WO2020233428A1 - 一种电动车辆的安全失效风险预测方法以及电动车辆 - Google Patents

一种电动车辆的安全失效风险预测方法以及电动车辆 Download PDF

Info

Publication number
WO2020233428A1
WO2020233428A1 PCT/CN2020/089420 CN2020089420W WO2020233428A1 WO 2020233428 A1 WO2020233428 A1 WO 2020233428A1 CN 2020089420 W CN2020089420 W CN 2020089420W WO 2020233428 A1 WO2020233428 A1 WO 2020233428A1
Authority
WO
WIPO (PCT)
Prior art keywords
safety
failure
risk
electric vehicle
safety failure
Prior art date
Application number
PCT/CN2020/089420
Other languages
English (en)
French (fr)
Inventor
张伟
Original Assignee
深圳市德塔防爆电动汽车有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳市德塔防爆电动汽车有限公司 filed Critical 深圳市德塔防爆电动汽车有限公司
Publication of WO2020233428A1 publication Critical patent/WO2020233428A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N5/00—Computing arrangements using knowledge-based models
    • G06N5/04—Inference or reasoning models
    • G06N5/046—Forward inferencing; Production systems
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0635—Risk analysis of enterprise or organisation activities

Definitions

  • the invention relates to a transportation tool, and more specifically, to a method for predicting the safety failure risk of an electric vehicle and an electric vehicle.
  • electric vehicles generally have electrical systems as high as hundreds of volts, which exceeds the safe voltage range of DC. If reasonable design and protection are not carried out, high voltage safety problems such as electric shocks may be caused.
  • electric vehicles include multiple components such as steering systems, braking systems, and safety control systems, and each component includes multiple components. The failure or malfunction of any component may cause loss of control or failure of the entire vehicle, thereby causing the driver or passengers to encounter danger.
  • the safety tree of electric vehicles is a systematic method to comprehensively solve the safety problems of electric vehicles. It is composed of safety failure top events, safety failure intermediate events, safety failure basic failure bottom events, related logic and data to establish a related logic system, and through the whole vehicle safety Demand analysis and vehicle system construction event model establishment tree diagram is a description of the logical relationship between different levels of events in the vehicle, such as the brake system, steering system, body parts and other subsystems or components for graphical representation and qualitative description.
  • the safety tree can accurately express the causal relationship and logic between safety failure top events and underlying basic failure events (process defects, external factors, etc.).
  • the vehicle safety status assessment is based on the real-time quantitative description of the vehicle safety status based on the safety tree. As the working hours of vehicles increase, the safety of vehicles gradually declines. This law will also be reflected in the increase in the risk of vehicle failure. Therefore, it is necessary to analyze the law of vehicle failure risk over time, and predict the future failure risk, so as to provide the necessary quantitative information basis for the safe operation and maintenance of vehicles.
  • the technical problem to be solved by the present invention is to provide a safe failure risk prediction method for electric vehicles in view of the above-mentioned defects of the prior art, which can analyze the law of vehicle failure risk over time, and predict the future failure risk. Provide the necessary quantitative information foundation for the safe operation and maintenance of vehicles.
  • the technical solution adopted by the present invention to solve its technical problems is to construct a method for predicting the safety failure risk of electric vehicles, including:
  • the step S2 further includes:
  • a Kalman filter algorithm is used to calculate the predicted safety failure risk value of the electric vehicle.
  • the step S21 further includes:
  • the step S22 further includes:
  • SC t-1 represents the first system safety factor of the electric vehicle at the first moment
  • ⁇ represents the rate of change of the failure risk of the electric vehicle
  • B( ⁇ t) represents the standard Brownian motion
  • ⁇ represents the electric vehicle The diffusion coefficient of the failure risk.
  • the step S23 further includes:
  • the Kalman filter algorithm is used to calculate the safety failure risk value formula of the electric vehicle in the interval (t, t+P step ⁇ t) based on the following formula
  • the step S2 further includes:
  • N represents the number of total security failures
  • n i represents the number of security failure bottom-level events corresponding to the i-th security failure.
  • the step S1 further includes:
  • Another technical solution adopted by the present invention to solve its technical problems is to construct a computer-readable storage medium on which a computer program is stored, and when the program is executed by a processor, the method for predicting the safety failure risk of electric vehicles is realized .
  • Another technical solution adopted by the present invention to solve its technical problem is to construct an electric vehicle, including a processor, and a computer program stored in the processor.
  • the program is executed by the processor, the electric vehicle Safety failure risk prediction method.
  • the method for predicting the safety failure risk of electric vehicles, the computer-readable storage medium, and the electric vehicle according to the present invention can analyze the law of vehicle failure risk over time, and predict the future failure risk, which is the safety of the vehicle. Operation and maintenance provide the necessary quantitative information foundation.
  • FIG. 1 is a flowchart of a first embodiment of a method for predicting a safety failure risk of an electric vehicle according to a preferred embodiment of the present invention
  • FIG. 2 is a schematic diagram of the classification of the complete vehicle safety failure data of the method for predicting the safety failure risk of an electric vehicle according to a preferred embodiment of the present invention
  • 3a-3c are schematic diagrams of part of the safety tree of the method for predicting the safety failure risk of an electric vehicle according to a preferred embodiment of the present invention.
  • FIG. 4 is a flowchart of the steps of calculating and predicting the safety failure risk value of the method for predicting the safety failure risk of an electric vehicle of the present invention
  • Fig. 5 is a flowchart of the steps of calculating the failure risk degree of the entire vehicle in the method for predicting the safety failure risk of the present invention.
  • the present invention relates to a safety failure risk prediction method for electric vehicles, including: S1, constructing a safety tree, the safety tree including multiple safety failure bottom events, safety failure intermediate events, safety failure top events, and the safety failure bottom events , The logical causality and safety importance between the safety failure intermediate event and the safety failure top-level event; S2, according to the safety tree, predict the safety failure risk of the electric vehicle.
  • the method for predicting the safety failure risk of electric vehicles, the computer-readable storage medium, and the electric vehicle according to the present invention can analyze the law of vehicle failure risk over time, and predict the future failure risk, which is the safety of the vehicle. Operation and maintenance provide the necessary quantitative information foundation.
  • Fig. 1 is a flowchart of a first embodiment of a method for predicting a safety failure risk of an electric vehicle according to a preferred embodiment of the present invention.
  • a safety tree is constructed, the safety tree includes multiple safety failure bottom-level events, safety failure intermediate events, safety failure top-level events, and the safety failure bottom-level events and the safety failure intermediate events , The logical causality between the top-level events of the safety failure and the degree of safety importance.
  • the data in the vehicle controller, the safety controller and the driving recorder of the electric vehicle are first transmitted to the platform database through the CAN bus. Then, the vehicle safety failure data of the electric vehicle is obtained from the data.
  • the vehicle safety failure data is mapped into different safety event groups, and the probability of each safety event group accounting for all safety failures is calculated.
  • the vehicle safety failure data can be mapped into multiple subsystems or components such as brake systems, steering systems, body parts, etc., so that the vehicle safety failure data can be included in different groups according to the principle of mapping classification. Among them, the probability of each security event group accounting for all security failures is calculated.
  • Fig. 2 is a schematic diagram of the classification of the entire vehicle safety failure data of the method for constructing a safety tree of an electric vehicle according to a preferred embodiment of the present invention.
  • the vehicle safety failure data can be mapped to structural safety events, electrical safety events, functional logic safety events, collision safety events, thermal safety events, and explosion-proof events.
