A method for evaluating the usability of an autonomous vehicle-to-vehicle interaction system

Abstract

The application provides a kind of availability evaluation method of autonomous vehicle interaction system.The method is aimed at the problem that interaction function is unavailable due to vehicle hardware and software failure and factors such as perception error, communication delay and transmission packet loss, divides the system into three levels of vehicle element, interaction subsystem and interaction system; offline simulation is carried out according to interaction category and structure category, the continuous available time and continuous unavailable time are counted, the instantaneous availability and steady-state availability of the function are calculated based on the alternate updating process and the index library is established; the vehicle element availability is calculated combined with the failure, repair and aging characteristics of the perception, transmission, decision and control modules; the interaction subsystem availability is obtained by coupling the function availability and the vehicle element availability based on the Bayes formula, and the instantaneous availability and average availability of the interaction system are obtained by shared vehicle deduplication and layer-by-layer aggregation, to realize the interpretable and reusable evaluation of the dynamic vehicle interaction system availability.

Description

Technical Field

[0001] This invention relates to the fields of autonomous driving and intelligent connected vehicle technology, and in particular to a method for assessing the availability of an autonomous vehicle-to-vehicle interaction system. Background Technology

[0002] With the continuous maturation of autonomous driving and intelligent connected vehicle technologies, ATS (Autonomous Transportation Systems) has become an important development direction in the future transportation field. ATS, with its core capabilities of autonomous perception, autonomous learning, autonomous decision-making, and autonomous response, breaks through the traditional driver-centric passive driving mode of transportation systems, driving the evolution of transportation systems towards an unmanned or minimally human-controlled proactive autonomous paradigm. Compared with traditional transportation systems, ATS significantly reduces reliance on human intervention, and its overall autonomous operation capability is greatly enhanced.

[0003] Autonomous vehicle-to-vehicle (V2V) interaction systems are the microscopic manifestation of Automatic Transportation Systems (ATS). They consist of multiple vehicles and their dynamic interactions, relying on the collaborative hardware and software support of onboard perception, transmission, decision-making, and control modules to achieve autonomous and coordinated operation in specific traffic scenarios. During actual system operation, it is difficult for hardware and software to completely avoid occasional failures and aging-related malfunctions. V2V interaction functions may also experience interaction conflicts due to insufficient algorithm accuracy, limited computing power, transmission interruptions, and interference from unforeseen events. Therefore, whether the system can remain available at all times and whether autonomous vehicles can safely interact with other vehicles at any time has become a critical issue that urgently needs to be addressed in the design and safety assurance of autonomous V2V interaction systems.

[0004] Availability is a comprehensive indicator that measures a system's ability to continuously maintain its expected functionality. It is defined as the probability that a system will be in a working state at a given moment or within a given time interval, provided that external resources are guaranteed. It is a core indicator for system evaluation and operational assurance. However, existing availability evaluation research mainly focuses on static topology systems, with modeling objects concentrated on the failure and repair processes at the hardware and software levels, making it difficult to directly apply to autonomous vehicle-to-vehicle interaction systems. Specifically, existing technologies have the following shortcomings: First, existing evaluation methods are difficult to adapt to the dynamic topology characteristics of autonomous vehicle-to-vehicle interaction systems. Vehicles, as mobile nodes in the road network, are in a continuous state of motion, frequently switching between scenarios such as platooning, coordinated lane changing, and intersection games. The system topology updates dynamically accordingly, and existing methods cannot achieve a comprehensive evaluation of multiple scenarios and multiple system interaction relationships within a unified framework. Second, traditional availability modeling only considers hardware and software failures as triggering conditions for unavailability, ignoring functional-level unavailability scenarios. In autonomous vehicle-to-vehicle interaction systems, even if the hardware and software of each vehicle are operating normally, uncertainties such as transmission delay, perception error, and control deviation may still lead to a decline or even failure of the interaction function. Existing methods lack a systematic means of characterizing the availability of the entire functional link of perception, transmission, decision-making, and control.

[0005] To address the aforementioned issues, existing availability assessment methods primarily target systems with fixed structures and states directly describable by hardware and software failures. These methods struggle to adapt to the dynamic participation of vehicles, continuously changing interaction relationships, and multi-source disturbances affecting functional implementation in autonomous vehicle-to-vehicle (V2V) systems. Especially when vehicle hardware and software operate normally, communication latency, perception errors, and control deviations can still render interactive functions unavailable. Therefore, there is an urgent need for a availability assessment method for autonomous V2V systems that can simultaneously consider the differences between the vehicle's basic operational capabilities and its interactive function implementation capabilities, and characterize the coupling relationship between the two. Summary of the Invention

[0006] Embodiments of the present invention provide a usability assessment method for an autonomous vehicle-to-vehicle interaction system, so as to effectively assess the usability of the autonomous vehicle-to-vehicle interaction system.

[0007] To achieve the above objectives, the present invention adopts the following technical solution.

[0008] A usability evaluation method for an autonomous vehicle-to-vehicle interaction system includes: Obtain the basic input information required to calculate the availability of the autonomous vehicle-to-vehicle interaction system. This basic input information includes test range information and vehicle parameters. An autonomous vehicle-to-vehicle interaction system model is constructed based on the aforementioned basic input information; Calculate the functional availability of the autonomous vehicle-to-vehicle interaction system based on the autonomous vehicle-to-vehicle interaction system model; Calculate the availability of elements in an autonomous vehicle-to-vehicle interaction system based on an autonomous vehicle-to-vehicle interaction system model; The availability of the autonomous vehicle-to-vehicle interaction system is calculated based on the functional availability and element availability of the autonomous vehicle-to-vehicle interaction system.

[0009] Preferably, the basic input information required to obtain the availability of the autonomous vehicle-to-vehicle interaction system includes test range information and vehicle parameters, including: To obtain the basic input information required to calculate the availability of the autonomous vehicle-to-vehicle interaction system, the basic input information includes test range information and vehicle parameters, wherein the test range information includes the test space range. Test time range Scope of test subjects and scope of interaction categories Where F, L, and I represent formation interaction, coordinated lane-changing interaction, and intersection game interaction, respectively; the vehicle parameters include the hardware parameters, software parameters, motion state parameters, and disturbance parameters of each vehicle element, and the disturbance parameter vector These represent the position measurement error, velocity measurement error, acceleration control error, system delay, and transmission packet loss rate, respectively; and let the current evaluation time be... Establish an initial empty functional availability index library. .

[0010] Preferably, the construction of the autonomous vehicle-to-vehicle interaction system model based on the basic input information includes: Based on vehicle operation data, vehicle elements, interaction edges, interaction subsystems, and interaction systems are identified to construct an autonomous vehicle-to-vehicle interaction system model. A single vehicle is defined as an element comprising a perception module, a transmission module, a decision-making module, and a control module. The perception and transmission modules are connected in parallel, and then connected in series with the decision-making and control modules. The interaction relationship between two vehicle elements is defined as an interaction edge. The smallest interaction structure consisting of several vehicle elements with direct interaction relationships and the interaction edges between them is defined as an interaction subsystem. According to the interaction subsystem The interaction category, number of participating vehicles, vehicle role composition, and set of interaction edges or conflict relationships are used to match the corresponding interaction structure category number for this interaction subsystem. The largest interactive structure, formed by merging several interconnected interactive subsystems that share vehicle elements, is defined as an interactive system. ; For any two interaction subsystems in the set of interaction subsystems and If the vehicle element sets of the two sets intersect, that is... If the two are related, they are classified into the same interaction system; if multiple interaction subsystems are continuously related by sharing vehicle elements, then even if some of the interaction subsystems do not directly share vehicle elements, these multiple interaction subsystems will be classified into the same interaction system; for a single interaction subsystem that does not share vehicle elements with other interaction subsystems, it will be treated as a separate interaction system.

