Evaluation method and device for automatic driving compliance

Abstract

The invention discloses an evaluation method and device for automatic driving compliance. A specific implementation mode of the method comprises the following steps: acquiring current driving data generated in a driving process of an automatic driving target vehicle; analyzing the preset compliance engine system configuration file to obtain an analysis result; and based on the analysis result, calling a mathematical function library to carry out compliance evaluation on the current driving data, and generating a compliance evaluation result. According to the embodiment, a legal language can be converted into a logic description and a mathematical equivalence relation; and a software code of an embedded language is generated according to the logic and mathematical relationship, then the current driving data of the target vehicle is evaluated based on the software code, and a compliance evaluation result is automatically generated. Therefore, the compliance of automatic driving can be automatically evaluated based on the current driving data of the target vehicle; the automatic driving requirement of the target vehicle is met, the method is suitable for different scenes, and the universality of scheme application is improved.

Description

Technical Field

[0001] This invention belongs to the field of autonomous driving technology, and particularly relates to an assessment method and device for autonomous driving compliance. Background Technology

[0002] With the development of autonomous vehicles, they will share transportation infrastructure and should adhere to the same traffic rules as human-driven vehicles. However, current traffic rules are compiled using natural language designed for human drivers, and much of this language is ambiguous, making it difficult for autonomous vehicles to accurately understand them. Therefore, digitizing current traffic rules to enable autonomous vehicles to understand and comply with them is crucial. Digitizing traffic rules also facilitates real-time compliance monitoring of autonomous vehicle behavior.

[0003] Currently, to monitor the compliance of autonomous vehicle behavior, a simulation testing method and system for traffic regulation compliance of autonomous vehicles has been proposed. This system designs atomic propositions according to classification principles and uses logic language to write and combine them into formal traffic regulations. By acquiring information about traffic participants, corresponding formal traffic regulations are defined, and the truth value of the atomic propositions in the traffic regulations is determined based on the acquired information, thus representing whether traffic participants are complying with the traffic regulations. However, most atomic propositions remain ambiguous for autonomous vehicles, and the system has excessively high requirements for information acquisition; much information is difficult to obtain or can only be obtained indirectly. Existing solutions are only suitable for offline compliance judgment of collected vehicle data and are too dependent on information acquisition, making them unsuitable for compliance monitoring of autonomous vehicles and failing to meet the needs of autonomous driving.

[0004] Another existing technology involves monitoring the compliance of autonomous driving vehicles. This method includes acquiring scenario information of the target vehicle during autonomous driving, retrieving trigger information corresponding to the compliance monitoring function from the scenario information, and determining whether the trigger information meets preset trigger conditions. If so, the method uses the compliance monitoring function to retrieve corresponding monitoring information from the scenario information to monitor the compliance of the target vehicle's autonomous driving. This monitoring method first analyzes the scenario information and then analyzes the relevant regulations that may appear and can be understood from the scenario information. Therefore, this method can only monitor specific application scenarios, has limitations, and lacks universality. Summary of the Invention

[0005] To address the aforementioned problems in the existing technology, embodiments of the present invention provide a method and apparatus for evaluating the compliance of autonomous driving. This method can automatically evaluate the compliance of autonomous driving based on the current driving data of the target vehicle, thus meeting the requirements of autonomous driving of the target vehicle.

[0006] According to a first aspect of the present invention, a method for evaluating the compliance of autonomous driving is provided. The method includes: acquiring current driving data generated by a target vehicle during driving; parsing a preset compliance engine system configuration file to obtain a parsing result; and, based on the parsing result, calling a mathematical function library to perform a compliance evaluation on the current driving data to generate a compliance evaluation result. The current driving data includes at least: environmental perception data, vehicle status data, and path planning.

[0007] Optionally, the method further includes: converting traffic regulations documents corresponding to the target vehicle model in the target country into machine-readable logical and formal language to generate a first condition library and a first rule library; establishing a mapping relationship between each legal field in the first condition library and the vehicle kinematics model to construct a condition analysis modeling system; associating the condition analysis modeling system with a mathematical model to output a system parameter table and a condition mathematical model library; determining the first condition library, the first rule library, and the system parameter table as a preset compliance engine system configuration file; and converting the condition mathematical model library into a code generation system to generate a mathematical function library.

[0008] Optionally, the step of converting the traffic regulations document corresponding to the target vehicle model in the target country into a machine-readable logical and formal language to generate a first condition library and a first rule library includes: decomposing the traffic regulations document corresponding to the target vehicle model in the target country into several traffic articles based on a regulation decomposer; for any target traffic article among the several traffic articles: performing content filling processing on the target traffic article using a semantic analysis model to output a standard traffic article; decomposing the standard traffic article into several first conditions based on the article decomposer; generating a first condition library corresponding to the traffic regulations document based on the several first conditions corresponding to each standard traffic article; and logically combining the first conditions in the first condition library based on a logical expression generator to generate a corresponding first rule library.

[0009] Optionally, the preset compliance engine system configuration file includes at least: a first condition library, a first rule library, and a system parameter table; the step of parsing the preset compliance engine system configuration file to obtain a parsing result includes: for any first rule in the first rule library: performing rule parsing processing on the first rule library and storing the rule parsing result according to the corresponding parsing value; the rule parsing result includes: at least two condition index numbers and at least one logical operator; selecting a first condition corresponding to each condition index number from the first condition library to generate a second condition library; for any first condition in the second condition library: performing condition parsing processing on the first condition and storing the condition parsing result according to the corresponding parsing value; the condition parsing result includes: at least one object and at least one function name; parsing the system parameter table and outputting the parsed system parameter table; and determining the parsed system parameter table, the rule parsing result, and the condition parsing result as the parsing result.

[0010] Optionally, the step of performing rule parsing processing on the first rule base and storing the rule parsing results according to the corresponding parsing values ​​includes: obtaining the rule index number of the first rule; based on the rule index number, reading and processing the fields in the first rule to obtain the current field; detecting the current field to obtain a first detection result; if the first detection result indicates that the current field is a logical operator, then converting the logical operator into a corresponding parsing value and storing it; if the first detection result indicates that the current field is not a logical operator, then storing the condition index number when determining that the current field is a condition index number.

[0011] Optionally, the step of converting the logical operator into a corresponding parsed value and storing it if the first detection result indicates that the current field is a logical operator includes: if the first detection result indicates that the current field is a NOT operator, then searching for the ending field of the NOT operator; if the search result indicates that the ending field of the NOT operator is parentheses, then converting the NOT operator and the adjacent parentheses into corresponding parsed values ​​and storing them; if the first detection result indicates that the current field is not a NOT operator, then determining whether the current field is an AND or OR field; if the determination result indicates that the current field is an AND or OR field, then converting the AND or OR field into a corresponding parsed value and storing it.