  • the above-mentioned inductive analysis process can use various methods known in the art, or use known methods to calculate the probability of each safety event group accounting for all safety failures, or use the respective measurement and collection empirical data of electric vehicle manufacturers.
  • a joint analysis method is used to classify the safety failure data of the entire vehicle in each safety event group to construct a safety tree.
  • the new joint analysis method is used for safety tree modeling. According to the actual situation of safety failure, one or more suitable analysis methods can be selected, so as to avoid using a certain model construction method alone.
  • the disadvantages of adaptation can be analyzed by applying method advantages in the actual application process to effectively simplify the selection process.
  • any security tree known in the art can be used, and any known security tree in the art can be used.
  • FIGS. 3a-3c are schematic diagrams of a partial safety tree constructed by the method for constructing a safety tree for an electric vehicle according to a preferred embodiment of the present invention. The following further describes the method for constructing a safety tree for an electric vehicle of the present invention based on FIGS. 3a-3b.
  • three safety failure intermediate events can be subdivided under structural safety events, namely, braking safety incidents, driving safety incidents, and steering safety incidents.
  • step S2 the safety failure risk of the electric vehicle is predicted according to the safety tree.
  • the risk of failure of the entire vehicle can be calculated based on the safety tree.
  • the system safety factor of the electric vehicle may be calculated according to the safety tree, and then the predicted safety failure risk value of the electric vehicle may be calculated according to the system safety system.
  • the method for predicting the safety failure risk of electric vehicles, the computer-readable storage medium, and the electric vehicle according to the present invention can analyze the law of vehicle failure risk over time, and predict the future failure risk, which is the safety of the vehicle. Operation and maintenance provide the necessary quantitative information foundation.
  • step S1 the first system safety factor of the electric vehicle at the first moment is calculated.
  • step S2 the first system safety factor of the electric vehicle at the first moment is calculated.
  • step S3 the standardized frequency of the occurrence of the safety failure intermediate event is counted within the set first time interval.
  • a service brake failure, a parking brake failure, and an abnormal hydraulic pressure can be regarded as an intermediate event of a safety failure.
  • the standardized frequency of each occurrence within a year can be counted.
  • the standardized frequency of occurrence of the safety failure intermediate event is converted to standard working conditions to obtain the occurrence frequency of the standard safety failure intermediate event.
  • the logic and probability of the effect of the underlying security failure event on the security failure intermediate event is the frequency of the same security failure intermediate event (referring to different security failure intermediate events caused by the same security failure underlying event), and
  • the impact probability of the underlying events of safety failure is weighted and combined to obtain the weighted standardized frequency of occurrence of intermediate safety failure events. Also taking the embodiment shown in FIG. 3b as an example, for the bottom layer of safety failure event of brake spring damage, it simultaneously corresponds to two homogenous safety failure intermediate events of service brake failure and parking brake failure.
  • the low-level safety failure event of abnormal brake pressure corresponds to the two intermediate safety failure events of the same source, the service brake failure and the parking brake failure.
  • the impact probability of brake spring damage on service brake failure and parking brake failure is 0.3% and 0.4%, respectively.
  • the risk level Li characterizes the safety-related consequences caused by the failure event (or the i-th safety failure intermediate event).
  • the specific value of i can be defined with reference to the security tree shown in FIGS. 3a-3c.
  • Risk level is a quantitative evaluation of the severity of consequences, which is usually defined quantitatively by experts based on business characteristics. There have been various risk levels for different electric vehicles in the field.
  • the weighted standardized frequency of the occurrence of safety failure intermediate events is converted to the working conditions to be calculated through the statistical regression analysis method, and the standardized frequency occurrence of all safety failure intermediate events is summed to obtain the The standardized frequency (and failure probability, unit: times/accumulated working hours (mileage)) of electrical system failure events in a given time interval under given working conditions.
  • Standardization of the frequency of occurrence of safety failures means that the frequency of occurrence of safety failures obtained by statistics under different environmental parameters is converted to a uniform specified environmental parameter to obtain an equivalent frequency that can be used for global analysis. According to the occurrence mechanism of the safety failure intermediate event, the working conditions affecting the number of occurrences of the safety failure intermediate event are analyzed.
  • the safety controller Based on the data recorded in the data in the vehicle controller, the safety controller and the driving recorder of the electric vehicle, the above analysis and judgment can be completed.
  • the risk weight q i corresponding to the standard safety failure intermediate event is calculated based on the occurrence frequency of the standard safety failure intermediate event.
  • the risk weight q i can be used to describe the parameter of the influence degree of the failure risk on the frequency of occurrence of the standard safety failure intermediate event.
  • the risk weight is the ratio of the standard safety failure intermediate event occurrence frequency to the maximum tolerable frequency; when the actual standard safety failure intermediate event occurrence frequency When it is greater than or equal to the highest tolerance frequency, the risk weight is equal to 1.
  • the maximum tolerance frequency is an important parameter used to normalize the risk weight of the intermediate event of a safety failure, which can be set by those skilled in the art based on experience. Through long-term observation and testing of electric vehicles, the highest tolerance frequency can be obtained. There have been various regulations in the field regarding the highest tolerance frequency of different safety failure intermediate events for different electric vehicles.
  • the risk level Li characterizes the safety-related consequences caused by the failure event (or the i-th safety failure intermediate event).
  • the specific value of i can be defined with reference to the security tree shown in FIGS. 3a-3c.
  • Risk level is a quantitative evaluation of the severity of consequences, which is usually defined quantitatively by experts based on business characteristics. There have been various risk levels for different electric vehicles in the field.
  • the first system safety factor is calculated based on the risk degree corresponding to all safety failure intermediate events of the electric vehicle
  • N represents the number of total security failures
  • n i represents the number of security failure bottom-level events corresponding to the i-th security failure.
  • the specific values of N and i can be defined with reference to the security tree shown in FIGS. 3a-3c.
  • step S2 the Wiener process is used to obtain the second system safety factor of the electric vehicle at the second moment based on Bayesian inference.
  • Wiener process is an important independent factor.
  • the incremental process is also called the Brownian motion process.
  • SC t-1 represents the first system safety factor of the electric vehicle at the first moment
  • ⁇ represents the rate of change of the failure risk of the electric vehicle
  • B( ⁇ t) represents the standard Brownian motion
  • ⁇ represents the electric vehicle The diffusion coefficient of the failure risk.
  • SCt represents the second system safety factor of the electric vehicle at the second time.
  • a Kalman filter algorithm is used to calculate the predicted safety failure risk value of the electric vehicle.
  • the expected maximum likelihood algorithm is used to estimate the change rate ⁇ of the failure risk of the electric vehicle and the spread of the failure risk of the electric vehicle based on the first system safety factor and the second system safety factor The coefficient ⁇ .
  • the expected maximum likelihood algorithm is one of the commonly used parameter estimation methods in statistics. It is usually known that a random sample satisfies a certain probability distribution, but the specific parameters are not clear. The parameter estimation is through several experiments , Observe the result, use the result to deduce the approximate value of the parameter.
  • the rate of change ⁇ of the failure risk of the electric vehicle and the diffusion coefficient ⁇ of the failure risk of the electric vehicle can be estimated.