[0011] Preferably, the calculation of the functional availability of the autonomous vehicle-to-vehicle interaction system based on the autonomous vehicle-to-vehicle interaction system model includes: For each interaction structure category identified at the current evaluation time t Compare it with the aforementioned functional availability index library Compare: If the interaction structure category Not in the index If a new interaction structure category is identified, offline simulation testing is conducted. This offline simulation testing includes: constructing a motion model according to the corresponding interaction category and setting the total number of tests and the duration of each test; determining the interaction function status at each time step in multiple offline tests based on the judgment index; statistically obtaining the continuous available duration sample set and the continuous unavailable duration sample set for this interaction structure category; performing distribution fitting on the two sample sets respectively; estimating candidate distribution parameters using the maximum likelihood method; and determining the probability density functions of the available duration distribution and the unavailable duration distribution based on the Akaike Information Criterion. , The probability density function , After performing a Laplace transform, input the equations established based on the alternating update process. And perform numerical inversion. We obtain the instantaneous availability of the functionality of this interaction structure category; using the final value theorem, we obtain the steady-state availability of the functionality of this interaction structure category. and The first The average duration of continuous availability and the average duration of continuous unavailability of the interaction structure are assigned to the category number of that interaction structure. The index is stored in the aforementioned functional availability index library. ; If the interaction structure category Already in the index If it is found to be a previously tested old interaction structure category, the index library will be called directly. The interactive structure is categorized by number. Instantaneous availability of index storage functionality Offline simulation testing will no longer be performed repeatedly.

[0012] Preferably, the calculation of the functional availability of the autonomous vehicle-to-vehicle interaction system based on the autonomous vehicle-to-vehicle interaction system model further includes: The interaction structure categories include: platooning interaction, cooperative lane-changing interaction, and intersection game interaction. The vehicle's interaction functions are defined by predictive and real-time execution indicators. Predictive indicators include time to collision (TTC), 2D-TTC, and estimated post-collision intrusion time (PPET). Real-time execution indicators include speed, acceleration, and trigger timeliness. The interaction function states of interactive subsystems with different interaction structures are determined time-by-time. The values ​​of the predictive indicators are extrapolated from the current vehicle state, and the disturbance error term is calculated using a first-order sensitivity propagation method, with determination performed in a probabilistic sense. The values ​​of the real-time execution indicators are directly given by the actual operating state at the current moment, and deterministic determination based on the realized value is adopted. An asymmetric hysteresis determination mechanism is introduced, and the out-of-bounds probability is calculated based on the safety margin of the indicator after disturbance relative to the threshold. With regression probability When the probability of an indicator going out of bounds reaches a preset risk level, the interactive function is switched from available to unavailable. When the regression probability of all indicators is not lower than the preset confidence level, the interactive function is restored from unavailable to available. Otherwise, the function status of the previous moment is maintained.

[0013] Preferably, the offline simulation test employs corresponding motion models for different interaction categories: for platooning interaction, a car-following model is used to ensure that vehicles in front and behind maintain longitudinal platooning and car-following throughout the entire test period, and the functional state duration of the platooning interaction is extracted throughout; for cooperative lane-changing interaction, a cooperative lane-changing model is used to ensure that lane-changing vehicles repeatedly initiate and complete lane changes during the test period, and only the duration from triggering to completion of each lane-changing interaction is extracted and included in the functional state duration; for intersection game interaction, an intersection game model is used to ensure that vehicles with spatiotemporal conflicts repeatedly enter the conflict zone to engage in game during the test period, and only the duration of each game interaction is extracted and included in the functional state duration.

[0014] Preferably, the calculation of the element availability of the autonomous vehicle-to-vehicle interaction system based on the autonomous vehicle-to-vehicle interaction system model includes: For each vehicle element participating in the evaluation Based on the hardware failure rate of its sensing module, transmission module, decision-making module, and control module. Hardware repair rate Software failure rate and software recovery rate Calculate the hardware availability of each module separately. Software availability The availability of each module is obtained by multiplying the results. Then, the availability of the perception module With the availability of transmission modules Combined in parallel and with decision module availability and control module availability By combining them in a series manner, the vehicle element at the current evaluation time can be obtained. Corresponding cumulative runtime Vehicle element availability: ; In the formula, Representing vehicle elements Availability; , These represent the availability of the vehicle's perception module and transmission module, respectively.

[0015] Preferably, when the interaction subsystem is within a finite duration, the instantaneous availability of the function of its corresponding interaction structure category is called with the duration of the interaction subsystem as the independent variable; when the evaluation object is a long-term operation scenario or a steady-state operation scenario, the steady-state availability of the function of its corresponding interaction structure category is called.

[0016] Preferably, the calculation of the availability of the autonomous vehicle-to-vehicle interaction system based on the functional availability and element availability of the autonomous vehicle-to-vehicle interaction system includes: For each interactive subsystem, determine the functional availability of the interactive structure category to which that subsystem belongs. Vehicle element availability of each vehicle element included in this interactive subsystem By coupling using Bayes' theorem, the interaction subsystem at the current evaluation time can be obtained. Availability :

[0017] For each interactive system, the availability of each interactive subsystem within that system is aggregated layer by layer. Furthermore, shared vehicle elements that appear repeatedly across multiple interactive subsystems are ranked according to their frequency of occurrence. After deduplication correction, the instantaneous availability of the interactive system at the current evaluation time t is obtained: .

[0018] Preferably, the method further includes: Compare the current assessment time With the end time of the test :like Then let Then, repeat the above process of constructing the autonomous vehicle-to-vehicle interaction system model, calculating the functional availability, element availability, and overall availability of the autonomous vehicle-to-vehicle interaction system; if Then stop time progression and proceed for each interactive system. The instantaneous availability of the interactive system recorded at each test moment The instantaneous availability curve of the interactive system as a function of test time is used to calculate the average availability of the interactive system by taking a time-weighted average of the instantaneous availability curve at each test moment within the duration interval. ; Using the system test time as an index, output the instantaneous availability of each interactive system at each test time. and average availability .

[0019] As can be seen from the technical solutions provided by the embodiments of the present invention above, the embodiments of the present invention provide an autonomous vehicle-to-vehicle interaction system availability assessment method that can simultaneously consider the differences between the vehicle's basic operating capabilities and the interactive function implementation capabilities, and characterize the coupling relationship between the two.

[0020] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and will become apparent from the description or may be learned by practice of the invention. Attached Figure Description

[0021] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 A flowchart illustrating a usability evaluation method for an autonomous vehicle-to-vehicle interaction system provided in this embodiment of the invention; Figure 2 This is a schematic diagram of the structure of an autonomous vehicle-to-vehicle interaction system provided in an embodiment of the present invention; Figure 3 A schematic diagram of vehicle position during a testing process provided in an embodiment of the present invention; Figure 4 A diagram showing the instantaneous availability calculation results of a function provided in an embodiment of the present invention; Figure 5 This is a diagram showing the calculation results of vehicle element availability provided in an embodiment of the present invention; Figure 6 This is a diagram illustrating the instantaneous availability calculation results of an interactive subsystem provided in an embodiment of the present invention. Figure 7 This is a diagram showing the instantaneous availability calculation results of an interactive system provided in an embodiment of the present invention. Detailed Implementation

[0023] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0024] Those skilled in the art will understand that, unless specifically stated otherwise, the singular forms “a,” “an,” “the,” and “the” used herein may also include the plural forms. It should be further understood that the term “comprising” as used in this specification means the presence of the stated features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. It should be understood that when we say an element is “connected” or “coupled” to another element, it can be directly connected or coupled to the other element, or there may be intermediate elements. Furthermore, “connected” or “coupled” as used herein can include wireless connections or couplings. The term “and / or” as used herein includes any and all combinations of one or more of the associated listed items.