[0012] Optionally, the step of performing condition parsing processing on the first condition and storing the condition parsing result according to the corresponding parsing value includes: obtaining the condition index number of the first condition; based on the condition index number, reading the fields in the first condition to obtain the current field; detecting the current field to obtain a second detection result; if the second detection result indicates that the current field is a valid mathematical operation, then storing the parsing value corresponding to the mathematical operation; wherein, the mathematical operation includes at least: mathematical operator name, and / or function name, and / or constant; if the second detection result indicates that the current field is not a valid mathematical operation, then storing the parsing value corresponding to the object when the current field is determined to be an object.

[0013] Optionally, the step of calling a mathematical function library to perform a compliance assessment on the current driving data based on the parsing result and generating a compliance assessment result includes: selecting a target object and a target function name corresponding to the current driving data from the parsing result; calling a mathematical function corresponding to the target function name from the mathematical function library; evaluating the target object using the mathematical function based on the parsed system parameter list to generate a real-time conditional assessment result; and performing a compliance assessment on the conditional assessment result based on the rule parsing result to generate a real-time rule assessment result.

[0014] Optionally, the method further includes: generating corresponding real-time condition evaluation lists and real-time rule evaluation lists based on the real-time condition evaluation results and real-time rule evaluation results corresponding to each current driving data; performing traffic regulation trigger analysis on the current driving data based on the real-time condition evaluation lists to generate a regulation trigger report; performing violation analysis on the real-time rule evaluation lists to output a compliance statistics report; and evaluating the current driving data based on the real-time rule evaluation lists to output an event record report.

[0015] According to a second aspect of the present invention, an assessment device for autonomous driving compliance is also provided. The device includes: a data acquisition module for acquiring current driving data generated by a target vehicle during autonomous driving; a parsing module for parsing a preset compliance engine system configuration file to obtain a parsing result; and an assessment module for using a mathematical function library to perform a compliance assessment on the current driving data based on the parsing result, thereby generating a compliance assessment result.

[0016] According to a third aspect of the present invention, an electronic device is also provided, comprising: a processor; a memory for storing executable instructions of the processor; the processor being configured to read the executable instructions from the memory and execute the instructions to implement the method as described in the first aspect.

[0017] According to a fourth aspect of the present invention, a computer-readable medium is also provided, on which a computer program is stored, wherein the program, when executed by a processor, implements the method described in the first aspect.

[0018] This invention provides a method and apparatus for assessing the compliance of autonomous driving. The method includes: first, acquiring current driving data generated by the target vehicle during its operation; second, parsing a preset compliance engine system configuration file to obtain a parsing result; and then, based on the parsing result, calling a mathematical function library to perform a compliance assessment on the current driving data, generating a compliance assessment result. This embodiment can transform legal language into logical descriptions and mathematical equivalence relations; and generate embedded language software code based on these logical and mathematical relations. Then, based on this software code, the current driving data of the target vehicle is assessed, automatically generating a compliance assessment result. Therefore, it can automatically assess the compliance of autonomous driving based on the current driving data of the target vehicle; this not only meets the needs of autonomous driving of the target vehicle but is also applicable to different scenarios, improving the universality of the solution application. Attached Figure Description

[0019] The following sections will describe some specific embodiments of the invention in detail by way of example and not limitation, with reference to the accompanying drawings. The same reference numerals in the drawings denote the same or similar parts or portions. Those skilled in the art should understand that these drawings are not necessarily drawn to scale. In the drawings:

[0020] Figure 1 This is a flowchart illustrating an embodiment of the method for assessing autonomous driving compliance provided by the present invention.

[0021] Figure 2 This is a schematic diagram of the architecture of a compliance engine system provided in an embodiment of the present invention;

[0022] Figure 3 This is a schematic diagram of the process of performing rule parsing on the first rule in one embodiment of the present invention;

[0023] Figure 4 This is a schematic diagram of the process of performing condition parsing on the first condition in one embodiment of the present invention;

[0024] Figure 5 This is a schematic diagram of the structure of outputting standard traffic law provisions based on content filling processing in one embodiment of the present invention;

[0025] Figure 6 This is a schematic diagram illustrating the results of breaking down and classifying the fields of standard traffic laws.

[0026] Figure 7 This is a schematic diagram of the physical model corresponding to function expression 2 in one embodiment of the present invention;

[0027] Figure 8 This is a schematic diagram of the system parameter table in one embodiment of the present invention;

[0028] Figure 9 This is a schematic diagram of the structure of the first rule base in one embodiment of the present invention;

[0029] Figure 10 This is a schematic diagram of the structure of the second condition library in one embodiment of the present invention;

[0030] Figure 11 A schematic diagram of the data flow in an assessment method for autonomous driving compliance provided in another embodiment of the present invention;

[0031] Figure 12 This is a schematic diagram of the structure of an evaluation device for autonomous driving compliance provided in an embodiment of the present invention. Detailed Implementation

[0032] To make the objectives, features, and advantages of this invention more apparent and understandable, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0033] like Figure 1 The diagram shown is a flowchart illustrating an assessment method for autonomous driving compliance provided in an embodiment of the present invention.

[0034] An assessment method for autonomous driving compliance includes at least the following steps:

[0035] S101, acquire the current driving data generated by the target vehicle during the driving process; wherein, the current driving data includes at least: environmental perception data, vehicle status data and path planning;

[0036] S102, parse the preset compliance engine system configuration file to obtain the parsing result;

[0037] S103, based on the parsing results, calls the mathematical function library to perform a compliance assessment on the current driving data and generates a compliance assessment result.

[0038] Specifically, the preset compliance engine system configuration file includes at least: a first condition library, a first rule library, and a system parameter table; wherein, the first condition library includes several first conditions, which are used to indicate logical judgments of specific scenarios derived from the processing, analysis, and calculation of objects in traffic laws; the first rules are used to indicate logical judgments based on the combination of at least two first conditions using logical operators, such as... Figure 9 As shown; the system parameter table is used to indicate the settings information of the compliant engine system itself, as well as constant information related to the autonomous vehicle itself, such as... Figure 8 As shown; the mathematical function library is used to indicate the set of basic computations constituted by mathematical or geometric descriptions invoked based on each first condition in the first condition library.

[0039] like Figure 8 As shown, the system parameter table stores the system's own settings, such as sampling frequency and maximum number of objects, as well as constants related to the vehicle itself, such as vehicle length, width, type, maximum braking force, and expected braking force when other vehicles brake suddenly. These parameters are assigned to corresponding system constants during system initialization, converted into a unified SI unit in the data processing unit (e.g., the maximum speed limit on highways: 120 km / h becomes 33.3 m / s), and stored in memory.

[0040] Here, compliance assessment results can be a real-time rule assessment list, and / or a real-time condition assessment list, and / or a regulatory trigger report, and / or a compliance statistics report, and / or an event log report.