  • the Kalman filter algorithm is used to calculate the safety failure risk value formula of the electric vehicle in the interval (t, t+P step ⁇ t) based on the following formula Among them, t represents the current moment, and P step ⁇ t] represents the number of failure risk prediction steps.
  • Kalman filtering is an algorithm that uses linear system state equations to perform optimal estimation of system state through system input and output observation data.
  • P step ⁇ t] can be obtained by averaging or weighting the time period during which the electric vehicle is most likely to fail based on experience obtained during the long-term operation and maintenance of the electric vehicle.
  • Those skilled in the art can select appropriate failure risk prediction step values according to actual needs.
  • Vehicle failure risk prediction refers to the prediction of the future based on the complete and real-time status of all safety failures of the entire vehicle. It is an important parameter of the entire vehicle that indicates the safety of the entire vehicle, thus affecting the product design, production and maintenance process. Optimization provides theoretical guidance. Implementing the method for predicting the safety failure risk of electric vehicles of the present invention can analyze the law of vehicle failure risk over time, and predict the future failure risk degree, providing the necessary quantitative information basis for the safe operation and maintenance of the vehicle .
  • Fig. 5 is a flowchart of the steps of calculating the failure risk degree of the entire vehicle in the method for predicting the safety failure risk of the present invention.
  • step S1 in the set first time interval, the standardized frequency of occurrence of the safety failure intermediate event is counted.
  • a service brake failure, a parking brake failure, and an abnormal hydraulic pressure can be regarded as an intermediate event of a safety failure.
  • the standardized frequency of each occurrence within a year can be counted.
  • step S2 the standardized frequency of occurrence of the safety failure intermediate event is converted to a standard working condition to obtain the occurrence frequency of the standard safety failure intermediate event.
  • the logic and probability of the effect of the security failure bottom-level event on the security failure intermediate event is the frequency of the same security failure intermediate event (referring to different security failure intermediate events caused by the same security failure bottom event) according to the safety
  • the influence probability of the bottom-level failure event is weighted and combined to obtain the standardized frequency of occurrence of the weighted safety failure intermediate event. Also taking the embodiment shown in FIG. 3b as an example, for the bottom layer of safety failure event of brake spring damage, it simultaneously corresponds to the two homogenous safety failure intermediate events of service brake failure and parking brake failure.
  • the low-level safety failure event of abnormal brake pressure corresponds to the two intermediate safety failure events of the same source, the service brake failure and the parking brake failure.
  • the impact probability of brake spring damage on service brake failure and parking brake failure is 0.3% and 0.4%, respectively.
  • the standardized frequency of occurrence of safety failure intermediate events is converted to the working conditions to be calculated through statistical regression analysis method, and the standardized frequency of occurrence of all safety failure intermediate events is summed to obtain a given working condition and a given time
  • Standardization of the frequency of occurrence of safety failures means that the frequency of occurrence of safety failures obtained by statistics under different environmental parameters is converted to a uniform specified environmental parameter to obtain an equivalent frequency that can be used for global analysis. According to the occurrence mechanism of the safety failure intermediate event, the working conditions affecting the number of occurrences of the safety failure intermediate event are analyzed.
  • the safety controller Based on the data recorded in the data in the vehicle controller, the safety controller and the driving recorder of the electric vehicle, the above analysis and judgment can be completed.
  • the risk weight q i corresponding to the standard safety failure intermediate event is calculated based on the occurrence frequency of the standard safety failure intermediate event.
  • the risk weight q i can be used to describe the parameter of the influence degree of the failure risk on the frequency of occurrence of the standard safety failure intermediate event.
  • the risk weight is the ratio of the standard safety failure intermediate event occurrence frequency to the maximum tolerable frequency; when the actual standard safety failure intermediate event occurrence frequency When it is greater than or equal to the highest tolerance frequency, the risk weight is equal to 1.
  • the maximum tolerance frequency is an important parameter used to normalize the risk weight of the intermediate event of a safety failure, which can be set by those skilled in the art based on experience. Through long-term observation and testing of electric vehicles, the highest tolerance frequency can be obtained. There have been various regulations in the field regarding the highest tolerance frequency of different safety failure intermediate events for different electric vehicles.
  • the risk level Li characterizes the safety-related consequences caused by the failure event (or the i-th safety failure intermediate event).
  • the specific value of i can be defined with reference to the security tree shown in FIGS. 3a-3c.
  • Risk level is a quantitative evaluation of the severity of consequences, which is usually defined quantitatively by experts based on business characteristics. There have been various risk levels for different electric vehicles in the field.
  • step S5 the vehicle failure risk is calculated according to the following formula based on the safety tree: Where N represents the number of total security failures, and n i represents the number of security failure bottom-level events corresponding to the i-th security failure. Also taking the embodiment shown in FIG. 3b as an example, for the bottom layer of safety failure event of brake spring damage, it simultaneously corresponds to the two homogenous safety failure intermediate events of service brake failure and parking brake failure. In the same way, the low-level safety failure event of abnormal brake pressure corresponds to the two intermediate safety failure events of the same source, the service brake failure and the parking brake failure. Therefore, the vehicle failure risk can be calculated according to the safety tree.
  • Implementing the method for predicting the safety failure risk of electric vehicles of the present invention can analyze the law of vehicle failure risk over time, and predict the future failure risk degree, providing the necessary quantitative information basis for the safe operation and maintenance of the vehicle .
  • the present invention can be implemented by hardware, software or a combination of software and hardware.
  • the present invention can be implemented in a centralized manner in at least one computer system, or implemented in a decentralized manner by different parts distributed in several interconnected computer systems. Any computer system or other equipment that can implement the method of the present invention is applicable.
  • the combination of commonly used software and hardware can be a general computer system with a computer program installed, and the computer system is controlled by installing and executing the program to make it run according to the method of the present invention.
  • the present invention can also be implemented by a computer program product.
  • the program contains all the features capable of implementing the method of the present invention. When it is installed in a computer system, the method of the present invention can be implemented.
  • the computer program in this document refers to any expression of a set of instructions that can be written in any programming language, code, or symbol.
  • the instruction set enables the system to have information processing capabilities to directly implement specific functions, or to perform After one or two steps, a specific function is realized: a) conversion into other languages, codes or symbols; b) reproduction in a different format.
  • the present invention also relates to a computer-readable storage medium on which a computer program is stored, and when the program is executed by a processor, the method for predicting the safety failure risk of an electric vehicle is realized.
  • the present invention also relates to an electric vehicle, including a processor, and a computer program stored in the processor, and the program is executed by the processor to realize the safety failure risk prediction method of the electric vehicle.
  • the method for predicting the safety failure risk of electric vehicles, the computer-readable storage medium, and the electric vehicle according to the present invention can analyze the law of vehicle failure risk over time, and predict the future failure risk, which is the safety of the vehicle. Operation and maintenance provide the necessary quantitative information foundation.