[0025] It will be understood by those skilled in the art that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. It should also be understood that terms such as those defined in general dictionaries should be understood to have the same meaning as in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless defined as herein.

[0026] To facilitate understanding of the embodiments of the present invention, the following will provide further explanation and description with reference to the accompanying drawings and several specific embodiments. These embodiments do not constitute a limitation on the embodiments of the present invention.

[0027] This invention provides a usability evaluation method for an autonomous vehicle-to-vehicle interaction system. The processing procedure of this method is as follows: Figure 1 As shown, the processing steps include the following: Step S1: Initialization. Obtain the basic input information required to calculate the availability of the autonomous vehicle-to-vehicle interaction system. This basic input information includes test range information, vehicle parameters, etc.

[0028] Step S2: Construct an autonomous vehicle-to-vehicle interaction system model based on the aforementioned basic input information; Step S3: Calculate the functional availability of the autonomous vehicle-to-vehicle interaction system based on the autonomous vehicle-to-vehicle interaction system model; Step S4: Calculate the availability of elements in the autonomous vehicle-to-vehicle interaction system based on the autonomous vehicle-to-vehicle interaction system model; Step S5: Calculate the availability of the autonomous vehicle-to-vehicle interaction system based on the functional availability and element availability of the autonomous vehicle-to-vehicle interaction system; Step S6: Determine if the test time ends; Step S7: Output the availability of the final autonomous vehicle-to-vehicle interaction system.

[0029] The above step S1 specifically includes: Figure 2 This is a schematic diagram of an autonomous vehicle-to-vehicle interaction system provided in an embodiment of the present invention. This step is used to determine the basic input information required for availability calculation. This basic input information includes test scope information, vehicle parameters, etc. The test scope information clarifies the space, time, and interaction categories covered by this test; the vehicle parameters describe the fault and repair characteristics of the vehicle's functional modules throughout their entire lifecycle, as well as the real-time operating status of the vehicle during the test.

[0030] The test scope information includes the test space scope, test time scope, test object scope, and interaction category scope. The test space scope represents the area being evaluated, such as road segments and intersections, and is denoted as... The test time range represents the start and end times of the test; its start time must be determined. and the end time The test duration is calculated as follows: Time step The test period is discretized. The test object range is the area within the test space that travels within the test time range. The collection of vehicles within is denoted as:

[0031] The interaction category is used to limit the interaction categories that need to be identified and evaluated in this test, and is denoted as:

[0032] In the formula, Indicates formation interaction, Indicates collaborative lane-changing interaction. This represents the interaction in a cross-junction game.

[0033] For each vehicle element in the candidate vehicle set The system collects vehicle geometric parameters, hardware and software reliability parameters, cumulative runtime at the start of the test, motion state parameters, and disturbance parameters. These represent the lengths of the vehicle body, These represent the vehicle body width, and the vehicle geometry parameters are used to calculate the actual net distance between vehicles, collision time, two-dimensional collision time, and the time relationship between the vehicle occupying the conflict area.

[0034] Set up vehicles At the start of the test The cumulative runtime is Each vehicle element contains four functional modules: a perception module ( ), transmission module ( ), Decision module ( ) and control module ( ). For vehicle number Each functional module has the following hardware parameters:

[0035] In the formula, Indicates vehicle The hardware parameters of the m-th module, This is due to the occasional hardware failure rate of this module. , These are the shape and scale parameters, respectively, of the module's hardware aging failure rate, which follows a Weibull distribution. The repair rate of this module's hardware. For vehicles The cumulative runtime.

[0036] Its software parameters are expressed as follows:

[0037] In the formula, Indicates vehicle The software parameters of the m-th module, , These represent the failure rate and recovery rate of the software subsystem of this module, respectively.

[0038] The motion state parameters are:

[0039] In the formula, Indicates vehicle At any moment The physical state, Indicates its longitudinal and transverse positions. Indicates its speed, This indicates its longitudinal and lateral acceleration. Indicates its heading angle, This indicates the lane number it is in.

[0040] In actual operation, the vehicle's motion state is still affected by factors such as position measurement error, speed measurement error, acceleration control error, system latency, and packet loss rate. To uniformly characterize these disturbances, the vehicle's disturbance parameter vector is defined as follows:

[0041] In the formula, , , , and These represent position measurement error, velocity measurement error, acceleration control error, system delay, and transmission packet loss rate, respectively.

[0042] Let the current assessment time And create an initial empty functional availability index library. .

[0043] Figure 3 This is a schematic diagram of the vehicle position during a test process provided in an embodiment of the present invention. See also... Figure 3 Test space range It includes a two-lane road in the same direction and an intersection downstream of it.

[0044] The test time range represents the start and end times of the test; its start time must be determined. and the end time The test duration is calculated as follows: Time step The test period is discretized, and the test object range is the vehicle traveling within the test space range within the test time range. The collection of vehicles within is denoted as:

[0045] The interaction category range is used to limit the interaction categories that need to be identified and evaluated in this test, denoted as:

[0046] In the formula, Indicates formation interaction, Indicates collaborative lane-changing interaction. This represents the interaction in a cross-junction game.

[0047] In this example, the cumulative total for each vehicle... runtime See Table 1; assuming all vehicles use the same modules and have identical hardware and software parameters, the differences in vehicle element availability are solely due to… The differences are due to the short duration of this test, and the aging effects that occurred during the testing phase can be disregarded.

[0048] Total running time of each vehicle Available test start time Down Replace the value with .

[0049] The hardware and software parameters for each module are shown in Tables 2 and 3, respectively.

[0050] Table 1. Cumulative Running Time of Each Vehicle

[0051] Table 2 Hardware parameters of each functional module

[0052] Table 3 Software parameters of each functional module

[0053] The motion state parameters of each vehicle during the operation are recorded as follows:

[0054] In the offline simulation of this example, the disturbance parameters of each vehicle... The same statistical family is used. The time delay and error parameters for the example vehicle are shown in Table 4. The distance perception error is truncated using a cumulative distribution probability of 0.45-0.55. Both speed perception error and acceleration control error are... Truncation, packet loss rate (PLR) and communication distance (Unit: m) A linear relationship exists:

[0055] Table 4. Delay and Error Parameters for Autonomous Vehicles

[0056] Let the current assessment time Establish an empty functional availability index library .

[0057] Step S2: Construct an autonomous vehicle-to-vehicle interaction system model.

[0058] This step identifies vehicle elements, interaction edges, interaction subsystems, and interaction systems based on vehicle operation data, and constructs an autonomous vehicle-to-vehicle interaction system model.

[0059] (1) Vehicle elements

[0060] Define a single vehicle as a vehicle element:

[0061] In the formula, Indicates the number is Vehicle elements.

[0062] Each vehicle element contains four functional modules: the perception module ( ), transmission module ( ), Decision module ( ) and control module ( The four modules are related as follows: the perception module is responsible for acquiring environmental information such as surrounding vehicles, lanes, and obstacles through onboard sensors; the transmission module is responsible for receiving status and intention information from other vehicles through vehicle-to-vehicle transmission. The perception and transmission modules together constitute the information acquisition layer, which provides redundancy in parallel, and any available channel can provide effective input for subsequent decisions. The decision module receives the output of the information acquisition layer and generates decision results such as longitudinal platooning, coordinated lane changing, or intersection game-theoretic passage order; the control module receives the output of the decision module and converts it into execution control quantities such as acceleration, braking, and steering. Therefore, the vehicle element internally forms a parallel relationship between the perception and transmission modules, and then forms a series relationship with the decision and control modules. As shown in Table 5, each core functional module is supported by both hardware and software, where the hardware provides the functional implementation carrier, and the software is responsible for the corresponding functional logic and algorithm implementation.

