02, respectively extract and obtain a data transmission feature and a data processing feature based on the first data transmission information and the first data processing information, perform reverse simulation generation of data based on the data transmission feature and the data processing feature, and obtain test data, where the test data contains second data transmission information and second data processing information with time stamps.

[0031] After obtaining the first data transmission information and the first data processing information under low data intensity conditions and high data intensity conditions, the data transmission characteristics and data processing characteristics are respectively obtained based on this information.

[0032] The data transmission characteristics can reflect the rules of data transmission under different conditions, such as the peak value and change frequency of data flow under low data intensity conditions and high data intensity conditions. The data processing characteristics can reflect the performance of the circuit board when processing data of different scales, such as processing efficiency and resource occupancy.

[0033] Next, based on the data transmission characteristics and data processing characteristics, reverse simulation generation of data is performed to obtain test data. Reverse simulation generation is a process of constructing simulated data using the data characteristics extracted from actual data. The generated test data includes the second data transmission information and the second data processing information with time stamps.

[0034] Among them, the time stamp ensures that the time sequence of the test data is consistent with the actual conditions, enabling the test data to accurately restore the changes in the actual data in the time dimension.

[0035] Through this step, the complex data in the actual conditions is converted into simulated data that can be used for standardized testing, solving the problems of single input of traditional test data and inability to simulate complex data changes, and providing effective data support for subsequent comprehensive testing of the circuit board performance.

[0036] S03, transmit the second data transmission information to the circuit board for autonomous driving according to the time stamp, and control the circuit board for autonomous driving to process the second data processing information according to the time stamp, and perform fault monitoring on the circuit board for autonomous driving based on preset indicators.

[0037] After obtaining the test data with time stamps, transmit the second data transmission information (each piece of data in the second data transmission information has a time stamp) to the circuit board for autonomous driving according to the corresponding time stamp, and control the circuit board for autonomous driving to process the second data processing information according to the time stamp.

[0038] The above operations strictly follow the time sequence of actual data transmission and processing, making the working state of the circuit board in the test process highly similar to the real operation scenario, thereby simulating the data interaction and processing process of the circuit board under actual conditions.

[0039] At the same time, perform fault monitoring on the circuit board for autonomous driving based on preset indicators. The preset indicators are set according to the requirements of autonomous driving for the circuit board performance. For example, if the data processing delay exceeds a certain threshold, the system resource occupancy continuously exceeds the upper limit, or there is an unresponsive state, it is determined as a fault.

[0040] By comparing the actual performance of the circuit board when processing test data with the preset indicators, it is possible to timely and accurately detect whether the circuit board has data processing delays, downtime, jamming and other fault problems, accurately locate performance bottlenecks and potential risk points, and achieve comprehensive and effective detection of the circuit board's data processing capabilities and potential fault hazards. It meets the high standards of circuit board performance testing for autonomous driving technology and overcomes the defect that traditional testing is difficult to effectively monitor circuit board failures under extreme data loads.

[0041] The present invention collects continuous high and low data intensity working condition data of real cars, simulates actual data changes to generate test data, and performs tests and monitoring by time stamp. This can comprehensively detect the data processing capabilities of the circuit board, improve test accuracy and reliability, and provide strong protection for the safe operation of the autonomous driving system.

[0042] In some embodiments, the receiving of multiple sets of target type driving conditions sent by a real car includes:

[0043] Determine the autonomous driving level that the autonomous driving circuit board is adapted to, and derive several adapted target operating condition types based on the autonomous driving level, including:

[0044] Determine the operating condition type with the highest operating condition intensity based on the autonomous driving level, and use the operating condition type and other operating condition types with lower operating condition intensity than the operating condition type as the target operating condition type;

[0045] A plurality of groups of target driving conditions that meet the target driving condition types are screened.

[0046] In the embodiment of the present invention, there are multiple levels of autonomous driving technology, from L0 to L5, and the performance requirements of circuit boards at different levels are significantly different.