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Strategic Management (AREA)
  • Theoretical Computer Science (AREA)
  • Economics (AREA)
  • Physics & Mathematics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • General Physics & Mathematics (AREA)
  • Marketing (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • Game Theory and Decision Science (AREA)
  • General Business, Economics & Management (AREA)
  • Development Economics (AREA)
  • Artificial Intelligence (AREA)
  • Educational Administration (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Electric Propulsion And Braking For Vehicles (AREA)
  • Traffic Control Systems (AREA)

Abstract

一种电动车辆的安全失效风险预测方法，包括：构建安全树，该安全树包括多个安全失效底层事件、安全失效中间事件、安全失效顶层事件以及该安全失效底层事件、该安全失效中间事件、该安全失效顶层事件之间的逻辑因果关系和安全重要程度（S1）；根据该安全树，预测该电动车辆的安全失效风险（S2）。该电动车辆的安全失效风险预测方法、计算机可读存储介质以及电动车辆，可以分析车辆失效风险随时间变化的规律，并对未来的失效风险度进行预测，为车辆的安全运维提供必要的定量化信息基础。

Description

一种电动车辆的安全失效风险预测方法以及电动车辆 技术领域
本发明涉及运输工具，更具体地说，涉及一种电动车辆的安全失效风险预测方法以及电动车辆。
背景技术
随着世界经济的快速发展和对环保意识的重视，汽车的普及率越来越高，同时对汽车尾气排放要求也越来越高，节能、安全、无污染的电动车辆是未来的发展趋势。然而，电动车辆一般有高达上百伏的电气系统，这就超过了直流的安全电压范围，如不进行合理的设计与防护，将可能带来人员电击等高压安全问题。此外，电动车辆包括诸如转向系统、制动系统、安全控制系统等多个组成部门，每个组成部分又包括多个组成部件。任何部件的失效或者故障都可能造成整个车辆的失控或者故障，从而导致驾驶者或者乘客遭遇危险。
而电动车辆的安全树是全面解决电动车辆安全问题的系统方法，是由通过安全失效顶事件、安全失效中间事件、安全失效基础故障底事件、相关逻辑和数据建立相关逻辑体系，通过整车安全需求分析和整车系统构建事件模型建立树状图，是对车辆不同层次事件之间逻辑关系的描述，针对例如制动系统、转向系统、车身零部件等多个子系统或部件进行图形表征和定性描述。安全树能够准确表达安全失效顶事件和底层基础故障事件(工艺缺陷、外部因素等)之间因果关系和逻辑。
整车安全状态评估是基于安全树对整车安全情况的实时定量描述。随着车 辆工作时间的增加，车辆的安全度是逐渐下降的，这一规律也将表现在车辆失效风险的增加上。因此，有必要分析车辆失效风险随时间变化的规律，并对未来的失效风险度进行预测，为车辆的安全运维提供必要的定量化信息基础。
发明内容
本发明要解决的技术问题在于，针对现有技术的上述缺陷，提供一种电动车辆的安全失效风险预测方法，可以分析车辆失效风险随时间变化的规律，并对未来的失效风险度进行预测，为车辆的安全运维提供必要的定量化信息基础。
本发明解决其技术问题所采用的技术方案是：构造一种电动车辆的安全失效风险预测方法，包括：
S1、构建安全树，所述安全树包括多个安全失效底层事件、安全失效中间事件、安全失效顶层事件以及所述安全失效底层事件、所述安全失效中间事件、所述安全失效顶层事件之间的逻辑因果关系和安全重要程度；
S2、根据所述安全树，预测所述电动车辆的安全失效风险。
在本发明所述的电动车辆的安全失效风险预测方法中，所述步骤S2进一步包括：
S21、计算第一时刻的所述电动车辆的第一系统安全系数；
S22、采用维纳过程基于贝叶斯推理获得第二时刻的所述电动车辆的第二系统安全系数；
S23、基于所述第一系统安全系数、所述第二系统安全系数，采用卡尔曼滤波算法计算所述电动车辆的预测安全失效风险值。
在本发明所述的电动车辆的安全失效风险预测方法中，所述步骤S21进一步包括：
S211、在设定第一时间区间内，统计所述安全失效中间事件发生的标准化频次；
S212、将所述安全失效中间事件发生的标准化频次换算到标准工作条件下以获得标准安全失效中间事件发生频次；
S213、基于所述标准安全失效中间事件发生频次计算所述标准安全失效中间事件对应的风险权值q i；
S214、基于所述标准安全失效中间事件对应的风险权值和风险等级Li计算所述标准安全失效中间事件对应的风险度R i＝q iL i，其中L i＝0，...，10；
S215、基于电动车辆的全部安全失效中间事件对应的风险度计算所述第一系统安全系数
Figure PCTCN2020089420-appb-000001
其中N表示全部安全失效的数量，n i表示第i安全失效对应的安全失效底层事件的数量。
在本发明所述的电动车辆的安全失效风险预测方法中，所述步骤S22进一步包括：
基于以下公式计算采用维纳过程基于贝叶斯推理获得第二时刻的所述电动车辆的第二系统安全系数SCt：SC t＝SC t-1+ηΔt+σB(Δt)
其中SC t-1表示第一时刻的所述电动车辆的第一系统安全系数；η表示所述电动车辆的失效风险的变化率，B(Δt)表示标准布朗运动，σ是表示所述电动车辆的失效风险的扩散系数。
在本发明所述的电动车辆的安全失效风险预测方法中，所述步骤S23进一步包括：
S231、采用期望极大似然算法基于所述第一系统安全系数和所述第二系统安全系数估算所述电动车辆的失效风险的变化率η和所述电动车辆的失效风险的扩散系数σ；
S232、采用卡尔曼滤波算法基于以下公式计算区间(t,t+P stepΔt]所述电动车辆的安全失效风险值公式
Figure PCTCN2020089420-appb-000002
在本发明所述的电动车辆的安全失效风险预测方法中，所述步骤S2进一步包括：
S2a、在设定第一时间区间内，统计所述安全失效中间事件发生的标准化频次；
S2b、将所述安全失效中间事件发生的标准化频次换算到标准工作条件下以获得标准安全失效中间事件发生频次；
S2c、基于所述标准安全失效中间事件发生频次计算所述标准安全失效中间事件对应的风险权值q i；
S2d、基于所述标准安全失效中间事件对应的风险权值和风险等级Li计算所述标准安全失效中间事件对应的风险度R i＝q iL i，其中L i＝0，...，10；
S2e、基于所述安全树根据以下公式计算整车失效风险度：
Figure PCTCN2020089420-appb-000003
其中N表示全部安全失效的数量，n i表示第i安全失效对应的安全失效底层事件的数量。
在本发明所述的电动车辆的安全失效风险预测方法中，所述步骤S1进一步包括：
S11.采集电动车辆的整车安全失效数据；
S12.将所述整车安全失效数据映射归类到不同的安全事件组别中，并分别统计各个安全事件组别频次数据；
S13.采用联合分析方法对各个安全事件组别中的所述整车安全失效数据进行分类构建安全树。
本发明解决其技术问题采用的另一技术方案是，构造一种计算机可读存储 介质，其上存储有计算机程序，所述程序被处理器执行时实现所述的电动车辆的安全失效风险预测方法。
本发明解决其技术问题采用的再一技术方案是，构造一种电动车辆，包括处理器，存储在所述处理器中的计算机程序，所述程序被处理器执行时实现所述的电动车辆的安全失效风险预测方法。
实施本发明的所述的电动车辆的安全失效风险预测方法、计算机可读存储介质以及电动车辆，可以分析车辆失效风险随时间变化的规律，并对未来的失效风险度进行预测，为车辆的安全运维提供必要的定量化信息基础。
附图说明
下面将结合附图及实施例对本发明作进一步说明，附图中：
图1是本发明的优选实施例的电动车辆的安全失效风险预测方法的第一实施例的流程图；
图2是本发明的优选实施例的电动车辆的安全失效风险预测方法的整车安全失效数据的归类示意图；
图3a-3c是本发明的优选实施例的电动车辆的安全失效风险预测方法的部分安全树的示意图；
图4是本发明的电动车辆的安全失效风险预测方法的计算预测安全失效风险值的步骤的流程图；
图5是本发明安全失效风险预测方法的计算整车失效风险度的步骤的流程图。
具体实施方式
为了使本发明的目的、技术方案及优点更加清楚明白，以下结合附图及实施例，对本发明进行进一步详细说明。应当理解，此处所描述的具体实施例仅仅用以解释本发明，并不用于限定本发明。
