WO2018129917A1 - 一种基于云的车辆故障诊断方法、装置及其系统 - Google Patents

一种基于云的车辆故障诊断方法、装置及其系统 Download PDF

Info

Publication number
WO2018129917A1
WO2018129917A1 PCT/CN2017/097124 CN2017097124W WO2018129917A1 WO 2018129917 A1 WO2018129917 A1 WO 2018129917A1 CN 2017097124 W CN2017097124 W CN 2017097124W WO 2018129917 A1 WO2018129917 A1 WO 2018129917A1
Authority
WO
WIPO (PCT)
Prior art keywords
monitoring data
fault
vehicle
feature vector
data
Prior art date
Application number
PCT/CN2017/097124
Other languages
English (en)
French (fr)
Inventor
刘真通
张永生
张伟
Original Assignee
华为技术有限公司
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 华为技术有限公司 filed Critical 华为技术有限公司
Priority to EP22177467.2A priority Critical patent/EP4119919A1/en
Priority to EP17890940.4A priority patent/EP3564647B1/en
Priority to JP2019538380A priority patent/JP6830540B2/ja
Priority to KR1020197023618A priority patent/KR102263337B1/ko
Publication of WO2018129917A1 publication Critical patent/WO2018129917A1/zh
Priority to US16/510,289 priority patent/US11468715B2/en
Priority to US17/939,438 priority patent/US12086165B2/en

Links

Images

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W50/00—Details of control systems for road vehicle drive control not related to the control of a particular sub-unit, e.g. process diagnostic or vehicle driver interfaces
    • B60W50/02—Ensuring safety in case of control system failures, e.g. by diagnosing, circumventing or fixing failures
    • B60W50/0205—Diagnosing or detecting failures; Failure detection models
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/28—Databases characterised by their database models, e.g. relational or object models
    • G06F16/284—Relational databases
    • G06F16/285—Clustering or classification
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W50/00—Details of control systems for road vehicle drive control not related to the control of a particular sub-unit, e.g. process diagnostic or vehicle driver interfaces
    • B60W50/08—Interaction between the driver and the control system
    • B60W50/14—Means for informing the driver, warning the driver or prompting a driver intervention
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
    • G01M17/00—Testing of vehicles
    • G01M17/007—Wheeled or endless-tracked vehicles
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B23/00—Testing or monitoring of control systems or parts thereof
    • G05B23/02—Electric testing or monitoring
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/22—Indexing; Data structures therefor; Storage structures
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24—Querying
    • G06F16/245—Query processing
    • G06F16/2455—Query execution
    • G06F16/24568—Data stream processing; Continuous queries
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
    • G—PHYSICS
    • G07—CHECKING-DEVICES
    • G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
    • G07C5/00—Registering or indicating the working of vehicles
    • G07C5/008—Registering or indicating the working of vehicles communicating information to a remotely located station
    • G—PHYSICS
    • G07—CHECKING-DEVICES
    • G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
    • G07C5/00—Registering or indicating the working of vehicles
    • G07C5/02—Registering or indicating driving, working, idle, or waiting time only
    • G—PHYSICS
    • G07—CHECKING-DEVICES
    • G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
    • G07C5/00—Registering or indicating the working of vehicles
    • G07C5/08—Registering or indicating performance data other than driving, working, idle, or waiting time, with or without registering driving, working, idle or waiting time
    • G07C5/0841—Registering performance data
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/10—Protocols in which an application is distributed across nodes in the network
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60Y—INDEXING SCHEME RELATING TO ASPECTS CROSS-CUTTING VEHICLE TECHNOLOGY
    • B60Y2200/00—Type of vehicle
    • B60Y2200/90—Vehicles comprising electric prime movers
    • B60Y2200/91—Electric vehicles
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60Y—INDEXING SCHEME RELATING TO ASPECTS CROSS-CUTTING VEHICLE TECHNOLOGY
    • B60Y2306/00—Other features of vehicle sub-units
    • B60Y2306/15—Failure diagnostics