[0063] Table 5: Hardware and software composition of each module.

[0064] Therefore, vehicle element availability is used to characterize the probability that key modules of vehicle elements involved in interaction, such as perception, transmission, decision-making, and control, are in a working state at a given moment or time interval. It is a comprehensive measure of the basic support capability of vehicles to perform interactive tasks.

[0065] Define the interaction relationship between two vehicle elements as an interaction edge:

[0066] In the formula, Indicates the connection of vehicle elements and Interaction edge, Number its two ends.

[0067] (2) Interaction Subsystem

[0068] It will consist of several vehicle elements that have direct interaction relationships. and the interaction edges between them The minimum interactive structure is defined as an interactive subsystem:

[0069] In the formula, Indicates the first An interactive subsystem. To form the set of interaction edges for this subsystem, , Vehicle elements constituting this subsystem For the formation subsystem, It consists of the preceding and following vehicles participating in longitudinal platooning; for a cooperative lane-changing subsystem, it typically includes lane-changing vehicles. Vehicle in front in the original lane Vehicle in front of the target lane and the vehicle behind in the target lane The intersection game subsystem consists of vehicles within the conflict zone that have potential spatiotemporal conflicts. Indicates the category number of the interaction structure.

[0070] The same interaction category may have different variations in the number of participating vehicles or the topology. Taking cooperative lane changing as an example: when there is a vehicle ahead in the original lane, the participating vehicles are lane-changing vehicle M, the vehicle ahead in the original lane F0, the vehicle ahead in the target lane F1, and the vehicle behind in the target lane R1, a total of four vehicles. The interaction relationship between M and F0, F1, and R1 needs to be assessed simultaneously. When there is no vehicle ahead in the original lane, the vehicle ahead in the original lane F0 is not present, and the participating vehicles are only M, F1, and R1, a total of three vehicles. To distinguish different structural variations under the same interaction category, let the interaction structure category to be calibrated be numbered as follows:

[0071] In the formula, This represents the total number of interaction structure categories. Each interaction structure category corresponds to a specific interaction type, the number of participating vehicles, the composition of vehicle roles, and the set of interaction edges or conflict relationships.

[0072] For the interaction subsystem Simultaneously record its formation time With the moment of disappearance And obtain the duration of this interactive subsystem:

[0073] For any time Interactive subsystem The duration is: .

[0074] (3) Interactive system

[0075] An interaction system is defined as the largest interaction structure formed by merging several interconnected interaction subsystems that share vehicle elements. For any two interaction subsystems in the set of interaction subsystems... and If its vehicle element set satisfies Then it is considered that the interaction subsystem and There is a connection between them, and they are grouped into the same interactive system.

[0076] If multiple interactive subsystems are linked together through shared vehicle elements, then even if some of these subsystems do not directly share vehicles, they are still grouped into the same interactive system. For example, interactive subsystems and Shared vehicle elements, and The elements of shared vehicles, , and Together they form an interactive system.

[0077] The merging process continues according to the above rules until no two different interaction systems share vehicle elements. For a single interaction subsystem that does not share vehicle elements with other interaction subsystems, it is treated as a separate interaction system.

[0078] Let the first The set of interactive subsystems contained in an interactive system is: Then the interactive system can be represented as:

[0079] In the formula, For the first A collection of vehicle elements for an interactive system. For the first The set of interaction edges for each interactive system is obtained by taking the union of the set of vehicle elements and the set of interaction edges for each interactive subsystem contained in the system:

[0080] This results in a set of interactive systems:

[0081] In the formula, This indicates the number of identified interactive systems. Since vehicle positions and inter-vehicle interactions change over time, the composition of the interactive subsystems and interactive systems can be dynamically updated at each discrete moment according to the aforementioned rules.

[0082] For the An interactive system Let its formation time and disappearance time be respectively:

[0083] Its duration is:

[0084] Step S3: Calculate the functional availability of the autonomous vehicle-to-vehicle interaction system based on the autonomous vehicle-to-vehicle interaction system model.

[0085] This invention introduces an availability index to evaluate the ability of an autonomous vehicle-to-vehicle (V2V) interaction system to continuously maintain its intended functions. This step involves constructing a functional availability index library for online operational evaluation of autonomous V2V interaction systems. For unknown interaction structure categories encountered during online evaluation, their functional availability models are obtained and stored through offline simulation testing. For interaction structure categories that have already been tested and stored in the index library, the corresponding functional availability results are directly retrieved, thus avoiding redundant simulation calculations. Since there may be structural variations within the same interaction category with different numbers of participating vehicles, vehicle role composition, interaction edge sets, or conflict relationship sets, they are numbered according to the interaction structure category. As a unified identifier for functional availability testing, sample aggregation, and index storage, step S3 includes the following sub-steps S3.1 to S3.6.

[0086] S3.1, Interaction Structure Category Determination

[0087] For each interaction structure category determined in step S2 at the current time Compare it with the functionality availability index. Comparison: like If it is determined to be a newly added interactive structure category, proceed to step S3.2 to perform offline testing on this category; if If it is determined to be an old interaction structure category that has already been tested, it will be called directly. China and Israel For indexing Proceed to step S4.

[0088] This decision ensures that offline tests for the same interaction structure category are performed at most once throughout the entire evaluation process, avoiding repeated simulations.

[0089] S3.2, Offline Test Initialization

[0090] For the newly added interactive structure category Set the total number of tests Duration of a single test And let the test number .

[0091] To ensure that the interaction process in offline testing conforms to the characteristics of real traffic flow, the traffic flow distribution is first calibrated based on the motion state parameters collected in step S1: summarizing the vehicle set. velocity at each moment Longitudinal acceleration and the inter-vehicle time distance calculated from the longitudinal position and speed of adjacent vehicles. The optimal distribution was determined using both the maximum likelihood method and the AIC criterion, resulting in the vehicle speed distribution. Acceleration distribution Workshop time interval distribution Its specific form is determined by measured data.

[0092] Each time a test scenario is generated, the initial speed of each vehicle is determined by... Sampling, the initial distance between adjacent vehicles is The acceleration of each vehicle before coordination during operation is determined by sampling and conversion based on vehicle speed. Sampling is used to drive the motion model.

[0093] To avoid the repeated calculation of hardware and software failures in the vehicle's perception, transmission, decision-making, and control modules during steps S3 and S4, the offline simulation test in this step assumes that all vehicle elements involved in the interaction are in a hardware and software usable state. The offline simulation test does not inject hardware and software failures of the vehicle modules; it only injects position measurement errors, speed measurement errors, acceleration control errors, communication latency, transmission packet loss, algorithm performance fluctuations, and traffic environmental disturbances. This is used to evaluate whether the interactive function can be correctly implemented under the condition that the vehicle's basic operational capabilities are guaranteed. Based on this, it is numbered according to the interaction structure category. Each motion model was constructed separately and offline testing was conducted independently. Each interaction structure category... This corresponds to the defined interaction category, number of participating vehicles, vehicle role composition, and set of interaction edges or conflict relationships. Specifically, it includes: (1) For platooning interaction, a suitable following model is adopted so that the vehicles in front and behind always maintain longitudinal platooning and following throughout the entire test period, and the functional state duration of platooning interaction is extracted throughout the process. (2) For the collaborative lane-changing interaction, the collaborative lane-changing model is adopted so that the lane-changing vehicle repeatedly initiates and completes the lane change during the test period. Only the duration from triggering to completion of each lane-changing interaction is extracted and included in the functional state duration. The non-lane-changing driving period within the lane-changing interval is not included. (3) For the intersection game interaction, the intersection game model is adopted, so that vehicles with spatiotemporal conflicts repeatedly enter the conflict zone to play the game during the test period, and only the duration of each game interaction is extracted and included in the functional state duration.