本发明涉及一种电动车辆的安全失效风险预测方法，包括：S1、构建安全树，所述安全树包括多个安全失效底层事件、安全失效中间事件、安全失效顶层事件以及所述安全失效底层事件、所述安全失效中间事件、所述安全失效顶层事件之间的逻辑因果关系和安全重要程度；S2、根据所述安全树，预测所述电动车辆的安全失效风险。实施本发明的所述的电动车辆的安全失效风险预测方法、计算机可读存储介质以及电动车辆，可以分析车辆失效风险随时间变化的规律，并对未来的失效风险度进行预测，为车辆的安全运维提供必要的定量化信息基础。
图1是本发明的优选实施例的电动车辆的安全失效风险预测方法的第一实施例的流程图。如图1所示，在步骤S1中，构建安全树，所述安全树包括多个安全失效底层事件、安全失效中间事件、安全失效顶层事件以及所述安全失效底层事件、所述安全失效中间事件、所述安全失效顶层事件之间的逻辑因果关系和安全重要程度。
在本发明的一个优选实施例中，首先通过CAN总线将所述电动车辆的整车控制器、安全控制器和行车记录仪中的数据传送到平台数据库。然后从所述数据中获取所述电动车辆的整车安全失效数据。将所述整车安全失效数据映射归类到不同的安全事件组别中，并计算各个安全事件组别占全部安全失效的概率。例如，可以将整车安全失效数据映射归类制动系统、转向系统、车身零部件等多个子系统或部件，这样就将所述整车安全失效数据按照映射归类的原理计入不同的组别当中，并且统计各个安全事件组别占全部安全失效的概率。
图2是本发明的优选实施例的电动车辆的安全树构建方法的整车安全失效数据的归类示意图。如图2所示，在本发明的一个优选实施例中，可以将所述整车安全失效数据分别映射到结构安全事件、电气安全事件、功能逻辑安全事件、碰撞安全事件、热安全事件、防爆安全事件、运营维修安全事件、环境安全事件和全生命周期安全事件。上述归纳分析过程可以采用本领域中已知的各种方法，也可以采用已知方法计算各个安全事件组别占全部安全失效的概率，还可以采用电动车辆制造商各自的测量和采集经验数据。最后采用联合分析方法对各个安全事件组别中的所述整车安全失效数据进行分类构建安全树。在本发明的优选实施例中，应用新型联合分析方法进行安全树建模，可以根据安全失效实际情况，选择合适的一种或者多种分析方法，以避免单独使用某种模型构建方法数据状况不适应的弊端，在实际运用过程中能应用方法优势针对性分析，有效简化选择过程。在本发明，可以采用本领域中已知的任何安全树，以及采用本领域中的任何已知安全树。在本发明的进一步的优选实施例中，在本公司申请的在先专利申请CN2019103168721“一种电动车辆的安全树构建方法以及电动车辆”中，公开了一种优选的安全树的构建方法，在此结合引用，以作参考。当然，在本发明的其他优选实施例中，还可以采用其他的安全树构建方法。
图3a-3c是本发明的优选实施例的电动车辆的安全树构建方法的构建的部分安全树的示意图。下面基于图3a-3b对本发明的电动车辆的安全树构建方法进一步说明如下。如图3a-3c所示，结构安全事件下面可以细分三个安全失效中间事件，即制动安全事件，行驶传动安全事件，和转向安全事件，我们可以分别对各个事件构建安全树。
在步骤S2中，根据所述安全树，预测所述电动车辆的安全失效风险。在 本发明的优选实施例中，可以根据所述安全树，计算整车失效风险度。在本发明的另一优选实施例中，可以根据所述安全树，计算电动车辆的系统安全系数，然后根据所述系统安全系统计算所述电动车辆的预测安全失效风险值。
实施本发明的所述的电动车辆的安全失效风险预测方法、计算机可读存储介质以及电动车辆，可以分析车辆失效风险随时间变化的规律，并对未来的失效风险度进行预测，为车辆的安全运维提供必要的定量化信息基础。
图4是本发明的电动车辆的安全失效风险预测方法的计算预测安全失效风险值的步骤的流程图。在步骤S1中，计算第一时刻的所述电动车辆的第一系统安全系数。在本发明的优选实施例中，首先在第一步中，在设定第一时间区间内，统计所述安全失效中间事件发生的标准化频次。如图3b所示，例如可以将行车制动故障、驻车制动故障，液压压力异常，分别作为一个安全失效中间事件，统计例如在一年之内，其分别发生的标准化频次。在第二步中，将所述安全失效中间事件发生的标准化频次换算到标准工作条件下以获得标准安全失效中间事件发生频次。在本步骤中，安全失效底层事件对安全失效中间事件的作用逻辑和影响概率，将同源安全失效中间事件(指不同的安全失效中间事件由同一个安全失效底层事件导致产生)的频次，按安全失效底层事件影响概率加权合并，得到加权后的安全失效中间事件发生的标准化频次。同样以图3b所示的实施例为例，对于制动弹簧损坏这一安全失效底层事件，其同时对应行车制动故障和驻车制动故障这两个同源安全失效中间事件。同理，制动压力异常这一安全失效底层事件同时对应行车制动故障和驻车制动故障这两个同源安全失效中间事件。而如图3b所示，制动弹簧损坏对于行车制动故障和驻车制动故障的影响概率分别是0.3％和0.4％。通过根据影响概率加权合并行车制动故障和驻车制动故障，就可以获得标准安全失效中间事件发生频次。 在本发明的进一步的优选实施例中，可以根据所述安全失效中间事件的风险等级获得其加权频次。例如在已知时间区间(t c，t c+Δt]内，安全失效中间事件i对应的标准安全失效中间事件S i(i＝1，...N)，则其相应的加权频次为
Figure PCTCN2020089420-appb-000004
其中，L i＝0，...，10。其中L i＝0，...，10。在本发明中，风险等级Li表征了失效事件(或第i个安全失效中间事件)所引起的与安全性相关的后果。在本发明的优选实施例中，i的具体取值，定义可以参照图3a-3c中示出的安全树。风险等级是对后果严重度的定量评价，通常由专家根据业务特点进行量化定义。本领域中已经有各种对于不同电动车辆的风险等级划分。
在第三步中，将安全失效中间事件发生的加权后的标准化频次，通过统计回归分析方法，换算到待计算的工作条件下，并对所有安全失效中间事件发生的标准化频次求和，得到给定工作条件下、给定时间区间内电气系统失效事件发生的标准化频次(及失效概率，单位是：次/累积工作时长(里程))。安全失效发生频次标准化是指将不同环境参数下统计得到的安全失效发生频次，换算至统一的规定环境参数下，得到可用于全局分析的等价发生频次。依据所述安全失效中间事件的发生机理，分析影响所述安全失效中间事件的事件发生数量的工作条件。比如可以根据道路状况，温度湿度，负载重量等等工作条件，分析影响所述安全失效中间事件的事件发生数量。在湿度大的情况下，发生制动安全事件、转向安全事件和行驶传动安全事件的数量可能较大。在道路状况差的情况下，发生行驶传动安全事件的数量可能较大。基于所述电动车辆的整车控制器、安全控制器和行车记录仪中的数据中记录的数据，可以完成上述分析判断。
随后在第四步中，基于所述标准安全失效中间事件发生频次计算所述标准安全失效中间事件对应的风险权值q i。风险权值q i可以用于描述所述标准安全 失效中间事件发生频次对失效风险影响程度的参数。当实际的所述标准安全失效中间事件发生频次小于最高容忍频次时，风险权值为所述标准安全失效中间事件发生频次同最高容忍频次的比值；当实际的所述标准安全失效中间事件发生频次大于或等于最高容忍频次时，风险权值等于1。最高容忍频次是用于安全失效中间事件归一化风险权值的重要参数，其可以由本领域技术人员根据经验设置。通过对电动车辆长期的观察，测试，可以获得该最高容忍频次。本领域中已经有各种对于不同电动车辆的不同安全失效中间事件的最高容忍频次的规定。
在第五步中，基于所述标准安全失效中间事件对应的风险权值和风险等级Li计算所述标准安全失效中间事件对应的风险度R i＝q iL i，其中L i＝0，...，10。在本发明中，风险等级Li表征了失效事件(或第i个安全失效中间事件)所引起的与安全性相关的后果。在本发明的优选实施例中，i的具体取值，定义可以参照图3a-3c中示出的安全树。风险等级是对后果严重度的定量评价，通常由专家根据业务特点进行量化定义。本领域中已经有各种对于不同电动车辆的风险等级划分。
在第六步中基于电动车辆的全部安全失效中间事件对应的风险度计算所述第一系统安全系数
Figure PCTCN2020089420-appb-000005
其中N表示全部安全失效的数量，n i表示第i安全失效对应的安全失效底层事件的数量。在本发明的优选实施例中，具体的N和i的具体取值，定义可以参照图3a-3c中示出的安全树。
如图4所示，在步骤S2中，采用维纳过程基于贝叶斯推理获得第二时刻的所述电动车辆的第二系统安全系数，本领域技术人员知悉，维纳过程是一个重要的独立增量过程，也称作布朗运动过程。其定义为，若一个随机过程{X(t),t>＝0}满足：⑴X(t)是独立增量过程；⑵任意 s,t>0,X(s+t)-X(s)～N(0,σ^2*t)，即X(s+t)-X(s)是期望为0，方差为σ^2*t的正态分布；⑶X(t)关于t是连续函数。则称{X(t),t>＝0}是维纳过程(Wiener process)或布朗运动。贝叶斯推理为已知的经典的统计归纳推理方法，其定义为已知一个事件集Bi(i＝1,2,...k)中每一Bi的概率P(Bi)，又知在Bi已发生的条件下事件A的条件概率P(A/Bi)，就可得出在给定A已发生的条件下任何Bi的条件概率(逆概率)P(Bi/A)。即P(Bi/A)＝P(Bi)P(A/Bi)/(P(B1)P(A/B1)+P(B2)P(A/B2)+...+P(Bn)P(A/Bn))。