Definitions

  • the present invention relates to the field of fault diagnosis, and in particular, to a cloud-based vehicle fault diagnosis method, apparatus and system thereof.
  • Car fault diagnosis is related to the safety of the vehicle and its drivers, and is a necessary measure to ensure the normal driving of the car.
  • the existing fault diagnosis techniques are mainly qualitative analysis methods (such as fault diagnosis based on expert systems, quantitative analysis methods (such as analytical model-based fault diagnosis, data-driven fault diagnosis); among them, data-driven fault diagnosis is used. Most of them are fault diagnosis techniques based on machine learning algorithms.
  • the prior art provides a cloud-based automobile fault detection system, which is based on cloud computing and can solve the secondary problem that the computing power of the single-chip microcomputer is insufficient or the detecting equipment is expensive and inconvenient to install on the automobile, but the system logic is simple, and There is no specific technical solution for how to accurately diagnose faults, and the accuracy of fault diagnosis is low, and it is difficult to ensure the safety of the vehicle.
  • the embodiment of the invention provides a cloud-based vehicle fault diagnosis method, system and device thereof, which can improve the accuracy of fault diagnosis and reduce the diagnosis time.
  • the first aspect provides a cloud-based vehicle fault diagnosis method, the method comprising: receiving monitoring data uploaded by a vehicle, wherein the monitoring data refers to data of a working state of a component or a functional system thereof monitored by the monitoring device; Parts refer to the components that make up the vehicle, such as: brakes, transmissions, compressors, tire pressure monitors, pumps, etc.; functional systems are a number of components that are used to achieve a certain function, such as: battery management system, a brake safety system, a power system, etc.; extracting a feature vector of the monitoring data from the monitoring data, the extracted feature vector is a set of numbers characterizing the monitoring data; for example, the feature vector is calculated by averaging or variance of the monitoring data The resulting set of mean or variance values corresponding to the raw data of the monitoring data, optionally, a set of numbers denoted as ⁇ A, B, C, D...Z ⁇ ; to monitor the components of the vehicle from which the data came or Functional system For the label, the feature vector of the monitoring data is classified and
  • the method before extracting the feature vector of the monitoring data from the monitoring data, the method further includes: parsing the received monitoring data to obtain the parsed monitoring Data; storing the parsed monitoring data in a tag by using a component or a functional system of the vehicle from which the monitoring data is derived; wherein the tag stored for the classification of the parsed monitoring data is stored with the tag stored for the feature vector
  • extracting the feature vector of the monitoring data from the monitoring data specifically includes: extracting a feature vector of the parsed monitoring data from the parsed monitoring data.
  • the method further comprises: periodically deleting the component from which the monitoring data represented by the most recently extracted feature vector is derived or The previously stored feature vector of the same functional system.
  • the second aspect provides a cloud-based vehicle fault diagnosis apparatus, including: a monitoring data receiving module, a data pre-processing module, a feature database, and a fault diagnosis module; and the monitoring data receiving module is configured to receive monitoring data uploaded by the vehicle, where The monitoring data is working state data of the component or function system monitored by the vehicle; the data pre-processing module is configured to extract a feature vector of the monitoring data from the monitoring data received by the monitoring data receiving module, The feature vector is a set of numbers representing the monitoring data; the feature database is configured to extract the data preprocessing module by using a component or a functional system of the vehicle from which the monitoring data is derived.
  • the feature vector classification is stored; the fault diagnosis module is configured to perform fault diagnosis on the feature vector classified and stored in the feature database based on a support vector machine algorithm.
  • the apparatus further includes: a central database; the central database is configured to: parse the monitoring data received by the monitoring data receiving module to obtain The parsed monitoring data; the parsed monitoring data is stored in a category by using a component or a functional system of the vehicle from which the monitoring data is derived; wherein the classified monitoring data is stored for the parsing
  • the label corresponds to the label stored for the classification of the feature vector; the data pre-processing module is specifically configured to: extract feature vectors of the parsed monitoring data from the monitoring data parsed by the central database.
  • the feature database is further configured to: periodically delete the monitoring data represented by the recently extracted feature vector The previously stored feature vector from the same component or functional system.
  • the central database deletes the monitoring data corresponding to the feature vector of the feature database deletion.