[0094] S3.3, Duration Statistics

[0095] This step determines the interactive function status of interactive subsystems with different interactive structures time-by-time based on the motion state parameters and disturbance parameters of each vehicle during the k-th offline test, and counts samples of continuous available time and continuous unavailable time, providing a data basis for subsequent fitting of available time distribution and unavailable time distribution.

[0096] (1) Functional availability determination

[0097] Interactive functions are determined by multiple criteria. Based on whether the criteria rely on future state predictions, they are categorized into predictive and real-time execution criteria. Predictive criteria include Time to Collision (TTC), 2D-TTC, and Predicted Post-Collision Intrusion (PPET), whose values ​​are extrapolated from the current vehicle state and are affected by measurement errors, system latency, and transmission packet loss, requiring probabilistic judgment. Real-time execution criteria include speed, acceleration, and trigger timeliness, whose values ​​are directly given by the actual operating state at the current moment.

[0098] To uniformly characterize the information delay caused by communication latency and packet loss, for vehicles Define equivalent delay:

[0099] In the formula, For communication cycle, It is used to uniformly describe the information delay caused by the combined effects of communication latency and packet loss.

[0100] The motion state vector for disturbance propagation is composed of the vehicle's two-dimensional position, two-dimensional velocity, and two-dimensional acceleration:

[0101] Let the first Class interaction structure in the first The first offline test The predictive judgment indicators are: The perturbation error term is then calculated using first-order sensitivity propagation:

[0102] In the formula, Indicates the first Class interaction structure in the first The set of vehicle elements in this test; An effective disturbance to the vehicle's motion state.

[0103] The effective disturbance to the vehicle's motion state is composed of both the state error and the state lag error caused by the equivalent time delay:

[0104] In the formula, Indicates vehicle The corresponding vehicle state error vector; This indicates the state lag error caused by the equivalent time delay.

[0105] Thus, the predictive judgment index values ​​under the effect of disturbance are obtained:

[0106] In the formula, For the effect of disturbance, the next The values ​​of predictive judgment indicators are determined. For real-time execution indicators, the impact of disturbances is already included in the actual operating state of the vehicle; therefore, the realized values ​​are directly used for deterministic judgment, and a separate disturbance propagation model is not established. .

[0107] To avoid frequent functional state switching caused by fluctuations in the judgment index around the threshold, an asymmetric hysteresis judgment mechanism is introduced. The safety margin of the index relative to the threshold after perturbation is defined as:

[0108] In the formula, the lower limit indicator refers to an indicator whose larger value indicates greater safety (e.g., ...). , , Upper limit indicators are those that are safer the smaller the value (such as speed and trigger timeliness), while range indicators are those that must be kept between the upper and lower limits (such as longitudinal acceleration and lateral acceleration). , , These are the corresponding thresholds.

[0109] Let the first The threshold scale for each judgment indicator is Define the unavailability trigger margin and recovery availability margin as follows:

[0110] In the formula, For the trigger margin of switching from an available state to an unavailable state, This refers to the recovery margin when recovering from an unavailable state to an available state. and These represent the unavailable trigger ratio and the restored available ratio, respectively.

[0111] Based on this, define the first The outbound probability and regression probability of each indicator:

[0112] In the formula, the probability of exceeding the boundary is... With regression probability For predictive indicators, the results are obtained from the statistical analysis of disturbance samples; for real-time execution indicators, since their values ​​are fixed, both indicators take the value of 0 or 1.

[0113] Based on a comprehensive evaluation of all criteria and after confirmation by the hysteresis loop mechanism, the first... Class interaction structure in the first The next test, time The functional state is determined by the following formula:

[0114] In the formula, This indicates whether the interactive function is available at any given time, with 1 indicating availability and 0 indicating inavailability. To determine the confidence level for probability. To allow for out-of-bounds risk levels. The first case represents a switch from available to unavailable, i.e., the probability of an indicator going out of bounds reaches the allowable risk level; the second case represents a recovery from unavailable to available, i.e., the regression probability of all indicators is not lower than the confidence level; the third case represents all indicators being within the hysteresis interval, maintaining the functional state of the previous moment.

[0115] The criteria for determining the interactive function and its threshold depend on the type of interactive subsystem. The following are the criteria for determining the functional availability of three types of interactive subsystems: platooning, lane changing, and intersection game.

[0116] The formation function is available only if all three conditions listed in Table 6 are met simultaneously. For the car behind With the car in front The collision time between them. Acceleration constraints ensure that the vehicle's longitudinal control does not exceed the physical and comfort boundaries; speed constraints ensure that the speed does not exceed the current lane speed limit. Among them, TTC is a predictive indicator, which uses probabilistic hysteresis determination; velocity and acceleration are real-time execution indicators, which use deterministic threshold hysteresis determination based on realized values.

[0117] Table 6. Conditions for Determining the Availability of Formation Function

[0118] The lane-changing function is available only if all the conditions listed in Table 7 are met simultaneously. This is the moment when the system first detects that the lane-changing conditions are met. When the system actually performs the lane change, the total response time of the decision-making and control modules can be expressed as: ; , These are the longitudinal and lateral accelerations of vehicles changing lanes. 2D-TTC is a predictive indicator, using probabilistic hysteresis for determination; trigger timeliness, speed, and acceleration are real-time execution indicators, using deterministic threshold hysteresis based on realized values. The determination criteria are shown in Table 7. The risk of longitudinal conflict only exists when M has not completely left its original lane. Assessment is only conducted before crossing the line; F1 and R1 are always at risk of collision with M throughout the lane change process and must be assessed at all times.

[0119] Table 7 Conditions for Determining the Availability of Lane Changing Function

[0120] The crossroads game function can only be available if all the conditions listed in Table 8 are met simultaneously. This is the moment when collision avoidance control is first detected. When the system actually initiates collision avoidance control, the response time can be expressed as: Among them, trigger timeliness is a real-time execution indicator, which adopts a deterministic threshold hysteresis judgment based on the realized value, requiring the system to initiate collision avoidance response in a timely manner after detecting a conflict risk; PPET is a predictive indicator, which adopts a probabilistic hysteresis judgment, and determines whether the collision avoidance result is safe by evaluating the time interval between the expected passage of the two vehicles through the conflict area.

[0121] Table 8. Conditions for Determining the Available States of the Crossroads Game Function

[0122] (2) Statistics on available and unavailable time

[0123] Complete the first Interaction structure in this test After determining the time-by-time functional state, its binary functional state sequence is obtained. , Indicates the interaction number Class interaction structure in the first The set of valid continuous discrete moments in this test.

[0124] The binary functional state sequence is segmented into consecutive segments with the same value along the time axis. If the state value remains continuously at 1, the time interval is defined as a continuous available process; if the state value remains continuously at 0, the time interval is defined as a continuous unavailable process.

[0125] Let the first The start and end times of each continuously available process are respectively and The corresponding continuous available time is:

[0126] In the formula, Indicates the first Class interaction structure in the first The first offline test obtained A continuous available duration sample.

[0127] Let the first The start and end times of the consecutive unavailable processes are respectively and The corresponding continuous unavailability duration is:

[0128] In the formula, Indicates the first Class interaction structure in the first The first offline test obtained A sample of consecutive unavailable durations.