基于此，获得如下公式：SCt：SC t＝SC t-1+ηΔt+σB(Δt)
其中SC t-1表示第一时刻的所述电动车辆的第一系统安全系数；η表示所述电动车辆的失效风险的变化率，B(Δt)表示标准布朗运动，σ是表示所述电动车辆的失效风险的扩散系数。SCt表示第二时刻的所述电动车辆的第二系统安全系数。
在步骤S3中，基于所述第一系统安全系数、所述第二系统安全系数，采用卡尔曼滤波算法计算所述电动车辆的预测安全失效风险值。首先，采用期望极大似然算法(EM)基于所述第一系统安全系数和所述第二系统安全系数估算所述电动车辆的失效风险的变化率η和所述电动车辆的失效风险的扩散系数σ。期望极大似然算法是统计学中惯用的参数估计的方法之一，其通常是在已知某个随机样本满足某种概率分布，但是其中具体的参数不清楚，参数估计就是通过若干次试验，观察其结果，利用结果推出参数的大概值。因此，利用该期望极大似然算法，通过多次实现，可以估算出所述电动车辆的失效风险的变化率η和所述电动车辆的失效风险的扩散系数σ。然后，采用卡尔曼滤波算法基于以下公式计算区间(t,t+P stepΔt]内所述电动车辆的安全失效风险值公式
Figure PCTCN2020089420-appb-000006
其中，t代表当前时刻，P stepΔt]表示失效风 险预测步数。卡尔曼滤波(Kalman filtering)一种利用线性系统状态方程，通过系统输入输出观测数据，对系统状态进行最优估计的算法。由于观测数据中包括系统中的噪声和干扰的影响，所以最优估计也可看作是滤波过程，其表达式为X(k)＝A X(k-1)+B U(k)+W(k)，因此基于该算法可以获得(t,t+P stepΔt]内所述电动车辆的安全失效风险值公式
Figure PCTCN2020089420-appb-000007
SC t表示t时刻的系统安全系数(安全系数预测算法中应用)，其估计值为
Figure PCTCN2020089420-appb-000008
本领域技术人员知悉，P stepΔt]可以根据电动车辆在长期的运营维护过程中，按照经验获得的，电动车辆最可能出现失效的时间周期平均或者加权获得。本领域技术人员可以根据实际需要选择适当的失效风险预测步数值。
车辆失效风险预测是指根据整车完整的实时的所有安全失效统一体现的现状对未来的预测，它是对整车安全性有指示意义的整车重要参数，从而对产品设计生产维保过程的优化提供理论指导。实施本发明的所述的电动车辆的安全失效风险预测方法，可以分析车辆失效风险随时间变化的规律，并对未来的失效风险度进行预测，为车辆的安全运维提供必要的定量化信息基础。
图5是本发明安全失效风险预测方法的计算整车失效风险度的步骤的流程图。如图5所示，在步骤S1中，在设定第一时间区间内，统计所述安全失效中间事件发生的标准化频次。如图3b所示，例如可以将行车制动故障、驻车制动故障，液压压力异常，分别作为一个安全失效中间事件，统计例如在一年之内，其分别发生的标准化频次。
在步骤S2中，将所述安全失效中间事件发生的标准化频次换算到标准工作条件下以获得标准安全失效中间事件发生频次。在本步骤中安全失效底层事件对安全失效中间事件的作用逻辑和影响概率，将同源安全失效中间事件(指不同的安全失效中间事件由同一个安全失效底层事件导致产生)的频次，按安 全失效底层事件影响概率加权合并，得到加权后的安全失效中间事件发生的标准化频次。同样以图3b所示的实施例为例，对于制动弹簧损坏这一安全失效底层事件，其同时对应行车制动故障和驻车制动故障这两个同源安全失效中间事件。同理，制动压力异常这一安全失效底层事件同时对应行车制动故障和驻车制动故障这两个同源安全失效中间事件。而如图3b所示，制动弹簧损坏对于行车制动故障和驻车制动故障的影响概率分别是0.3％和0.4％。通过根据影响概率加权合并行车制动故障和驻车制动故障，就可以获得标准安全失效中间事件发生频次。然后将安全失效中间事件发生的标准化频次，通过统计回归分析方法，换算到待计算的工作条件下，并对所有安全失效中间事件发生的标准化频次求和，得到给定工作条件下、给定时间区间内电气系统失效事件发生的标准化频次(及失效概率，单位是：次/累积工作时长(里程))。安全失效发生频次标准化是指将不同环境参数下统计得到的安全失效发生频次，换算至统一的规定环境参数下，得到可用于全局分析的等价发生频次。依据所述安全失效中间事件的发生机理，分析影响所述安全失效中间事件的事件发生数量的工作条件。比如可以根据道路状况，温度湿度，负载重量等等工作条件，分析影响所述安全失效中间事件的事件发生数量。在湿度大的情况下，发生制动安全事件、转向安全事件和行驶传动安全事件的数量可能较大。在道路状况差的情况下，发生行驶传动安全事件的数量可能较大。基于所述电动车辆的整车控制器、安全控制器和行车记录仪中的数据中记录的数据，可以完成上述分析判断。
随后在步骤S3中，基于所述标准安全失效中间事件发生频次计算所述标准安全失效中间事件对应的风险权值q i。风险权值q i可以用于描述所述标准安全失效中间事件发生频次对失效风险影响程度的参数。当实际的所述标准安全失效中间事件发生频次小于最高容忍频次时，风险权值为所述标准安全失效中 间事件发生频次同最高容忍频次的比值；当实际的所述标准安全失效中间事件发生频次大于或等于最高容忍频次时，风险权值等于1。最高容忍频次是用于安全失效中间事件归一化风险权值的重要参数，其可以由本领域技术人员根据经验设置。通过对电动车辆长期的观察，测试，可以获得该最高容忍频次。本领域中已经有各种对于不同电动车辆的不同安全失效中间事件的最高容忍频次的规定。
在步骤S4中，基于所述标准安全失效中间事件对应的风险权值和风险等级Li计算所述标准安全失效中间事件对应的风险度R i＝q iL i，其中L i＝0，...，10。在本发明中，风险等级Li表征了失效事件(或第i个安全失效中间事件)所引起的与安全性相关的后果。在本发明的优选实施例中，i的具体取值，定义可以参照图3a-3c中示出的安全树。风险等级是对后果严重度的定量评价，通常由专家根据业务特点进行量化定义。本领域中已经有各种对于不同电动车辆的风险等级划分。
在步骤S5中，基于所述安全树根据以下公式计算整车失效风险度：
Figure PCTCN2020089420-appb-000009
其中N表示全部安全失效的数量，n i表示第i安全失效对应的安全失效底层事件的数量。同样以图3b所示的实施例为例，对于制动弹簧损坏这一安全失效底层事件，其同时对应行车制动故障和驻车制动故障这两个同源安全失效中间事件。同理，制动压力异常这一安全失效底层事件同时对应行车制动故障和驻车制动故障这两个同源安全失效中间事件。因此，整车失效风险度可以根据安全树进行计算。R s的最小值为R min＝0，对应于无风险；R s的最大值为R max＝10(n 1+…+n N)，对应于最大风险。
实施本发明的所述的电动车辆的安全失效风险预测方法，可以分析车辆失效风险随时间变化的规律，并对未来的失效风险度进行预测，为车辆的安全运 维提供必要的定量化信息基础。
因此，本发明可以通过硬件、软件或者软、硬件结合来实现。本发明可以在至少一个计算机系统中以集中方式实现，或者由分布在几个互连的计算机系统中的不同部分以分散方式实现。任何可以实现本发明方法的计算机系统或其它设备都是可适用的。常用软硬件的结合可以是安装有计算机程序的通用计算机系统，通过安装和执行程序控制计算机系统，使其按本发明方法运行。
本发明还可以通过计算机程序产品进行实施，程序包含能够实现本发明方法的全部特征，当其安装到计算机系统中时，可以实现本发明的方法。本文件中的计算机程序所指的是：可以采用任何程序语言、代码或符号编写的一组指令的任何表达式，该指令组使系统具有信息处理能力，以直接实现特定功能，或在进行下述一个或两个步骤之后实现特定功能：a)转换成其它语言、编码或符号；b)以不同的格式再现。
因此本发明还涉及一种计算机可读存储介质，其上存储有计算机程序，所述程序被处理器执行时实现所述的电动车辆的安全失效风险预测方法。
本发明还涉及电动车辆，包括处理器，存储在所述处理器中的计算机程序，所述程序被处理器执行时实现所述的电动车辆的安全失效风险预测方法。
实施本发明的所述的电动车辆的安全失效风险预测方法、计算机可读存储介质以及电动车辆，可以分析车辆失效风险随时间变化的规律，并对未来的失效风险度进行预测，为车辆的安全运维提供必要的定量化信息基础。
虽然本发明是通过具体实施例进行说明的，本领域技术人员应当明白，在不脱离本发明范围的情况下，还可以对本发明进行各种变换及等同替代。另外，针对特定情形或材料，可以对本发明做各种修改，而不脱离本发明的范围。因此，本发明不局限于所公开的具体实施例，而应当包括落入本发明权利要求范 围内的全部实施方式。
以上所述仅为本发明的较佳实施例而已，并不用以限制本发明，凡在本发明的精神和原则之内所作的任何修改、等同替换和改进等，均应包含在本发明的保护范围之内。