[0041] For example, when the evaluation result is a real-time rule evaluation list and a real-time condition evaluation list;

[0042] The step of performing a compliance assessment on the current driving data based on the parsing results and generating a compliance assessment result includes: selecting a target object and target function name corresponding to the current driving data from the parsing results; calling the mathematical function corresponding to the target function name from the mathematical function library; evaluating the target object using the mathematical function based on the parsed system parameter list to generate a real-time conditional assessment result; and performing a compliance assessment on the conditional assessment result based on the rule parsing results to generate a real-time rule assessment result. Based on the real-time conditional assessment result and real-time rule assessment result corresponding to each piece of current driving data, a corresponding real-time conditional assessment list and a real-time rule assessment list are generated, respectively.

[0043] For example, when the evaluation results are a real-time rule evaluation list, a real-time condition evaluation list, a regulatory trigger report, a compliance statistics report, and an event log report;

[0044] The process of using a mathematical function library to perform a compliance assessment on the current driving data based on the parsing results, and generating a compliance assessment result, includes: selecting a target object and target function name corresponding to the current driving data from the parsing results; calling a mathematical function corresponding to the target function name from the mathematical function library; evaluating the target object using the mathematical function based on the parsed system parameter list, and generating a real-time condition assessment result; and performing a compliance assessment on the condition assessment result based on the rule parsing results, generating a real-time rule assessment result. Based on the real-time condition assessment result and real-time rule assessment result corresponding to each piece of current driving data, corresponding real-time condition assessment lists and real-time rule assessment lists are generated respectively; traffic regulation trigger analysis is performed on the current driving data based on the real-time condition assessment lists, generating a regulation trigger report; violation analysis is performed on the real-time rule assessment lists, outputting a compliance statistics report; and an event record report is output based on the real-time rule assessment lists for the current driving data assessment.

[0045] This embodiment first loads a preset compliance engine system configuration file into the target vehicle or related hardware, then receives the current driving data output by the corresponding environmental perception module of the autonomous driving software stack, and analyzes and processes the current driving data to determine the violations in the current driving scenario; thus, it can automatically assess the compliance of autonomous driving based on the current driving data of the target vehicle; it not only meets the needs of autonomous driving, but also can be applied to different scenarios, improving the universality of the solution application.

[0046] In a preferred embodiment of this example, the evaluation results are fed back to the user, thereby improving the user experience.

[0047] In a preferred embodiment of this example, traffic regulations corresponding to the target vehicle model in the target country are converted into machine-readable logical and formal language to generate a first condition library and a first rule library; a mapping relationship is established between each legal field in the first condition library and the vehicle kinematics model to construct a condition analysis modeling system; the condition analysis modeling system is associated with a mathematical model to output a system parameter table and a condition mathematical model library; the first condition library, the first rule library, and the system parameter table are determined as a preset compliance engine system configuration file; the condition mathematical model library is converted into a code generation system to generate a mathematical function library.

[0048] For example, all non-measurable variable objects, such as dynamic objects like pedestrians, motor vehicles, and trucks, and static objects like zebra crossings, solid lines, dashed lines, no-entry zones, and traffic signs, are converted into corresponding geometric shapes and described using appropriate algebraic methods. For instance, pedestrians can be represented by rectangles, and motor vehicles and trucks can be described by rectangular hexahedrons, using the coordinates of the geometric shape's center and its length, width, and height. Solid lines on the ground can be described using polylines, using the coordinates of multiple points on the polyline. Based on this substitution, the two objects in the traffic law's rule "cars must not cross solid lines" (i.e., solid lines and cars) are converted into polylines and rectangles, respectively. The physical determination of whether a car crosses a solid line then becomes a computational geometry problem of whether the polyline lies within the rectangle.

[0049] For example, the step of converting traffic regulations documents corresponding to the target vehicle model in the target country into machine-readable logical and formal language to generate a first condition library and a first rule library includes: decomposing the traffic regulations documents corresponding to the target vehicle model in the target country into several traffic articles based on a regulation decomposer; for any target traffic article among the several traffic articles: performing content filling processing on the target traffic article using a semantic analysis model to output a standard traffic article; decomposing the standard traffic article into several first conditions based on the article decomposer; generating a first condition library corresponding to the traffic regulations document based on the several first conditions corresponding to each standard traffic article; and logically combining the first conditions in the first condition library based on a logical expression generator to generate a corresponding first rule library.

[0050] This section includes traffic regulations such as: Road Traffic Management Regulations, Road Traffic Law, UNECER 157 Automated Lane Keeping System Regulations, judicial interpretations, and case law. Target countries include: China, the United States, the United Kingdom, and Germany; target vehicle types include: large trucks, cars, buses, and coaches.

[0051] For example: based on the understanding of traffic regulations and the analysis of target traffic provisions using a semantic analysis model, some information of the target traffic provisions can be supplemented to output standard traffic provisions; such as... Figure 5 As shown, the blue part in the following text is the text supplemented by the semantic analysis model based on the understanding of the context. It completes the operational state described by the target traffic law, which is essential for subsequent judgments.

[0052] like Figure 6 As shown, a standard traffic law is decomposed into two conditions based on a law decomposer; each condition is parsed into an object and logic (i.e., a function expression).

[0053] Condition 1: Is there a vehicle (motor vehicle, lane) in the same lane ahead? == True;

[0054] Condition 2: Is (maintaining a safe distance from the vehicle in front (vehicle in front, vehicle behind) == True) == True;

[0055] Is there a vehicle (motor vehicle, lane) in the same lane ahead? (True) Whether (maintaining a safe distance from the vehicle in front (vehicle in front, vehicle behind) == True) == True → Violation of regulations = True; where ∧, , → are logical operators; "NOT" represents... The "AND" operator is represented by ∧, the "condition" operator is represented by →, and the "whether (...) == true" operator is a comparison operator.

[0056] At this point, the logical meaning changes from "when condition 1 is met, the system must execute or make the scenario described by condition 2" to "when condition 1 is met and condition 2 is not met, the system decision violates the relevant regulations".

[0057] For each condition, the system will search for similar conditions in the previously defined condition library and provide a corresponding formal language description, as shown in the following example:

[0058] The conditions described in the text: When there are motor vehicles traveling in the same lane =>;

[0059] The condition described in the formal language is: whether (there exists a vehicle (motor vehicle, lane) ahead) = true;

[0060] The conditions described in the text are: The following vehicle should maintain a distance sufficient to allow for emergency braking.

[0061] Formal language description of the conditions: whether or not) maintaining a safe distance from the vehicle in front (vehicle in front, vehicle behind) = true) = true;

[0062] When no directly applicable conditions are available, a recommended expression for the new conditions is generated based on the knowledge from the first condition library. Then, the existing formalized expression conditions are associated with known physical formulas, vehicle kinematics formulas, and mathematical models, and further transformed into a geometric model to construct a machine-understandable formal expression, i.e., a computable mathematical expression.

[0063] like Figure 7 As shown: The physical model for maintaining a safe distance from the vehicle in front (front vehicle, rear vehicle) is that the actual distance from your vehicle to the vehicle in front is greater than or equal to a safe distance x sufficient to take emergency braking measures. 前车 ≥x 安全 Here, "automobile" is used to indicate the target vehicle for autonomous driving.