[0047] In this regard, the present invention first determines the level of autonomous driving that the circuit board for autonomous driving is adapted to, because the circuit board faces different data processing tasks and different degrees of complexity in working conditions at different levels of autonomous driving, and correspondingly, the performance requirements for data processing of the circuit board at different levels of autonomous driving are also different. Among them, the level of autonomous driving that the circuit board for autonomous driving is adapted to can be determined by the designer based on the design and application objectives of the circuit board for autonomous driving, that is, the level of autonomous driving that the circuit board for autonomous driving is subsequently used to perform is pre-specified by the designer.

[0048] Meanwhile, corresponding driving condition types are pre-configured for different levels of autonomous driving, and each driving condition type is characterized by the level of driving condition intensity. For example, in the driving condition scenario corresponding to the L5 level, more sensor fusion data needs to be transmitted and processed, and more autonomous decisions need to be made, so the driving condition intensity is the highest. In contrast, in the driving condition scenario corresponding to the L2 level, much less sensor fusion data needs to be transmitted and processed, and the driving condition intensity is significantly lower than that of the L5 level. In fact, the driving condition intensity gradually increases from L0 to L5. It can be understood that the driving condition intensity can be quantitatively represented by the data processing intensity.

[0049] Based on the above corresponding relationship, the autonomous driving level of the circuit board for autonomous driving to be tested can be determined to obtain the corresponding driving condition types (driving condition type 1, driving condition type 2, …) and their associated driving condition intensities, and this driving condition intensity is used as the highest intensity, and all driving condition types lower than this highest driving condition intensity are selected. For example, if the autonomous driving level of the circuit board for autonomous driving to be tested is L3, then all the driving condition types corresponding to the L0-L3 levels are used as the target driving condition types. It can be understood that although only the L2 and L3 levels have been realized at present, there are also test data for the L4 and L5 levels of autonomous driving under ideal conditions, and they are also included. Of course, all of these data should be legally obtained, and sensitive privacy information should be erased as much as possible, which will not be elaborated here.

[0050] Taking this highest driving condition type and other driving condition types with lower driving condition intensities as the target driving condition types can comprehensively cover different working intensity scenarios that the circuit board may encounter in actual applications. This includes both low-intensity driving conditions that the circuit board can easily handle and high-intensity driving conditions that require full data processing, ensuring the integrity and comprehensiveness of the test data and providing more sufficient data support for accurately evaluating the performance of the circuit board.

[0051] A large amount of driving condition data generated by real vehicle operation will be transmitted and stored in the database, and it will be marked based on the autonomous driving level of the vehicle when the data is generated. Then, after determining the target driving condition types, multiple groups of target type driving conditions that match the target driving condition types are accurately selected from the database.

[0052] In some embodiments, based on the first data transmission information and the first data processing information, data transmission characteristics and data processing characteristics are respectively obtained, and based on the data transmission characteristics and the data processing characteristics, reverse simulation generation of data is performed to obtain test data, including:

[0053] Using a convolutional network to respectively extract features from the first data transmission information and the first data processing information to obtain the data transmission characteristics and the data processing characteristics;

[0054] Quantify the data traffic change feature, transmission frequency distribution feature, and transmission path priority feature in the data transmission features, and combine the processing efficiency fluctuation range feature, algorithm response time feature, and resource occupancy ratio feature in the data processing features to construct a data simulation generation model;

[0055] Through the data simulation generation model, according to the preset random perturbation rules, generate multiple groups of simulation data within the value ranges of the data transmission features and the data processing features;

[0056] According to the time sequence and duration of the low data intensity working condition and the high data intensity working condition, add corresponding time marks to the multiple groups of simulation data to form test data containing the second data transmission information and the second data processing information with time marks.

[0057] In the embodiments of the present invention, in view of the fact that the convolutional network (CNN) has strong feature extraction capabilities and can automatically mine key features from complex data. The present invention is configured to input the first data transmission information and the first data processing information into the convolutional network model respectively, and extract the data transmission features and the data processing features through structures such as convolutional layers, pooling layers, and fully connected layers.

[0058] Among them, the first data transmission information, for example, includes data records transmitted by a millimeter-wave radar to a circuit board within a period of time, including fields such as timestamps, data volumes, and transmission intervals; the first data processing information includes the operation logs when the circuit board processes these data, such as the processing start time, end time, CPU occupancy rate, etc. Organize these two types of data into a matrix form as the input of the convolutional network.