Claims (9)

  1. 一种电动车辆的安全失效风险预测方法，其特征在于，包括：
    S1、构建安全树，所述安全树包括多个安全失效底层事件、安全失效中间事件、安全失效顶层事件以及所述安全失效底层事件、所述安全失效中间事件、所述安全失效顶层事件之间的逻辑因果关系和安全重要程度；
    S2、根据所述安全树，预测所述电动车辆的安全失效风险。
  2. 根据权利要求1所述的电动车辆的安全失效风险预测方法，其特征在于，所述步骤S2进一步包括：
    S21、计算第一时刻的所述电动车辆的第一系统安全系数；
    S22、采用维纳过程基于贝叶斯推理获得第二时刻的所述电动车辆的第二系统安全系数；
    S23、基于所述第一系统安全系数、所述第二系统安全系数，采用卡尔曼滤波算法计算所述电动车辆的预测安全失效风险值。
  3. 根据权利要求1所述的电动车辆的安全失效风险预测方法，其特征在于，所述步骤S21进一步包括：
    S211、在设定第一时间区间内，统计所述安全失效中间事件发生的标准化频次；
    S212、将所述安全失效中间事件发生的标准化频次换算到标准工作条件下以获得标准安全失效中间事件发生频次；
    S213、基于所述标准安全失效中间事件发生频次计算所述标准安全失效中间事件对应的风险权值q i；
    S214、基于所述标准安全失效中间事件对应的风险权值和风险等级Li计 算所述标准安全失效中间事件对应的风险度R i＝q iL i，其中L i＝0，...，10；
    S215、基于电动车辆的全部安全失效中间事件对应的风险度计算所述第一系统安全系数
    Figure PCTCN2020089420-appb-100001
    其中N表示全部安全失效的数量，n i表示第i安全失效对应的安全失效底层事件的数量。
  4. 根据权利要求3所述的电动车辆的安全失效风险预测方法，其特征在于，所述步骤S22进一步包括：
    基于以下公式计算采用维纳过程基于贝叶斯推理获得第二时刻的所述电动车辆的第二系统安全系数SCt：SC t＝SC t-1+ηΔt+σB(Δt)
    其中SC t-1表示第一时刻的所述电动车辆的第一系统安全系数；η表示所述电动车辆的失效风险的变化率，B(Δt)表示标准布朗运动，σ是表示所述电动车辆的失效风险的扩散系数。
  5. 根据权利要求4所述的电动车辆的安全失效风险预测方法，其特征在于，所述步骤S23进一步包括：
    S231、采用期望极大似然算法基于所述第一系统安全系数和所述第二系统安全系数估算所述电动车辆的失效风险的变化率η和所述电动车辆的失效风险的扩散系数σ；
    S232、采用卡尔曼滤波算法基于以下公式计算区间(t,t+P stepΔt]所述电动车辆的安全失效风险值公式
    Figure PCTCN2020089420-appb-100002
  6. 根据权利要求2所述的电动车辆的安全失效风险预测方法，其特征在于，所述步骤S2进一步包括：
    S2a、在设定第一时间区间内，统计所述安全失效中间事件发生的标准化频次；
    S2b、将所述安全失效中间事件发生的标准化频次换算到标准工作条件下 以获得标准安全失效中间事件发生频次；
    S2c、基于所述标准安全失效中间事件发生频次计算所述标准安全失效中间事件对应的风险权值q i；
    S2d、基于所述标准安全失效中间事件对应的风险权值和风险等级Li计算所述标准安全失效中间事件对应的风险度R i＝q iL i，其中L i＝0，...，10；
    S2e、基于所述安全树根据以下公式计算整车失效风险度：
    Figure PCTCN2020089420-appb-100003
    其中N表示全部安全失效的数量，n i表示第i安全失效对应的安全失效底层事件的数量。
  7. 根据权利要求1-6中任意一项权利要求所述的电动车辆的安全失效风险预测方法，其特征在于，所述步骤S1进一步包括：
    S11.采集电动车辆的整车安全失效数据；
    S12.将所述整车安全失效数据映射归类到不同的安全事件组别中，并分别统计各个安全事件组别频次数据；
    S13.采用联合分析方法对各个安全事件组别中的所述整车安全失效数据进行分类构建安全树。
  8. 一种计算机可读存储介质，其上存储有计算机程序，其特征在于，所述程序被处理器执行时实现根据权利要求1-7中任意一项权利要求所述的电动车辆的安全失效风险预测方法。
  9. 一种电动车辆，其特征在于，包括处理器，存储在所述处理器中的计算机程序，所述程序被处理器执行时实现根据权利要求1-7中任意一项权利要求所述的电动车辆的安全失效风险预测方法。
PCT/CN2020/089420 2019-05-17 2020-05-09 一种电动车辆的安全失效风险预测方法以及电动车辆 WO2020233428A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910415051.3 2019-05-17
CN201910415051.3A CN111144606B (zh) 2019-05-17 2019-05-17 一种电动车辆的安全失效风险预测方法以及电动车辆