[0129] Therefore, the first Class interaction structure in the first The available time sample set and the unavailable time sample set in this offline test are represented as follows:

[0130] In the formula, Indicates the first Class interaction structure in the first Number of samples of continuous availability time in the offline test. This indicates the number of samples with continuous unavailability. When a valid interaction period ends, if the functional state has not yet switched at the end, the dwell time of that end period is not fully observed and is treated as a right-censored sample, which is not included in the subsequent distribution parameter estimation.

[0131] Step S3.4, Determining the Termination of the Test

[0132] If the test sequence number satisfies Then the test sequence number will be updated. Then return to step S3.3 and continue executing step S3. If the updated test sequence number is not satisfied, then it is determined that all K tests have been completed, the test loop is terminated, and the process proceeds to step S3.5 to calculate the functional availability.

[0133] In this embodiment, k=10, and a total of 10 offline tests were executed.

[0134] S3.5 Calculate the functional availability of the autonomous vehicle-to-vehicle interaction system

[0135] Since the judgment criteria and conflict relationships of different structural variants under the same interaction category may be different, this invention is based on the interaction structure category numbering. Summarize the duration samples. For all... In this test, they belonged to the same interaction structure category. The interactive subsystem has the following available duration sample set and unavailable duration sample set:

[0136] In the formula, and They represent the first The set of samples of continuous available duration and continuous unavailable duration for interactive structures.

[0137] For sample set and Distribution fitting was performed separately, the maximum likelihood method was used to estimate the candidate distribution parameters, and the optimal distribution model was determined according to the AIC criterion. , The first The available duration distribution function and unavailable duration distribution function of the interactive structure. , These are their probability density functions.

[0138] Table 9 shows the fitting results of the duration distribution for different interaction structure categories.

[0139] Table 9. Fitting results of duration distribution for different interaction structure categories

[0140] From the available / unavailable duration probability density function in Table 9 , According to its Laplace transform , Substitution

[0141] Then, the Abate-Whitt-Euler numerical inversion algorithm is used to obtain... The results are shown in Figure 4.

[0142] according to Therefore, there is

[0143] (1) Calculation of instantaneous functional availability

[0144] set up This indicates that all vehicle elements comprising this interactive subsystem are at time [time]. All events are in an available state, meaning the vehicles involved in the interaction possess the basic operational capabilities required to perform the interaction task; assuming... This indicates an event in which the corresponding interactive function can be implemented normally, provided that the vehicle elements are available. The instantaneous availability of a class-interaction structure is defined as: the conditional probability that the interaction structure will achieve the expected interaction function, assuming all participating vehicle elements are available.

[0145] When the initial state is an available state satisfy:

[0146] Performing a Laplace transform, we get:

[0147] In the formula, , , They are respectively , , The Laplace transform of .

[0148] Numerical inversion of the above equation yields the first... Instantaneous availability of functionality in interactive structures:

[0149] In the formula, This represents the inverse Laplace transform operator. In practical calculations, the Abate-WhittEuler numerical inversion algorithm can be used to solve it.

[0150] (2) Calculation of functional steady-state availability

[0151] According to the final value theorem, the steady-state availability of the function is:

[0152] In the formula, and The first The average duration of continuous availability and the average duration of continuous unavailability for interactive structures.

[0153] For each interaction structure category Summarize its instantaneous availability function and steady-state availability:

[0154] S3.6, Write Functionality Availability Index Repository

[0155] The obtained functional availability results are categorized by interaction structure type. For indexing, Stored as an entry in the functionality availability index library .

[0156]

[0157] The results of the instantaneous availability calculation are as follows: Figure 4 As shown. Figure 4 The instantaneous availability curve of functions over time serves as a functional availability index. The reusable results are used in subsequent steps at any current evaluation time. The following is based on the duration of the interaction subsystem. Make the call.

[0158] Step S4: Calculation of the availability of elements in the autonomous vehicle-to-vehicle interaction system

[0159] As the test progresses, the vehicle's cumulative runtime is updated synchronously with the current evaluation time. If the test time... and test start time All are measured in seconds, while the hardware / software failure rate, repair rate, and Weibull scale parameters are measured in hours. Therefore, the vehicle... The cumulative runtime at the current evaluation time t is:

[0160] In the formula, The unit is hours. Used to convert additional runtime during the test process from seconds to hours.

[0161] For the set of vehicle elements involved in the evaluation Each vehicle element The availability of its elements is calculated using the following formula:

[0162] In the formula, Representing vehicle elements Availability; , These represent the availability of the vehicle's perception module and transmission module, respectively. The two are combined in parallel to reflect the redundancy and backup relationship of the vehicle's information acquisition methods (active perception and inter-vehicle communication). , These represent the availability of the vehicle's decision-making module and control module, respectively, and are combined with the aforementioned parallel results in series.

[0163] The availability of each module is obtained by multiplying its hardware availability by its software availability:

[0164] In the formula, , They represent the first The hardware availability and software availability of each module, m=1,2,3,4 correspond to the sensing, transmission, decision-making and control modules respectively.

[0165] (1) Hardware availability calculation

[0166] The failure rate of module hardware consists of two parts: incidental failure rate and aging failure rate. Hardware lifespan exhibits significant wear and aging characteristics; therefore, the Weibull model is used to describe its aging failure rate.

[0167] In the formula, Indicates the first Each module hardware at time failure rate, For occasional failure rate, , These represent the shape and scale parameters of the Weibull distribution, respectively. (Module) The hardware repair rate , For module The average hardware repair time is... .

[0168] Assuming the hardware is in a usable state at the factory, according to the Markov equation, the module is obtained. The analytical expression for instantaneous hardware availability is:

[0169] (2) Software availability calculation

[0170] Similarly, the software error recovery process employs an exponential process, with a failure rate... With recovery rate The availability of the following module software:

[0171] Step S5: Availability Calculation of Autonomous Vehicle-to-Vehicle Interaction System

[0172] This step is used to calculate the availability of the interactive subsystem and the instantaneous availability of the interactive system based on the availability of interactive functions, the availability of vehicle elements, and the shared vehicle relationships in the interactive system.

[0173] For each interactive subsystem, the conditional constraints of vehicle element status on the realization of interactive functions are considered. The event that all vehicle elements are available is taken as a prerequisite for the realization of interactive functions. Based on Bayesian probability relations, a coupling model between vehicle element availability and conditional function availability is established to obtain the actual availability of the interactive subsystem at the current evaluation time.

[0174] For each interactive subsystem Considering the conditional constraints of vehicle element states on the implementation of interactive functions, the available states of vehicle elements and the available states of interactive functions are probabilistically coupled. Let... This indicates that all vehicle elements comprising this interactive subsystem are at time [time]. All events are in an available state, meaning the vehicles involved in the interaction possess the basic operational capabilities required to perform the interaction task; assuming... This indicates an event in which the corresponding interactive function can be implemented normally when the vehicle elements are available.

[0175] Since the state of vehicle elements is a fundamental condition for the implementation of interactive functions, there is a conditional dependency between the two, rather than an independent relationship. Therefore, the actual available events of the interactive subsystem can be represented as a joint event of available events for vehicle elements and available events for interactive functions:

[0176] Based on Bayesian probability relationships, the interactive subsystem at the current evaluation time... The availability can be expressed as:

[0177] For a set containing vehicle elements Interactive subsystem Let its corresponding interaction structure category number be Then at time Functionality availability is Vehicle elements Availability is denoted as Under the condition that the states of each vehicle element are independent of each other, we have:

[0178] In the formula, Representing the interactive subsystem At any moment The availability status; Indicates the first The functional availability status of the class interaction structure; Representing vehicle elements Cumulative runtime The available status is as follows; Representing the interactive subsystem The cumulative runtime of each vehicle element is the set of values ​​in the formula. In this formula, the availability of interactive functions is determined by the duration of the interactive subsystem, and the availability of vehicle elements is determined by the cumulative runtime of each vehicle itself.