[0059] The calculation formula for the convolution operation is:

[0060]

[0061] Among them, is the value of the output feature map of the th convolutional layer at the position, is the convolutional kernel weight, is the value of the corresponding position of the input feature map of the th layer, is the bias. Through multiple layers of convolution and pooling operations, gradually extract the deep features of the data.

[0062] The convolutional network can be built using TensorFlow or PyTorch, and no specific limitation is required.

[0063] Then, among the extracted data transmission characteristics, there are data traffic change characteristics, transmission frequency distribution characteristics, and transmission path priority characteristics. Among the data processing characteristics, there are processing efficiency fluctuation range characteristics, algorithm response time characteristics, and resource occupancy ratio characteristics. Quantify these characteristics (for example, normalization processing), convert them into numerical data, and then integrate them into a complete feature matrix. Based on the quantified and integrated feature matrix, construct a data simulation generation model that can generate simulated data. This model can be constructed based on a statistical model, a machine learning model, or a deep learning model.

[0064] Taking the Gaussian mixture model (GMM) as an example, introduce the structure and the model construction process of the data simulation generation model:

[0065] The Gaussian mixture model is essentially a probability model composed of multiple Gaussian distributions, and its structure can be understood as the weighted superposition of multiple single Gaussian distributions. When used for data simulation generation of an autonomous driving circuit board, each single Gaussian distribution corresponds to a data distribution pattern in the data feature space. For example, for the data traffic change rate in the data transmission characteristics, there are stable traffic change patterns during normal driving, fluctuating traffic change patterns under complex road conditions, etc. Each pattern can be represented by a single Gaussian distribution.

[0066] The probability density function of the Gaussian mixture model is:

[0067]

[0068] Among them, represents the number of Gaussian distributions, that is, the number of "mixture components" in the model; is the weight of the th Gaussian distribution, satisfying , and , which reflects the proportion of the th distribution in the entire mixture model; is a Gaussian distribution with a mean of and a covariance matrix of . determines the central position of this Gaussian distribution, determines the shape and dispersion degree of the distribution. When processing circuit board data, and will be adjusted according to the distribution characteristics of the data transmission characteristics (data traffic change characteristics, transmission frequency distribution characteristics, transmission path priority characteristics) and data processing characteristics (processing efficiency fluctuation range characteristics, algorithm response time characteristics, resource occupancy ratio characteristics).

[0069] The construction process of the Gaussian mixture model (GMM):

[0070] Collect the data features extracted and quantified from the first data transmission information and the first data processing information. For example, combine features such as data traffic change rate, transmission frequency probability, and average processing efficiency into a feature matrix. Suppose N groups of data features are obtained, and each group of features contains D dimensions, then the dimension of the feature matrix is N×D.

[0071] Initialize parameters: The number of Gaussian distributions K and the weight of each Gaussian distribution need to be initialized. , mean and covariance matrix . The random initialization method can be adopted. For example, randomly generate weights that satisfy and , and randomly set the initial values of the mean and covariance matrix.

[0072] Parameter estimation: Use the Expectation-Maximization (EM) algorithm to iteratively optimize the model parameters. In the E-step (expectation step), calculate the probability that each data point belongs to each Gaussian distribution according to the current model parameters; in the M-step (maximization step), re-estimate the model parameters ( , and ) to maximize the likelihood function of the data. By continuously repeating the E-step and M-step until the model parameters converge, that is, the parameter change is less than the preset threshold.

[0073] Model training: Input the prepared feature matrix into the Gaussian mixture model and train it according to the EM algorithm. During the training process, the model will gradually learn the distribution law of the data features, adjust the parameters of each Gaussian distribution, so that the model can better fit the data. For example, through training, the model will determine the central position (mean ) and the degree of dispersion (covariance matrix ) corresponding to different data traffic change patterns, and the probability (weight ) that each pattern appears in the actual data.

[0074] Model verification and adjustment: After training, use a part of the data that has not participated in training (verification set) to verify the model. Evaluate the fitting effect of the model by calculating indicators such as the likelihood value and mean square error of the verification set data under the model. If the model effect is not ideal, the number of Gaussian distributions K can be adjusted, the parameters can be re-initialized, or other optimization algorithms can be adopted, and then retrained and verified until a satisfactory model is obtained.