[0064] Further transformation into a more refined mathematical model (variation) can lead to, for example, Figure 6 The physical and geometric variables are defined as shown.

[0065] Is it true (maintaining a safe distance from the vehicle in front, vehicle behind)?

[0066] Function expression 1: Maintain a safe distance from the vehicle in front (vehicle in front, vehicle behind);

[0067] x 安全 =Safe following distance (front vehicle and rear vehicle)

[0068] Equation (1);

[0069] if the car in front, x 前车 ≥x 安全 ;

[0070] Function expression 2: Safe following distance (front vehicle, rear vehicle);

[0071]

[0072] in,

[0073] x 安全 A safe distance sufficient to take emergency braking measures;

[0074] t 反应 =t 硬 +t 软 Reaction time;

[0075] t 硬 : The time required for the actuator of the corresponding vehicle braking system to generate maximum braking force from receiving a braking signal to the brake disc;

[0076] t 软 This corresponds to the time required for the driver of the vehicle to recognize an emergency and give a braking signal. In the case of a human driver, this is the human reaction time, while in the case of autonomous driving software, it corresponds to the time required by the environmental perception software stack and the autonomous driving decision-making software stack.

[0077] a 前车预期刹车力 : The maximum deceleration that can be applied when the vehicle in front brakes;

[0078] a 自车可用最大刹车力 The planned maximum deceleration that the vehicle will apply during emergency braking;

[0079] v 自车 The current speed of the vehicle (or the vehicle behind it);

[0080] v 前车 The current speed of the vehicle in front;

[0081] x 静止 The longitudinal distance between two vehicles when both vehicles in front and behind brake to zero speed.

[0082] In a preferred embodiment of this example, the method further includes: performing signal verification processing on the current driving data; if the verification result indicates that the current driving data passes the verification, then performing signal conversion processing on the current driving data to generate converted driving data; performing noise reduction processing on the converted driving data to obtain noise-reduced driving data. Then, based on the parsing result, a mathematical function library is called to perform a compliance assessment on the noise-reduced driving data, generating a compliance assessment result. This improves the accuracy of the compliance assessment of the current driving data.

[0083] In a preferred embodiment of this example, the preset compliance engine system configuration file includes at least: a first condition library, a first rule library, and a system parameter table.

[0084] For any first rule in the first rule base: perform rule parsing processing on the first rule base, and store the rule parsing results according to the corresponding parsing values; the rule parsing results include: at least two condition index numbers and at least one logical operator; select a first condition corresponding to each condition index number from the first condition base to generate a second condition base; for any first condition in the second condition base: perform condition parsing processing on the first condition, and store the condition parsing results according to the corresponding parsing values; the condition parsing results include: at least one object and at least one function name; parse the system parameter table and output the parsed system parameter table; determine the parsed system parameter table, the rule parsing results, and the condition parsing results as the parsing result. Therefore, this embodiment, by initializing the compliance engine system configuration file in the compliance engine, facilitates the compliance engine's effective evaluation of current driving data, improving the accuracy of compliance evaluation of the target vehicle during autonomous driving.

[0085] In a preferred embodiment of this example, the step of calling a mathematical function library to perform a compliance assessment on the current driving data based on the parsing result and generating a compliance assessment result includes: selecting a target object and a target function name corresponding to the current driving data from the parsing result; calling a mathematical function corresponding to the target function name from the mathematical function library; evaluating the target object using the mathematical function based on the parsed system parameter list to generate a real-time conditional assessment result; and performing a compliance assessment on the conditional assessment result based on the rule parsing result to generate a real-time rule assessment result.

[0086] like Figure 2 The diagram shown is an architectural schematic of a compliance engine system in one embodiment of the present invention.

[0087] The compliance engine system includes: a middleware layer, and compliance engine modules that communicate with the middleware layer; the compliance engine modules include: a data processing unit, a rule parser, a condition parser, a compliance evaluator, and a function retrieval unit; the compliance evaluator includes a condition evaluator and a rule evaluator.

[0088] The middle layer is used to perform signal verification and signal conversion processing on the raw driving data generated by the target vehicle during the driving process of autonomous driving, and send the converted driving data to the data processing unit.

[0089] The data processing unit is used to perform noise reduction processing on the converted driving data and send the noise-reduced driving data to the condition evaluator respectively;

[0090] The rule parser is used to parse the rules in the first rule base and output the rule parsing results; and sends the rule parsing results to the condition parser and the rule evaluator respectively.

[0091] The condition parser is used to query the first condition corresponding to the rule parsing result from the first condition library, generate the second condition library, perform condition parsing processing on the first condition in the second condition library, and output the condition parsing result; then the condition parsing result is sent to the condition evaluator and the function calling unit respectively.

[0092] The condition evaluator is used to select the target object and the name of the target function corresponding to the noise-reduced driving data from the parsing results; based on the parsed system parameter list, it evaluates the target object using the mathematical function corresponding to the name of the target function to generate a real-time condition evaluation result; and then sends the real-time condition evaluation result to the rule evaluator.

[0093] The function retrieval unit is used to retrieve the mathematical function corresponding to the name of the target function in the condition parsing result from the mathematical function library, and send the retrieved mathematical function to the condition evaluator.

[0094] The rule evaluator is used to perform compliance evaluation on the real-time condition evaluation result based on the rule parsing result, and generate a real-time rule evaluation result.

[0095] It should be noted that the final output of the rule evaluator is a rule evaluation expressed in "true" / "false" format, used for reporting or graphical display. The final output of the condition evaluator is a condition evaluation expressed in "true" / "false" format, used for reporting or graphical display.

[0096] like Figure 3 The diagram shown is a flowchart illustrating the rule parsing process for the first rule in one embodiment of the present invention.

[0097] The first rule base is parsed, and the parsing results are stored according to the corresponding parsed values; this includes at least the following steps:

[0098] S301, obtain the rule index number of the first rule;

[0099] S302, based on the rule index number, read and process the fields in the first rule to obtain the current field;

[0100] S303, Perform a test on the current field and obtain the first test result;

[0101] S304, if the first detection result indicates that the current field is a logical operator, then the logical operator is converted into the corresponding parsed value and stored;

[0102] S305, if the first detection result indicates that the current field is not a logical operator, then the condition index number is stored when the current field is determined to be a condition index number.

[0103] For example, if the detection result indicates that the current field is a logical operator, then converting the logical operator into a corresponding parsed value for storage includes: if the detection result indicates that the current field is a NOT operator, then searching for the ending field of the NOT operator; if the search result indicates that the ending field of the NOT operator is parentheses, then converting the NOT operator and the adjacent parentheses into corresponding parsed values ​​for storage; if the detection result indicates that the current field is not a NOT operator, then determining whether the current field is an AND or OR field; if the determination result indicates that the current field is an AND or OR field, then converting the AND or OR field into a corresponding parsed value for storage.