Publications (1)

Publication Number Publication Date
WO2020233428A1 true WO2020233428A1 (zh) 2020-11-26

Family

ID=70516777

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/089420 WO2020233428A1 (zh) 2019-05-17 2020-05-09 一种电动车辆的安全失效风险预测方法以及电动车辆

Country Status (2)

Country Link
CN (1) CN111144606B (zh)
WO (1) WO2020233428A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113865885A (zh) * 2021-09-26 2021-12-31 青岛迈金智能科技股份有限公司 一种用于检测自行车损耗的方法及装置
CN115796589A (zh) * 2022-12-05 2023-03-14 三亚学院 纯电动汽车三电系统风险隐患灵敏度分析方法

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115511236B (zh) * 2021-06-22 2023-08-11 中国石油化工股份有限公司 石油化工过程安全风险动态评估方法及装置

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2012003713A (ja) * 2010-06-21 2012-01-05 Hitachi Solutions Ltd 障害原因判定ルール変化検知装置及びプログラム
US8813025B1 (en) * 2009-01-12 2014-08-19 Bank Of America Corporation Customer impact predictive model and combinatorial analysis
CN104392072A (zh) * 2014-12-12 2015-03-04 北京交通大学 一种基于动态故障树的设备集成系统可靠性分析方法
CN105022893A (zh) * 2015-08-11 2015-11-04 河北工业大学 交流充电桩系统故障分析方法
CN105652805A (zh) * 2014-11-14 2016-06-08 中国科学院沈阳计算技术研究所有限公司 基于简化fta方法的数控装置硬件安全保护方法及装置
CN106529581A (zh) * 2016-10-24 2017-03-22 杭州电子科技大学 一种基于贝叶斯网络的桥式起重机故障诊断方法
CN106598029A (zh) * 2016-12-21 2017-04-26 北京交通大学 基于性能退化的列控车载设备可靠性预测方法
CN106647724A (zh) * 2017-02-15 2017-05-10 北京航空航天大学 一种基于车辆异常数据监测的t‑box信息安全检测及防护方法

Family Cites Families (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8260493B2 (en) * 2010-02-17 2012-09-04 GM Global Technology Operations LLC Health prognosis for complex system using fault modeling
CN103049346B (zh) * 2012-12-11 2015-03-18 工业和信息化部电子第五研究所 基于失效物理的元器件故障树构建方法和系统
US20210133871A1 (en) * 2014-05-20 2021-05-06 State Farm Mutual Automobile Insurance Company Autonomous vehicle operation feature usage recommendations
CN105426692B (zh) * 2015-12-10 2018-06-05 青岛农业大学 基于数据驱动的海洋平台多阶段任务系统可靠性估计方法
CN105573855A (zh) * 2015-12-22 2016-05-11 浙江师范大学 一种用于故障树决策图分析的模块化事件排序方法
CN106444703B (zh) * 2016-09-20 2018-12-07 西南石油大学 基于故障模式发生概率的动设备运行状态模糊评价及预测方法
DE102018120328A1 (de) * 2017-08-24 2019-02-28 Infineon Technologies Ag Sicherheitseinrichtung mit erweiterter Zuverlässigkeit
CN107909268A (zh) * 2017-11-15 2018-04-13 吉林大学 一种公交车失火事故风险评估的方法
CN109094482B (zh) * 2018-07-25 2021-08-24 哈尔滨工业大学 车辆运行风险评估信息采集系统、风险评估系统及方法

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8813025B1 (en) * 2009-01-12 2014-08-19 Bank Of America Corporation Customer impact predictive model and combinatorial analysis
JP2012003713A (ja) * 2010-06-21 2012-01-05 Hitachi Solutions Ltd 障害原因判定ルール変化検知装置及びプログラム
CN105652805A (zh) * 2014-11-14 2016-06-08 中国科学院沈阳计算技术研究所有限公司 基于简化fta方法的数控装置硬件安全保护方法及装置
CN104392072A (zh) * 2014-12-12 2015-03-04 北京交通大学 一种基于动态故障树的设备集成系统可靠性分析方法
CN105022893A (zh) * 2015-08-11 2015-11-04 河北工业大学 交流充电桩系统故障分析方法
CN106529581A (zh) * 2016-10-24 2017-03-22 杭州电子科技大学 一种基于贝叶斯网络的桥式起重机故障诊断方法
CN106598029A (zh) * 2016-12-21 2017-04-26 北京交通大学 基于性能退化的列控车载设备可靠性预测方法
CN106647724A (zh) * 2017-02-15 2017-05-10 北京航空航天大学 一种基于车辆异常数据监测的t‑box信息安全检测及防护方法

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113865885A (zh) * 2021-09-26 2021-12-31 青岛迈金智能科技股份有限公司 一种用于检测自行车损耗的方法及装置
CN113865885B (zh) * 2021-09-26 2024-04-26 青岛迈金智能科技股份有限公司 一种用于检测自行车损耗的方法及装置
CN115796589A (zh) * 2022-12-05 2023-03-14 三亚学院 纯电动汽车三电系统风险隐患灵敏度分析方法
CN115796589B (zh) * 2022-12-05 2023-09-29 三亚学院 纯电动汽车三电系统风险隐患灵敏度分析方法

Also Published As

Publication number Publication date
CN111144606A (zh) 2020-05-12
CN111144606B (zh) 2020-09-15

Similar Documents

Publication Publication Date Title
WO2020233428A1 (zh) 一种电动车辆的安全失效风险预测方法以及电动车辆
KR102263337B1 (ko) 클라우드 기반 차량 고장 진단 방법, 장치 및 시스템
US20190384257A1 (en) Automatic health indicator learning using reinforcement learning for predictive maintenance
KR101713985B1 (ko) 예지 정비 방법 및 장치
US11868101B2 (en) Computer system and method for creating an event prediction model
US20120245791A1 (en) Apparatus and method for predicting mixed problems with vehicle
DE102020124693B4 (de) Adaptives Prognosesystem und Prognoseverfahren für Fahrzeuge
CA2389274C (en) Apparatus and method for performance and fault data analysis
WO2020211844A1 (zh) 基于安全树概率和安全重要度的电动车辆安全控制方法和电动车辆
WO2020211846A1 (zh) 一种基于安全树模型的电动车辆运维优化方法
DE102016004534A1 (de) Verfahren zur Fahrzeugdiagnose
WO2020233429A1 (zh) 一种电动车辆的安全状态评估方法以及电动车辆
CN110084919B (zh) 一种电动车辆的安全树构建方法以及电动车辆
Davari et al. A fault detection framework based on lstm autoencoder: A case study for volvo bus data set
Kopp et al. Early reliability estimation in automotive industry
US20220284740A1 (en) Method for determining the operating state of vehicle components
CN111143752B (zh) 一种电动车辆的安全重要度的计算方法以及电动车辆
Kilian et al. Best Practices for Advanced Modelling of Safety Mechanisms in an FTA
CN110077234B (zh) 一种电动车辆安全树更新方法和电动车辆
CN111144681A (zh) 一种电动车辆的关键重要度的计算方法以及电动车辆
CN110223416B (zh) 一种电动车辆的原始数据分析方法以及电动车辆
CN116579677B (zh) 一种高速铁路电务车载设备全生命周期管理方法及系统
KR20230157806A (ko) 클러스터링 알고리즘에 기반하여 차량 내 모터의 고장 여부를 판단하는 방법 및 그를 이용한 시스템
WO2024013879A1 (ja) 診断制御装置及び診断制御方法
WO2024004313A1 (ja) 不具合診断システム、不具合診断装置および不具合診断方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20808714

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20808714

Country of ref document: EP

Kind code of ref document: A1