[0075] After the above construction process, the Gaussian mixture model can effectively learn the distribution law of the data features of the autonomous driving circuit board, and provide a reliable data simulation generation model for generating simulated data according to the preset random perturbation rules in the future.

[0076] In some embodiments, the random perturbation rule is specifically as follows:

[0077] Obtain the driving environment information corresponding to the target type of driving condition. The driving environment information includes the number of other vehicles with vehicle-to-vehicle communication functions and / or roadside devices with vehicle-to-road communication functions in the preset area where the real vehicle is located. Determine the minimum value of the random perturbation signal based on the number.

[0078] Wherein, the minimum value is positively correlated with the number.

[0079] In the embodiments of the present invention, the data transmission characteristics and data processing characteristics on which the foregoing simulated data is generated are actually based on the data actually transmitted (received and transmitted externally) and processed by the real vehicle. These data do not consider some potential situations, such as whether there are other vehicles or roadside devices (intelligent traffic lights, roadside base stations) with communication functions in the surrounding area. Specifically, when the real vehicle enters a high data intensity driving condition, it is generally because the current road condition has become complex. At this time, the vehicle controller triggers the activation of a higher road condition detection mode and a more refined data processing strategy. Similarly, other vehicles or roadside devices around the real vehicle may also enter a high data intensity driving condition when facing the same road condition.

[0080] In view of the above situation, the present invention sets that the real vehicle transmits the target type of driving condition together with the driving environment information in the preset area (such as within 20 meters or 50 meters around) at that time to the database. The driving environment information includes the number of other vehicles with vehicle-to-vehicle communication functions or roadside devices with vehicle-to-road communication functions in the preset area where the real vehicle is located. Among them, in the presence of complex road conditions, other vehicles with driving assistance systems have a higher probability of needing to communicate with surrounding vehicles to obtain information such as the position, speed, and expected driving path of the real vehicle, so as to guide their own driving decisions; the roadside devices will transmit information such as road conditions and traffic rules to the real vehicle and also obtain information such as the position, speed, and expected driving path of the real vehicle.

[0081] Therefore, after obtaining the number of other vehicles with in-vehicle communication functions and the number of roadside devices with vehicle-road communication functions within the preset area, the present invention determines the minimum value of the intensity of the random perturbation signal based on these quantities. And the minimum value is set to be positively correlated with this quantity because the more devices there are in the preset area, the higher the uncertainties such as data transmission fluctuations (mainly increases) and processing task changes faced by the actual operation of the autonomous driving circuit board. For example, when there are 20 vehicles with in-vehicle communication functions and 10 roadside devices in the preset area, compared with the situation where there are only 5 vehicles and 3 roadside devices, the probability that the circuit board receives a larger amount of data and the data changes more frequently is higher. At this time, in order to more realistically simulate the working state of the circuit board in a complex environment, a higher minimum value of the random perturbation signal needs to be set. In actual operation, the minimum value can be determined by establishing a mathematical model or looking up a preset mapping table, so as to ensure that the generated simulation data can better fit the operation of the circuit board in the actual complex environment and effectively test the ability of the circuit board to cope with data transmission and processing uncertainties.

[0082] It can be understood that the random perturbation signal can be at least one of the following interference signals:

[0083] Data flow fluctuation signal: Simulates the sudden increase or decrease in the amount of data received by the circuit board due to changes in the number of in-vehicle communication and vehicle-road communication devices in the actual scenario. For example, during the traffic peak period, a large number of vehicles and roadside devices transmit data simultaneously, causing the amount of data received by the circuit board to far exceed the normal level instantaneously, and the data flow fluctuation signal reflects this sudden change in the amount of data.

[0084] Data transmission frequency change signal: Reflects the irregular change in the data transmission interval. Under complex road conditions, the communication frequency between vehicles and surrounding devices will change due to environmental changes. It may originally transmit data once per second and suddenly change to transmit once every half second, or the transmission interval becomes irregular. This signal can simulate the uncertainty of this transmission frequency.

[0085] Data processing priority change signal: Simulates the dynamic adjustment of the data processing priorities of various types of data by the circuit board under different working conditions. When an emergency obstacle is detected ahead, the processing priority of the sensor data related to the obstacle will be increased, while the processing priority of some non-emergency entertainment system data will be decreased. The data processing priority change signal is used to simulate such changes in priorities.