[0179] The above model shows that the availability of interactive functions is not independent of the absolute availability probability of the vehicle's operating state, but rather a conditional probability of function realization after factors such as perception errors, communication delays, control deviations, and environmental disturbances have affected the vehicle's basic operating conditions. Through Bayesian probability relationships, the combined impact of the vehicle's basic capabilities and interactive function realization capabilities on the actual availability state of the interactive subsystem can be simultaneously characterized.

[0180] In actual evaluation, when the interaction subsystem is within a finite duration, the instantaneous availability of the corresponding interaction structure category is called using the duration of the interaction subsystem as the independent variable; when the evaluation object is a long-term operation scenario or a steady-state operation scenario, the steady-state availability of the corresponding interaction structure category is called and coupled with the availability of the participating vehicle elements to obtain the availability of the interaction subsystem.

[0181] For the An interactive system Vehicle elements At any moment Below the interactive system The number of times it appears in is:

[0182] In the formula, As an indicator function, when vehicle elements Belongs to the interaction subsystem vehicle element set The value is 1 if the condition is met, and 0 otherwise.

[0183] Interactive System The instantaneous availability is determined by the current set of effective interactive subsystems, the duration of each interactive subsystem, and the cumulative running time of participating vehicles, and is expressed as: In the formula, Representing an interactive system At any moment The instantaneous availability is calculated as follows: the numerator is the product of the availability of the currently effective interactive subsystems; the denominator is used to eliminate the portion of shared vehicle elements that is repeatedly counted in multiple interactive subsystems. For vehicle elements belonging to only one interactive subsystem, .

[0184] Step S6: Determine if the test time ends.

[0185] Complete the current moment After step S5, compare the current evaluation time. With the end time of the test .like If the test time has not yet reached the end time, then let If the test time has reached its termination time, proceed to step S2; otherwise, proceed to step S7.

[0186] After completing S5 at t=0s, the current evaluation time is... With the end time of the test , Therefore, .like If the test time has not yet reached the end time, then let Then return to step S2 and re-execute S2 at the new time t=0.1s: It persists, and its formation time remains unchanged. It has been a long time The structure category remains b=1,2,3. Proceed to S3.1 for judgment: At this point... Therefore All already exist (Old structure category), direct call The corresponding Proceed to S4. S4 calculates the availability of elements at each time step; S5... Call Coupled with element availability S6 Judgment ,make Return to S2. Repeat this process. All time periods use the old structure category and are directly reused. , obtain every moment until The test time has reached the termination time, exit the loop and proceed to step S7.

[0187] Therefore, the whole Offline testing during the evaluation process is only conducted during the offline testing phase. Execute once (to obtain) The remaining 50 test moments are all reused to avoid repeated simulations.

[0188] During the above cycle, each time the current evaluation time is completed... Steps S4 and S5 below record the vehicle element availability and the instantaneous availability of the interaction subsystem at that moment, respectively. This is done as the evaluation time progresses according to... The recorded vehicle element availability sequence, interaction subsystem instantaneous availability sequence, and interaction system instantaneous availability sequence are then processed sequentially to form... Figure 5 , Figure 6 The results of the time variation are shown.

[0189] Step S7: Output the availability of the autonomous vehicle-to-vehicle interaction system.

[0190] After the outermost loop terminates, for each interactive system Instantaneous availability values ​​recorded at each test moment Its duration varies with system testing time Curve of instantaneous availability change over time , and according to

[0191] After the test period ends, for each interactive system... The instantaneous availability of the interactive system recorded at each discrete test moment is summarized to form a sequence of instantaneous availability of the interactive system as a function of test time:

[0192] In this embodiment, the interactive system is described separately. and interactive system Output its test period The instantaneous availability sequence corresponding to each discrete test moment:

[0193] The instantaneous availability sequence constitutes the interactive system. Instantaneous availability versus time curve within the test time range like Figure 7 As shown.

[0194] And calculate the interaction system separately. and interactive system Average availability during the test period :

[0195] in for In its duration The set of test times within the system. The final result is based on the system test time. Using the index, output the instantaneous availability of the entire interactive system at each test time. and average availability .

[0196] The calculation results are shown in Table 10.

[0197] Table 10 Average Availability of Interactive Systems

[0198] Finally, the instantaneous availability is output for each interactive system. ( Figure 7 ) and average availability .

[0199] In summary, compared with the prior art, the present invention has the following beneficial effects: (1) This invention divides the autonomous vehicle-to-vehicle interaction system into three levels: vehicle elements, interaction subsystems, and the interaction system. The availability of vehicle elements and functional availability are calculated separately, and coupled based on Bayes' theorem. The availability of the interaction system is obtained through deduplication of shared vehicles and layer-by-layer aggregation. Thus, it can avoid the evaluation bias caused by judging the availability of the system solely based on the integrity of the vehicle's hardware and software, and adapt to the dynamic, open, and random operating characteristics of the vehicle-to-vehicle interaction system.

[0200] (2) This invention incorporates position measurement error, velocity measurement error, acceleration control error, system delay, and transmission packet loss rate into the functional state determination, establishes the transmission relationship between disturbance parameters and functional determination indicators, and combines probabilistic determination and asymmetric hysteresis determination mechanisms to determine the interactive functional state. Thus, it can quantify the impact of multi-source disturbances on the availability of interactive functions and reduce frequent switching of functional states caused by fluctuations in determination indicators near the threshold.

[0201] (3) This invention conducts offline simulation tests according to interaction categories and interaction structure categories, and establishes a functional availability index library. For interaction structure categories that have been tested, the corresponding results can be directly called, thereby avoiding repeated simulation calculations and improving the computational efficiency and result reusability of availability assessment.

[0202] Those skilled in the art will understand that the accompanying drawings are merely schematic diagrams of one embodiment, and the modules or processes shown in the drawings are not necessarily essential for implementing the present invention.

[0203] As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that the present invention can be implemented by means of software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in various embodiments or some parts of the embodiments of the present invention.

[0204] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, for apparatus or system embodiments, since they are basically similar to method embodiments, the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments. The apparatus and system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0205] The above description is merely a preferred 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 usability evaluation method for an autonomous vehicle-to-vehicle interaction system, characterized in that, include: Obtain the basic input information required to calculate the availability of the autonomous vehicle-to-vehicle interaction system. This basic input information includes test range information and vehicle parameters. An autonomous vehicle-to-vehicle interaction system model is constructed based on the aforementioned basic input information; Calculate the functional availability of the autonomous vehicle-to-vehicle interaction system based on the autonomous vehicle-to-vehicle interaction system model; Calculate the availability of elements in an autonomous vehicle-to-vehicle interaction system based on an autonomous vehicle-to-vehicle interaction system model; The availability of the autonomous vehicle-to-vehicle interaction system is calculated based on the functional availability and element availability of the autonomous vehicle-to-vehicle interaction system.

2. The method according to claim 1, characterized in that, The basic input information required to obtain the availability of the autonomous vehicle-to-vehicle interaction system includes test range information and vehicle parameters, including: To obtain the basic input information required to calculate the availability of the autonomous vehicle-to-vehicle interaction system, the basic input information includes test range information and vehicle parameters, wherein the test range information includes the test space range. Test time range Scope of test subjects and scope of interaction categories Where F, L, and I represent formation interaction, coordinated lane-changing interaction, and intersection game interaction, respectively; the vehicle parameters include the hardware parameters, software parameters, motion state parameters, and disturbance parameters of each vehicle element, and the disturbance parameter vector These represent the position measurement error, velocity measurement error, acceleration control error, system delay, and transmission packet loss rate, respectively; and let the current evaluation time be... Establish an initial empty functional availability index library. .