[0104] The specific steps are as follows:

[0105] S1, Obtain the rule index number of the first rule;

[0106] S2, based on the rule index number, read the fields in the first rule; if the reading result indicates that the current field does not exist, then determine the previous current field as the end of the first rule and store the parsing result; if the reading result indicates that the current field exists, then execute step S3.

[0107] S3, check if the current field is a NOT operator; if yes, proceed to step S4; if no, proceed to step S6.

[0108] S4: Search if parentheses appear immediately after the NOT operator; if yes, proceed to step S5; otherwise, return an error message.

[0109] S5, convert the NOT operator and adjacent parentheses into corresponding parsed values ​​and store them; continue executing step S2;

[0110] S6, determine whether the current field is an AND or OR field; if yes, proceed to step S7; if no, proceed to step S8.

[0111] S7, convert the AND or OR field into the corresponding parsed value and store it; continue to execute step S2;

[0112] S8. Determine if the current field is a left parenthesis; if yes, proceed to step S9; otherwise, proceed to step S10.

[0113] S9, use a counter to increment the number of left parentheses by 1, and if a right parenthesis corresponding to the left parenthesis is found after the search, convert the complete parenthesis into a parsed value and store it; continue to execute step S2;

[0114] S10, determine if the current field is a right parenthesis; if yes, proceed to step S11; if no, proceed to step S12.

[0115] S11: Decrement the number of left parentheses by 1 using a counter, and continue executing step S2 if the number of left parentheses is greater than or equal to 0; return an error message if the number of left parentheses is less than 0.

[0116] S12, determine whether the current field is a condition index number; if yes, store the condition index number; if no, return an error message.

[0117] During software initialization, the compliance engine reads the first rule from the first rule base line by line, converting the first rule base into a truth table. Simultaneously, it extracts all conditions from the first rule base and constructs a second condition base based on the first condition base, storing it in memory. Then, it parses the first conditions in the second condition base and outputs the parsing results. In each clock cycle, based on the current driving data, the parsing results, and the mathematical function library, it calculates the condition evaluation results. Finally, the compliance engine maps these condition evaluation results to the truth table to calculate the rule evaluation results corresponding to the current driving data.

[0118] like Figure 4 The diagram shown is a schematic flowchart of the condition parsing process for the first condition in one embodiment of the present invention.

[0119] The first condition is parsed, and the parsing result is stored according to the corresponding parsed value; this includes at least the following steps:

[0120] S401, obtain the condition index number of the first condition;

[0121] S402, based on the condition index number, read and process the fields in the first condition to obtain the current field;

[0122] S403, Perform a detection on the current field to obtain the second detection result;

[0123] S404, if the second detection result indicates that the current field is a valid mathematical operation quantity, then the parsed value corresponding to the mathematical operation quantity is stored; wherein, the mathematical operation quantity includes at least: mathematical operator name, and / or function name, and / or constant;

[0124] S405, if the second detection result indicates that the current field is not a valid mathematical operation, then when the current field is determined to be an object, the parsed value corresponding to the object is stored.

[0125] The specific steps are as follows:

[0126] S1, Obtain the condition index number of the first condition;

[0127] S2, based on the condition index number, read the fields in the first condition; if the reading result indicates that the current field does not exist, then determine the previous current field as the end of the first condition and store the parsing result; if the reading result indicates that the current field exists, then execute step S3.

[0128] S3, check if the current field is a left parenthesis; if not, proceed to step S4; if yes, use a counter to increment the number of left parentheses by 1 and proceed to step S2.

[0129] S4, check if the current field is a right parenthesis; if so, use a counter to decrement the number of left parentheses by 1, and continue to execute step S2 if the number of remaining left parentheses is greater than or equal to 0; if the number of remaining left parentheses is less than 0, return an error message; otherwise, execute step S5.

[0130] S5, determine if the current field is a mathematical operator name; if yes, find and store the index number corresponding to the mathematical operator name, and continue to execute step S2; if no, execute step S6.

[0131] S6, determine whether the current field is a function name; if yes, convert the function name into the corresponding parsed value and store it, then continue to execute step S2; if no, execute step S7.

[0132] S7. Determine whether the current field is a constant; if yes, convert the constant into the corresponding parsed value and store it, then continue with step S2; if no, proceed to step S8.

[0133] S8. Determine if the current field is an object; if so, convert the object to the corresponding parsed value and store it, then continue with step S2; otherwise, determine if the current field is invalid and return an error message.

[0134] like Figure 3 and Figure 4 The parser flowchart is shown. Generally, parsing is based on reading space-separated words and resolving function names, logical operators, mathematical operator names, objects, and constants, etc.

[0135] The parsing process begins with a text file containing conditions and rules. The condition text file is as follows: Figure 10 The second condition library is shown below; the second condition library consists of multiple lines of first conditions, each line starting with a condition ID, followed by the condition name and its input. For example, a first condition is: "1: There is a vehicle (motor vehicle, lane) in front in the same lane == True".

[0136] Rule text file such as Figure 9 The first rule base is shown; the first rule base consists of multiple lines of first rules. Each first rule is defined as a logical combination condition, so each line in the file begins with a rule ID, followed by a condition index number ID composed of logical operators (such as "AND", "OR", "NOT"), for example, the first rule: "1:1 AND NOT (2)". In this example, 1 and 2 are the condition IDs defined in the first condition base.

[0137] The condition parser first reads the first condition of each line in the second condition library and extracts the first condition ID. Then, it reads the fields and checks if the current field is a function name followed by an open bracket; if so, it marks the current field as a function name; it then continues reading the fields following the function name until it reaches the closing bracket. The fields following the function name are used as input to the function, which can be objects or parameters. In this example, vehicles and lanes are objects.

[0138] The rule parser reads the first rule; it first obtains the rule index number ID corresponding to the first rule, and then reads the field as a condition ID or logical operator.

[0139] like Figure 11 The following is a detailed description of this embodiment in conjunction with specific scenarios.

[0140] An assessment method for autonomous driving compliance includes at least the following steps:

[0141] S1 retrieves the traffic regulations documents corresponding to the target vehicle model in the target country from the global traffic regulations and autonomous driving regulations database; traffic regulations documents include: road traffic management regulations, road traffic laws, UNECER 157 regulations on automatic lane keeping systems, judicial interpretations and precedents; and sends the traffic regulations documents to the natural language parsing and quantitative analysis system.

[0142] S2, in the natural language parsing and quantitative analysis system, the regulation decomposer decomposes the traffic regulation documents corresponding to the target vehicle model in the target country into several traffic articles; for any target traffic article among the several traffic articles: the semantic analysis model performs content filling processing on the target traffic article and outputs a standard traffic article; the regulation decomposer decomposes the standard traffic article into several first conditions; based on the several first conditions corresponding to each standard traffic article, a first condition library corresponding to the traffic regulation document is generated; the logical expression generator performs logical combination on the first conditions in the first condition library to generate a corresponding first rule library.