[0086] Data transmission path switching signal: Reflects the random change in the data transmission path during the transmission process. Due to factors such as the network environment and device status, the data may originally be transmitted through a certain path and suddenly switch to another path. This signal can simulate the impact of the uncertainty brought by the data transmission path switching on the performance of the circuit board.

[0087] In some embodiments, the fault monitoring of the circuit board for autonomous driving based on preset metrics includes:

[0088] After the circuit board for autonomous driving is tested based on the second data transmission information and the second data processing information, if no fault is detected, a new random perturbation rule is formulated based on the increased random perturbation signal, and multiple groups of new test data are generated based on the new random perturbation rule;

[0089] The subsequent test of the circuit board for autonomous driving is carried out using the new test data until a fault is detected.

[0090] In the embodiments of the present invention, another purpose of testing the circuit board for autonomous driving may also be to test the upper limit of data transmission and processing of the circuit board for autonomous driving. Therefore, after the test is completed using the test data, if no fault is detected, the random perturbation signal is increased, and new test data representing a higher working condition intensity is formulated, and so on until a fault is detected. At this time, the performance upper limit of the circuit board for autonomous driving is determined. Of course, the test can be divided into multiple rounds, and the test results are integrated, which can improve the accuracy of determining the performance upper limit.

[0091] Such as Figure 2 As shown, the embodiments of the present invention also disclose a performance test system for a circuit board for autonomous driving. The system includes at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor. The processor executes the computer program to implement the following method steps:

[0092] Receiving multiple groups of target type driving working conditions sent by a real vehicle, where the target type driving working conditions include a low data intensity working condition and a high data intensity working condition with adjacent time; both the low data intensity working condition and the high data intensity working condition include first data transmission information and first data processing information related to the circuit board;

[0093] Based on the first data transmission information and the first data processing information, data transmission characteristics and data processing characteristics are respectively obtained, and reverse simulation generation of data is performed based on the data transmission characteristics and the data processing characteristics to obtain test data, where the test data includes second data transmission information and second data processing information with time stamps;

[0094] Transmitting the second data transmission information to the circuit board for autonomous driving according to the time stamp, and controlling the circuit board for autonomous driving to process the second data processing information according to the time stamp, and performing fault monitoring on the circuit board for autonomous driving based on preset metrics.

[0095] An embodiment of the present invention also discloses an electronic device, including: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, where the processor executes the computer program to implement the method as described in the foregoing embodiment.

[0096] An embodiment of the present invention also discloses a computer storage medium storing a computer program, where the computer program is executed by a processor to implement the method as described in the foregoing embodiment.

[0097] An embodiment of the present invention also discloses a computer program product, when the computer program product runs on a terminal, enabling the terminal to execute to implement the method as described in the foregoing embodiment.

[0098] The present invention is described with reference to the flowcharts and / or block diagrams of methods, systems, electronic devices, computer storage media, and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, and the combination of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for implementing the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0099] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device, and the instruction device implements the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0100] As mentioned above, the above is only a preferred embodiment of the present invention and is not used to limit the protection scope of the present invention.

Claims

1. A performance testing method for a circuit board used in autonomous driving, characterized in that , including the following steps: Receiving multiple sets of target type driving conditions sent by a real vehicle, where the target type driving conditions include a low data intensity condition and a high data intensity condition with adjacent time; both the low data intensity condition and the high data intensity condition include first data transmission information and first data processing information related to a circuit board; Based on the first data transmission information and the first data processing information, respectively extracting to obtain a data transmission feature and a data processing feature, and based on the data transmission feature and the data processing feature, performing reverse simulation generation of data to obtain test data, where the test data includes second data transmission information and second data processing information with time stamps; Transmitting the second data transmission information to the circuit board for autonomous driving according to the time stamp, and controlling the circuit board for autonomous driving to process the second data processing information according to the time stamp, and performing fault monitoring on the circuit board for autonomous driving based on a preset index.