3. The method according to claim 2, characterized in that, The construction of an autonomous vehicle-to-vehicle interaction system model based on the aforementioned basic input information includes: Based on vehicle operation data, vehicle elements, interaction edges, interaction subsystems, and interaction systems are identified to construct an autonomous vehicle-to-vehicle interaction system model. A single vehicle is defined as an element comprising a perception module, a transmission module, a decision-making module, and a control module. The perception and transmission modules are connected in parallel, and then connected in series with the decision-making and control modules. The interaction relationship between two vehicle elements is defined as an interaction edge. The smallest interaction structure consisting of several vehicle elements with direct interaction relationships and the interaction edges between them is defined as an interaction subsystem. According to the interaction subsystem The interaction category, number of participating vehicles, vehicle role composition, and set of interaction edges or conflict relationships are used to match the corresponding interaction structure category number for this interaction subsystem. The largest interactive structure, formed by merging several interconnected interactive subsystems that share vehicle elements, is defined as an interactive system. ; For any two interaction subsystems in the set of interaction subsystems and If the vehicle element sets of the two sets intersect, that is... If the two are related, they are classified into the same interaction system; if multiple interaction subsystems are continuously related by sharing vehicle elements, then even if some of the interaction subsystems do not directly share vehicle elements, these multiple interaction subsystems will be classified into the same interaction system; for a single interaction subsystem that does not share vehicle elements with other interaction subsystems, it will be treated as a separate interaction system.

4. The method according to claim 3, characterized in that, The calculation of the functional availability of the autonomous vehicle-to-vehicle interaction system based on the autonomous vehicle-to-vehicle interaction system model includes: For each interaction structure category identified at the current evaluation time t Combine it with the aforementioned functional availability index library Compare: If the interaction structure category Not in the index If a new interaction structure category is identified, offline simulation testing is conducted. This offline simulation testing includes: constructing a motion model according to the corresponding interaction category and setting the total number of tests and the duration of each test; determining the interaction function status at each time step in multiple offline tests based on the judgment index; statistically obtaining the continuous available duration sample set and the continuous unavailable duration sample set for this interaction structure category; performing distribution fitting on the two sample sets respectively; estimating candidate distribution parameters using the maximum likelihood method; and determining the probability density functions of the available duration distribution and the unavailable duration distribution based on the Akaike Information Criterion. , ; the probability density function , After performing a Laplace transform, input the equations established based on the alternating update process. And perform numerical inversion. We obtain the instantaneous availability of the functionality of this interaction structure category; using the final value theorem, we obtain the steady-state availability of the functionality of this interaction structure category. and The first The average duration of continuous availability and the average duration of continuous unavailability of the interaction structure are assigned to the category number of that interaction structure. The index is stored in the aforementioned functional availability index library. ; If the interaction structure category Already in the index If it is found to be a previously tested old interaction structure category, the index library will be called directly. The interactive structure is categorized by number. Instantaneous availability of index storage functionality Offline simulation testing will no longer be performed repeatedly.

5. The method according to claim 4, characterized in that, The calculation of the functional availability of the autonomous vehicle-to-vehicle interaction system based on the autonomous vehicle-to-vehicle interaction system model also includes: The interaction structure categories include: platooning interaction, cooperative lane-changing interaction, and intersection game interaction. The vehicle's interaction functions are defined by predictive and real-time execution indicators. Predictive indicators include time to collision (TTC), 2D-TTC, and estimated post-collision intrusion time (PPET). Real-time execution indicators include speed, acceleration, and trigger timeliness. The interaction function states of interactive subsystems with different interaction structures are determined time-by-time. The values ​​of the predictive indicators are extrapolated from the current vehicle state, and the disturbance error term is calculated using a first-order sensitivity propagation method, with determination performed in a probabilistic sense. The values ​​of the real-time execution indicators are directly given by the actual operating state at the current moment, and deterministic determination based on the realized value is adopted. An asymmetric hysteresis determination mechanism is introduced, and the out-of-bounds probability is calculated based on the safety margin of the indicator after disturbance relative to the threshold. With regression probability When the probability of an indicator going out of bounds reaches a preset risk level, the interactive function is switched from available to unavailable. When the regression probability of all indicators is not lower than the preset confidence level, the interactive function is restored from unavailable to available. Otherwise, the function status of the previous moment is maintained.

6. The method according to claim 4, characterized in that, The offline simulation test uses corresponding motion models for different interaction categories: for platooning interaction, a car-following model is used to keep the vehicles in front and behind in longitudinal platooning throughout the test period, and the functional state duration of platooning interaction is extracted throughout the process. For collaborative lane-changing interactions, a collaborative lane-changing model is used to enable lane-changing vehicles to repeatedly initiate and complete lane changes during the test period. Only the duration from triggering to completion of each lane-changing interaction is extracted and included in the functional state duration. For intersection game interaction, an intersection game model is used to make vehicles with spatiotemporal conflicts repeatedly enter the conflict zone to play the game during the test period. Only the duration of each game interaction is extracted and included in the functional state duration.

7. The method according to claim 5, characterized in that, The calculation of the element availability of the autonomous vehicle-to-vehicle interaction system based on the autonomous vehicle-to-vehicle interaction system model includes: For each vehicle element participating in the evaluation Based on the hardware failure rate of its sensing module, transmission module, decision-making module, and control module. Hardware repair rate Software failure rate and software recovery rate Calculate the hardware availability of each module separately. Software availability The availability of each module is obtained by multiplying the results. Then, the availability of the perception module With the availability of transmission modules Combined in parallel and with decision module availability and control module availability By combining them in a series manner, the vehicle element at the current evaluation time can be obtained. Corresponding cumulative runtime Vehicle element availability: ； In the formula, Representing vehicle elements Availability; , These represent the availability of the vehicle's perception module and transmission module, respectively.

8. The method according to claim 3, characterized in that, When the interaction subsystem is within a finite duration, the instantaneous availability of the function of its corresponding interaction structure category is called with the duration of the interaction subsystem as the independent variable; when the evaluation object is a long-term operation scenario or a steady-state operation scenario, the steady-state availability of the function of its corresponding interaction structure category is called.

9. The method according to claim 3, characterized in that, The calculation of the availability of the autonomous vehicle-to-vehicle interaction system based on the functional availability and element availability of the autonomous vehicle-to-vehicle interaction system includes: For each interactive subsystem, determine the functional availability of the interactive structure category to which that subsystem belongs. Vehicle element availability of each vehicle element included in this interactive subsystem By coupling using Bayes' theorem, the interaction subsystem at the current evaluation time can be obtained. Availability : For each interactive system, the availability of each interactive subsystem within that system is aggregated layer by layer. Furthermore, shared vehicle elements that appear repeatedly across multiple interactive subsystems are ranked according to their frequency of occurrence. After deduplication correction, the instantaneous availability of the interactive system at the current evaluation time t is obtained: 。 10. The method according to claim 7, characterized in that, The method further includes: Compare the current assessment time With the end time of the test :like Then let Then, repeat the above process of constructing the autonomous vehicle-to-vehicle interaction system model, calculating the functional availability, element availability, and overall availability of the autonomous vehicle-to-vehicle interaction system; if Then stop time progression and proceed for each interactive system. The instantaneous availability of the interactive system recorded at each test moment The instantaneous availability curve of the interactive system as a function of test time is used to calculate the average availability of the interactive system by taking a time-weighted average of the instantaneous availability curve at each test moment within the duration interval. ; Using the system test time as an index, output the instantaneous availability of each interactive system at each test time. and average availability .

Images