[0143] A mapping relationship is established between each legal field in the first condition library and the vehicle kinematics model to construct a condition analysis modeling system; the condition analysis modeling system is associated with a mathematical model to output a system parameter table and a condition mathematical model library; the first condition library, the first rule library, and the system parameter table are determined as the preset compliance engine system configuration file; the condition mathematical model library is converted into a code generation system to generate a mathematical function library.

[0144] The natural language parsing and quantitative analysis system sends the preset compliance engine system configuration file to the compliance engine system.

[0145] S3, the autonomous vehicle inputs current driving data as an input signal to the compliant engine system; the current driving data includes at least: environmental perception data collected by the environmental perception module, and automatic software path planning;

[0146] In S4, within the compliance engine system, the parser parses the preset compliance engine system configuration file and sends the rule parsing results to the rule evaluator; it also sends the condition parsing results and the parsed system parameter table to the condition evaluator; the condition evaluator calls mathematical functions from the mathematical function library based on the condition parsing results; the condition evaluator performs condition evaluation based on the condition parsing results, input signals, mathematical functions, and the parsed system parameter table, and outputs real-time condition evaluation results; the condition evaluator sends the real-time condition evaluation results to both the rule evaluator and the compliance report generator; the rule evaluator performs rule evaluation based on the rule parsing results and the real-time condition evaluation results, outputs real-time rule evaluation results, and sends the real-time rule evaluation results to the compliance report generator. Simultaneously, the rule evaluator sends real-time violation alerts to the autonomous vehicle based on the real-time rule evaluation results.

[0147] The compliance engine system sends input signals to the compliance report generator.

[0148] S5, in the compliance report generator, a list of real-time condition assessment results (Boolean type) is generated based on the real-time condition assessment results; a list of real-time rule assessment results (Boolean type) is also generated based on the real-time rule assessment results. The traffic trigger analyzer obtains the real-time condition assessment list and performs traffic rule trigger analysis on the input signal based on the real-time condition assessment list, generating a regulation trigger report; the violation analysis statistician obtains the real-time rule assessment table and performs violation analysis on the real-time rule assessment list, outputting a compliance statistics report; the event recorder obtains the real-time rule assessment table and the input signal, evaluates the input signal based on the real-time rule assessment list, and outputs an event record report.

[0149] The compliance report generator sends regulatory trigger reports, compliance statistics reports, and event log reports to the compliance cloud respectively.

[0150] S6, Compliance Cloud sent the above reports to OEMs, operators, regulatory authorities, and insurance companies respectively.

[0151] This embodiment, based on a pre-defined text processing workflow and machine learning algorithms, transforms the legal language of traffic laws into machine-readable logical and formalized language (combining geometry, algebra, and logical expressions). Simultaneously, relevant case precedents are transformed into traffic scenarios and output in a specific descriptive manner. Secondly, the system automatically generates corresponding mathematical and geometric expressions based on the logical and formalized language, converts them into code, and embeds them into the compliance engine system in the form of rules, generating a compliance judgment software module and automatically generating embedded plugins for interfaces corresponding to different simulation software or hardware chips. Finally, the generated software module, in the form of a SoC, is installed in the vehicle or related hardware. By receiving and analyzing the output signals of the corresponding environmental perception modules in the autonomous driving software stack, as well as the planned paths or decisions of the autonomous driving software, it can not only determine which traffic regulations are relevant in the current driving scenario and record and statistically analyze them in real time; it can also determine whether the autonomous driving software's decisions violate traffic regulations, output corresponding feedback, and record data on traffic violations. This embodiment's method can also dynamically evaluate the compliance and safety of the autonomous driving software stack and provide feedback and reports; furthermore, it can combine map information to dynamically evaluate the vehicle hardware's working status and provide feedback and reports. This satisfies the target vehicle's autonomous driving requirements.

[0152] Because the method in this embodiment transforms the legal language in traffic laws into machine-readable logical and formalized language, and generates embedded language software code based on the logical and mathematical relationships, it has universality and is applicable to all autonomous driving scenarios.

[0153] like Figure 12 The diagram shown is a structural schematic of an evaluation device for autonomous driving compliance provided in an embodiment of the present invention.

[0154] An evaluation device for autonomous driving compliance, the device 120 includes: a data acquisition module 121 for acquiring current driving data generated by a target vehicle during autonomous driving; a parsing module 122 for parsing a preset compliance engine system configuration file to obtain a parsing result; and an evaluation module 123 for using a mathematical function library to evaluate the current driving data based on the parsing result and generating a compliance evaluation result.

[0155] In a preferred embodiment of this invention, the device further includes: a first generation module, used to convert traffic regulations documents corresponding to the target vehicle model in the target country into machine-readable logical and formal language to generate a first condition library and a first rule library; a construction module, used to establish a mapping relationship between each legal field in the first condition library and the vehicle kinematics model to construct a condition analysis modeling system; an association module, used to associate the condition analysis modeling system with a mathematical model to output a system parameter table and a condition mathematical model library; a determination module, used to determine the first condition library, the first rule library, and the system parameter table as a preset compliance engine system configuration file; and a second generation module, used to convert the condition mathematical model library into a code generation system to generate a mathematical function library.

[0156] In a preferred embodiment of this example, the first generation module includes: a regulation decomposition unit, used to decompose traffic regulation documents corresponding to a target vehicle model in a target country into several traffic articles based on a regulation decomposition device; a content filling unit, used to perform content filling processing on any target traffic article among the several traffic articles using a semantic analysis model, outputting a standard traffic article; and decomposing the standard traffic article into several first conditions based on the article decomposition device; a article decomposition unit, used to generate a first condition library corresponding to the traffic regulation document based on the several first conditions corresponding to each standard traffic article; and a generation unit, used to logically combine the first conditions in the first condition library based on a logical expression generator to generate a corresponding first rule library.

[0157] In a preferred embodiment of this example, the preset compliance engine system configuration file includes at least: a first condition library, a first rule library, and a system parameter table; the parsing module includes: a rule parsing processing unit, used to perform rule parsing processing on the first rule library for any first rule in the first rule library, and store the rule parsing result according to the corresponding parsing value; the rule parsing result includes: at least two condition index numbers and at least one logical operator; a selection unit, used to select a first condition corresponding to each condition index number from the first condition library to generate a second condition library; a condition parsing processing unit, used to perform condition parsing processing on any first condition in the second condition library, and store the condition parsing result according to the corresponding parsing value; the condition parsing result includes: at least one object and at least one function name; a parameter parsing processing unit, used to parse the system parameter table and output the parsed system parameter table; and a determination unit, used to determine the parsed system parameter table, the rule parsing result, and the condition parsing result as the parsing result.

[0158] In a preferred embodiment of this example, the rule parsing processing unit includes: an acquisition subunit for acquiring the rule index number of the first rule; a reading processing subunit for reading and processing fields in the first rule based on the rule index number to acquire the current field; a detection subunit for detecting the current field to obtain a first detection result; a conversion processing subunit for converting the logical operator into a corresponding parsed value and storing it if the first detection result indicates that the current field is a logical operator; and a determination subunit for storing the condition index number when determining that the current field is a condition index number if the first detection result indicates that the current field is not a logical operator.