2. The performance testing method of a circuit board for autonomous driving according to claim 1, characterized in that: The receiving of multiple sets of target type driving conditions sent by a real vehicle includes: Determining the autonomous driving level adapted to the circuit board for autonomous driving, and based on the autonomous driving level, obtaining several adapted target condition types, including: Based on the autonomous driving level, obtaining the condition type with the highest condition intensity, and taking this condition type and other condition types with condition intensity lower than this condition type as the target condition types; Screening multiple sets of the target type driving conditions that meet each of the target condition types.

3. A performance testing method for a circuit board used in autonomous driving according to claim 1, characterized in that: Based on the first data transmission information and the first data processing information, respectively extracting to obtain a data transmission feature and a data processing feature, and based on the data transmission feature and the data processing feature, performing reverse simulation generation of data to obtain test data, including: Using a convolutional network to respectively extract features from the first data transmission information and the first data processing information to obtain the data transmission feature and the data processing feature; Quantifying the data flow change feature, transmission frequency distribution feature, and transmission path priority feature in the data transmission feature, and combining the processing efficiency fluctuation range feature, algorithm response time feature, and resource occupancy ratio feature in the data processing feature to construct a data simulation generation model; Through the data simulation generation model, according to a preset random perturbation rule, generating multiple sets of simulation data within the value ranges of the data transmission feature and the data processing feature; According to the time sequence and duration of the low data intensity condition and the high data intensity condition, adding corresponding time stamps to the multiple sets of simulation data to form test data including second data transmission information and second data processing information with time stamps.

4. A performance testing method for a circuit board used in autonomous driving according to claim 3, characterized in that: The specific random perturbation rule is: Obtaining the driving environment information corresponding to the target type driving conditions, where the driving environment information includes the number of other vehicles with vehicle-to-vehicle communication functions and / or roadside devices with vehicle-to-road communication functions in a preset area where the real vehicle is located, and determining the minimum value of the random perturbation signal based on the number; Among them, the minimum value is positively correlated with the number.

5. A performance testing method for a circuit board for autonomous driving according to claim 4, characterized in that: Fault monitoring of the circuit board for autonomous driving based on preset metrics includes: After testing the circuit board for autonomous driving based on the second data transmission information and the second data processing information, if no fault is detected, a new random perturbation rule is formulated based on the increased random perturbation signal, and multiple groups of new test data are generated based on the new random perturbation rule; Subsequent tests of the circuit board for autonomous driving are performed using the new test data until a fault is detected.

6. A performance testing system for a circuit board used in autonomous driving, the system comprising at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, characterized in that: The processor executes the computer program to implement the following method steps: Receiving multiple groups of target type driving conditions sent by a real vehicle, where the target type driving conditions include a low data intensity condition and a high data intensity condition with adjacent time; both the low data intensity condition and the high data intensity condition include first data transmission information and first data processing information related to the circuit board; Based on the first data transmission information and the first data processing information, data transmission characteristics and data processing characteristics are respectively obtained, and reverse simulation generation of data is performed based on the data transmission characteristics and the data processing characteristics to obtain test data, where the test data includes second data transmission information and second data processing information with time stamps; Transmitting the second data transmission information to the circuit board for autonomous driving according to the time stamp, and controlling the circuit board for autonomous driving to process the second data processing information according to the time stamp, and performing fault monitoring on the circuit board for autonomous driving based on preset metrics.

7. The performance test system for a circuit board used in autonomous driving according to claim 6, characterized in that: The receiving multiple groups of target type driving conditions sent by a real vehicle includes: Determining the autonomous driving level adapted to the circuit board for autonomous driving, and obtaining several adapted target condition types based on the autonomous driving level, including: Obtaining the condition type with the highest condition intensity based on the autonomous driving level, and using this condition type and other condition types with condition intensity lower than this condition type as the target condition types; Screening multiple groups of the target type driving conditions that meet each of the target condition types.

8. An electronic device, comprising: At least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, characterized in that: the processor executes the computer program to implement the method according to any one of claims 1-5.

9. A computer storage medium storing a computer program, characterized in that: The computer program is executed by the processor to implement the method according to any one of claims 1-5.

10. A computer program product, comprising a computer program stored on a non-transitory computer-readable medium, characterized in that: When the computer program is executed by the processor, it implements the method according to any one of claims 1-5.

Images