[0159] In a preferred embodiment of this example, the conversion processing subunit includes: a first conversion unit, configured to: if the first detection result indicates that the current field is a NOT operator, search for the ending field of the NOT operator; if the search result indicates that the ending field of the NOT operator is parentheses, convert the NOT operator and the adjacent parentheses into corresponding parsed values ​​and store them; and a second conversion unit, configured to: if the first detection result indicates that the current field is not a NOT operator, determine whether the current field is an AND or OR field; if the determination result indicates that the current field is an AND or OR field, convert the AND or OR field into corresponding parsed values ​​and store them.

[0160] In a preferred embodiment of this example, the condition parsing processing unit includes: an acquisition subunit for acquiring the condition index number of the first condition; a reading subunit for reading and processing the fields in the first condition based on the condition index number to acquire the current field; a detection subunit for detecting the current field to obtain a second detection result; a conversion processing subunit for storing the parsed value corresponding to the mathematical operation if the second detection result indicates that the current field is a valid mathematical operation; wherein the mathematical operation includes at least: mathematical operator names, and / or function names, and / or constants; and a determination subunit for storing the parsed value corresponding to the object when the current field is determined to be an object if the second detection result indicates that the current field is not a valid mathematical operation.

[0161] In a preferred embodiment of this example, the evaluation module includes: a selection unit, used to select a target object and a target function name corresponding to the current driving data from the parsing results; a calling unit, used to call a mathematical function corresponding to the target function name from a mathematical function library; a condition evaluation unit, used to evaluate the target object using the mathematical function based on the parsed system parameter list, and generate a real-time condition evaluation result; and a rule evaluation unit, used to perform a compliance evaluation on the condition evaluation result based on the rule parsing result, and generate a real-time rule evaluation result.

[0162] In a preferred embodiment of this invention, the device further includes: a third generation module, configured to generate corresponding real-time condition evaluation lists and real-time rule evaluation lists based on the real-time condition evaluation results and real-time rule evaluation results corresponding to each of the current driving data; a traffic regulation trigger analysis module, configured to perform traffic regulation trigger analysis on the current driving data based on the real-time condition evaluation lists and generate a regulation trigger report; a violation analysis module, configured to perform violation analysis on the real-time rule evaluation lists and output a compliance statistics report; and an event evaluation module, configured to evaluate the current driving data based on the real-time rule evaluation lists and output an event record report.

[0163] The aforementioned apparatus can execute the autonomous driving compliance assessment method provided in an embodiment of the present invention, and possesses the corresponding functional modules and beneficial effects for executing the autonomous driving compliance assessment method. Technical details not described in detail in this embodiment can be found in the autonomous driving compliance assessment method provided in an embodiment of the present invention.

[0164] The present invention also provides an electronic device, comprising: a processor; a memory for storing executable instructions of the processor; the processor being configured to read the executable instructions from the memory and execute the instructions to implement the assessment method for autonomous driving compliance described in the present invention.

[0165] In addition to the methods and apparatus described above, embodiments of this application may also be computer program products, which include computer program instructions that, when executed by a processor, cause the processor to perform the steps in the methods according to various embodiments of this application described in the "Exemplary Methods" section above.

[0166] The computer program product can be written in any combination of one or more programming languages ​​to perform the operations of the embodiments of this application. The programming languages ​​include object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0167] Furthermore, embodiments of this application may also be computer-readable storage media storing computer program instructions thereon, which, when executed by a processor, cause the processor to perform the steps of the methods according to the following embodiments of this application as described in the "Exemplary Methods" section above.

[0168] The computer-readable storage medium may be any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may, for example, include, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatuses, or devices, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: electrical connections having one or more wires, portable disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0169] The basic principles of this application have been described above with reference to specific embodiments. However, it should be noted that the advantages, benefits, and effects mentioned in this application are merely examples and not limitations, and should not be considered as essential features of each embodiment of this application. Furthermore, the specific details disclosed above are for illustrative and facilitative purposes only, and are not limitations. These details do not limit the application to the necessity of employing the aforementioned specific details for implementation.

[0170] The block diagrams of devices, apparatuses, devices, and systems involved in this application are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, devices, and systems can be connected, arranged, and configured in any manner. Words such as “comprising,” “including,” “having,” etc., are open-ended terms meaning “including but not limited to,” and are used interchangeably with them. The terms “or” and “and” as used herein refer to the terms “and / or,” and are used interchangeably with them unless the context clearly indicates otherwise. The term “such as” as used herein refers to the phrase “such as but not limited to,” and is used interchangeably with it.

[0171] It should also be noted that in the apparatus, equipment, and methods of this application, the components or steps can be disassembled and / or recombined. These disassemblies and / or recombinations should be considered as equivalent solutions of this application.

[0172] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use this application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein can be applied to other aspects without departing from the scope of this application. Therefore, this application is not intended to be limited to the aspects shown herein, but rather to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0173] The above description has been given for purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of this application to the forms disclosed herein. Although numerous exemplary aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations thereof.

[0174] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of those different embodiments or examples.

[0175] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0176] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for assessing compliance with autonomous driving regulations, characterized in that, include: Acquire the current driving data generated by the target vehicle during the driving process; wherein, the current driving data includes at least: environmental perception data collected by the environmental perception module, vehicle status data, and path planning; The traffic regulations documents corresponding to the target vehicle model in the target country are converted into machine-readable logical and formal language to generate a first condition library and a first rule library; each legal field in the first condition library is mapped to a vehicle kinematics model to construct a condition analysis modeling system; the condition analysis modeling system is associated with a mathematical model to output a system parameter table and a condition mathematical model library; the condition mathematical model library is converted into a code generation system to generate a mathematical function library. For any first rule in the first rule base: perform rule parsing processing on the first rule, and store the rule parsing result according to the corresponding parsing value; the rule parsing result includes: at least two condition index numbers and at least one logical operator; select a first condition corresponding to each condition index number from the first condition base to generate a second condition base; for any first condition in the second condition base: perform condition parsing processing on the first condition, and store the condition parsing result according to the corresponding parsing value; parse the system parameter table and output the parsed system parameter table; determine the parsed system parameter table, the rule parsing result, and the condition parsing result as the parsing result; Based on the analysis results, a mathematical function library is invoked to perform a compliance assessment on the current driving data, generating a compliance assessment result.

2. The method according to claim 1, characterized in that, The condition resolution result includes: at least one object and at least one function name; Based on the parsing results, a mathematical function library is invoked to perform a compliance assessment on the current driving data, generating a compliance assessment result; including: Select the target object and target function name corresponding to the current driving data from the condition parsing results; Call the mathematical function corresponding to the name of the target function from the mathematical function library; Based on the parsed system parameter list, the target object is evaluated using the mathematical function to generate a real-time conditional evaluation result; Based on the results of parsing several rules, a compliance assessment is performed on the real-time condition assessment results to generate real-time rule assessment results.

3. The method according to claim 1, characterized in that, The step of converting traffic regulations corresponding to the target vehicle model in the target country into a machine-readable logical and formal language to generate a first condition library and a first rule library includes: Based on the regulation decomposition tool, the traffic regulations documents corresponding to the target vehicle model in the target country are decomposed into several traffic law articles; For any target traffic law among the aforementioned traffic laws: the semantic analysis model performs content filling processing on the target traffic law and outputs a standard traffic law; The standard traffic law is broken down into several first conditions using a law decomposer. Based on several first conditions corresponding to each of the aforementioned standard traffic laws and regulations, a first condition library corresponding to traffic law documents is generated. The logical expression generator performs logical combinations on the first conditions in the first condition library to generate the corresponding first rule library.

4. The method according to claim 1, characterized in that, The step of parsing the first rule base and storing the parsing results according to the corresponding parsed values ​​includes: Obtain the rule index number of the first rule; Based on the rule index number, the fields in the first rule are read and processed to obtain the current field; The current field is tested to obtain the first test result; If the first detection result indicates that the current field is a logical operator, then the logical operator is converted into a corresponding parsed value and stored. If the first detection result indicates that the current field is not a logical operator, then the condition index number is stored when the current field is determined to be a condition index number.

5. The method according to claim 4, characterized in that, If the first detection result indicates that the current field is a logical operator, then the logical operator is converted into a corresponding parsed value and stored; including: If the first detection result indicates that the current field is a NOT operator, then search for the ending field of the NOT operator. If the search result indicates that the ending field of the NOT operator is parentheses, then convert the NOT operator and the adjacent parentheses into corresponding parsed values ​​and store them. If the first detection result indicates that the current field is not a NOT operator, then determine whether the current field is an AND or OR field; if the determination result indicates that the current field is an AND or OR field, then convert the AND or OR field into the corresponding parsed value and store it.

6. The method according to claim 1, characterized in that, The step of performing condition parsing on the first condition and storing the condition parsing result according to the corresponding parsed value includes: Obtain the condition index number of the first condition; Based on the condition index number, the fields in the first condition are read and processed to obtain the current field; The current field is then tested to obtain a second test result; If the second detection result indicates that the current field is a valid mathematical operation, then the parsed value corresponding to the mathematical operation is stored to generate a conditional parsing result; wherein, the mathematical operation includes at least: mathematical operator name, and / or function name, and / or constant; If the second detection result indicates that the current field is not a valid mathematical operation, then when the current field is determined to be an object, the parsed value corresponding to the object is stored.

7. The method according to claim 1, characterized in that, Also includes: Acquire the raw driving data generated by the target vehicle during its autonomous driving process; The original driving data is subjected to signal verification processing; If the verification result indicates that the original driving data passes the verification, then the original driving data is subjected to signal conversion processing to generate converted driving data; The converted driving data is then subjected to noise reduction processing to obtain the current driving data.

8. The method according to claim 2, characterized in that, Also includes: Based on the real-time condition evaluation results and real-time rule evaluation results corresponding to each current driving data, a corresponding real-time condition evaluation list and a real-time rule evaluation list are generated respectively. Based on the real-time condition assessment list, the current driving data is analyzed for traffic regulations triggering to generate a regulation triggering report. The real-time rule evaluation list is analyzed for violations, and a compliance statistics report is output. The current driving data is evaluated based on the real-time rule evaluation list, and an event log report is output.

9. A compliance engine system for autonomous driving compliance assessment, characterized in that, The preset compliance engine system configuration file includes at least: a first condition library, a first rule library, and a system parameter table; the compliance engine system includes: a middle layer, and a compliance engine module that communicates with the middle layer; the compliance engine module includes: a data processing unit, a rule parser, a condition parser, a condition evaluator, a rule evaluator, and a function call unit; The middle layer is used to perform signal verification and signal conversion processing on the raw driving data generated by the target vehicle during the driving process of autonomous driving, and send the converted driving data to the data processing unit. The data processing unit is used to perform noise reduction processing on the converted driving data and send the noise-reduced current driving data to the condition evaluator; The rule parser is used to parse the rules in the first rule base and output the rule parsing results; and sends the rule parsing results to the condition parser and the rule evaluator respectively. A condition parser is used to query a first condition from a first condition library that corresponds to the rule parsing result, generate a second condition library, perform condition parsing processing on the first condition in the second condition library, and output the condition parsing result; then, the condition parsing result is sent to the condition evaluator and the function calling unit respectively; wherein, the condition parsing result includes: at least one object and at least one function name; The function retrieval unit is used to retrieve the mathematical function corresponding to the name of the target function in the condition parsing result from the mathematical function library, and send the retrieved mathematical function to the condition evaluator; The condition evaluator is used to select the target object and the name of the target function corresponding to the current driving data from the condition parsing result; based on the parsed system parameter list, it evaluates the target object using the mathematical function corresponding to the name of the target function, generates a real-time condition evaluation result, and sends the real-time condition evaluation result to the rule evaluator. A rule evaluator is used to perform compliance evaluation on the real-time condition evaluation result based on the rule parsing result, and generate a real-time rule evaluation result.

10. An assessment device for autonomous driving compliance, characterized in that, The acquisition module is used to acquire the current driving data generated by the target vehicle during the driving process of the autonomous driving system; wherein, the current driving data includes at least: environmental perception data acquired by the environmental perception module, vehicle status data, and path planning. The first generation module is used to convert the traffic regulations documents corresponding to the target vehicle model in the target country into machine-readable logical and formal language, and generate the first condition library and the first rule library. The construction module is used to establish a mapping relationship between each legal field in the first condition library and the vehicle kinematics model, and to build a condition analysis modeling system; The association module is used to associate the conditional analysis and modeling system with the mathematical model, and output the system parameter table and the conditional mathematical model library; The determination module is used to convert the conditional mathematical model library into a code generation system, generating a mathematical function library; The parsing module is configured to: perform rule parsing processing on any first rule in the first rule base, and store the rule parsing result according to the corresponding parsing value; the rule parsing result includes: at least two condition index numbers and at least one logical operator; select a first condition corresponding to each condition index number from the first condition base to generate a second condition base; for any first condition in the second condition base: perform condition parsing processing on the first condition, and store the condition parsing result according to the corresponding parsing value; parse the system parameter table and output the parsed system parameter table; the parsed system parameter table, the rule parsing result, and the condition parsing result are determined as the parsing result; The evaluation module is used to perform a compliance evaluation on the current driving data based on the parsing results by calling a mathematical function library, and generate a compliance evaluation result.

11. A computer-readable medium having a computer program stored thereon, which, when executed by a processor, implements the method as claimed in any one of claims 1 to 8.

12. An electronic device, comprising: processor; Memory used to store the processor's executable instructions; The processor is configured to read the executable instructions from the memory and execute the instructions to implement the method as described in any one of claims 1 to 8.

Images