  • a third aspect provides a cloud-based vehicle fault diagnosis system, the system comprising: the second aspect or the first implementation of the second aspect or the apparatus, vehicle in the second implementation manner of the second aspect;
  • the monitored data is uploaded to the device of the second aspect or the first implementation of the second aspect or the second implementation of the second aspect; the second aspect or the first implementation or the second aspect of the second aspect
  • the device in the second implementation of the aspect performs fault diagnosis based on the received data.
  • the feature vector of the monitoring data from different components or functional systems is classified and stored according to the support vector machine algorithm, and the feature vectors stored in the classification are diagnosed in parallel, which not only shortens the diagnosis time but also avoids Different data in the data transfer process affect each other and improve the accuracy of fault diagnosis.
  • FIG. 1 is a schematic diagram of a cloud-based fault diagnosis system according to an embodiment of the present invention
  • FIG. 2 is a schematic diagram of a cloud diagnostic apparatus according to an embodiment of the present invention.
  • FIG. 3 is a schematic diagram of a central database according to an embodiment of the present invention.
  • FIG. 4 is a schematic diagram of a periodic storage unit of a central database according to an embodiment of the present invention.
  • FIG. 5 is a schematic diagram of a feature database according to an embodiment of the present invention.
  • FIG. 6 is a schematic diagram of a periodic storage unit of a feature database according to an embodiment of the present invention.
  • FIG. 7 is a flowchart of a cloud-based fault diagnosis method according to an embodiment of the present invention.
  • FIG. 8 is a schematic diagram of a parallel calculation according to an embodiment of the present invention.
  • FIG. 9 is a flowchart of an offline training and a method for testing a fault classification model according to an embodiment of the present invention.
  • FIG. 10 is a flowchart of a method for internal decision making of a system degrading decision machine according to an embodiment of the present invention.
  • FIG. 11 is a flow chart of a battery pack current sensor fault diagnosis method according to an embodiment of the present invention.
  • Embodiments of the present invention provide a cloud-based vehicle fault diagnosis system, which can be used for real-time or/and online diagnosis of vehicle faults, statistical vehicle function system/parts fault data, etc., as shown in FIG. 1, the system includes the following The components are: cloud diagnostic device 1000, vehicle 2000, vehicle or component manufacturer 3000, maintenance service provider 4000, and other devices 5000.
  • the cloud diagnostic device 1000 implements data interaction with the vehicle 2000, the vehicle or component manufacturer 3000, the service provider 4000, and other devices 5000 through wireless communication technology; optionally, the system does not impose any restrictions on the wireless communication technology. Is one or more of the wireless communication technologies under any protocol.
  • the cloud diagnostic device 1000 can diagnose and locate faults occurring in the vehicle in real time based on the monitoring data uploaded by the vehicle 2000, and manage and count the fault data; further, the processed or statistical fault data can be sent to the corresponding vehicle 2000, the entire vehicle, or Component manufacturer 3000, service provider 4000 or other equipment 5000; for example, component manufacturer A wants to know the fault condition of component A produced by it, cloud diagnostic device 1000 can send the fault data of component A obtained by statistics to Parts Manufacturer A, the fault data of the component A includes but is not limited to the number of vehicles in which the component A fails, the number of failures of the component A of a certain vehicle, and the like; for example, the vehicle B wants to know the fault condition of the real-time operation.
  • the cloud diagnostic device 1000 can transmit the fault data of the vehicle B obtained according to the monitored data uploaded in real time by the vehicle B to the vehicle B.
  • the fault data of the vehicle B includes, but is not limited to, the safety factor of the vehicle B and the fault of a certain component. Tips and so on.
  • the vehicle 2000 is used to indicate one or more vehicles connected to the cloud diagnostic device 1000 by wireless communication technology, and does not specifically refer to a certain vehicle in motion; the vehicle 2000 is configured with a monitoring sensing device for monitoring vehicle running data or zero.
  • the running data of the component may be uploaded to the cloud diagnostic device 1000 according to the fault diagnosis requirement or the instruction setting, and the monitoring data uploaded by the vehicle 2000 is further processed by the cloud diagnostic device 1000.
  • the vehicle or component manufacturer 3000, the service provider 4000, and other equipment 5000 are not necessary components of the system, and they acquire/receive data related to the fault from the cloud diagnostic device 1000 based on their respective needs, for analyzing a certain The probability, frequency, and impact on the vehicle/functional system/components.
  • the embodiment of the invention can solve the limitation of the single-chip computer single-chip computing capability, improve the accuracy of the fault diagnosis, and realize the unified management of numerous faults of many vehicles based on the cloud.
  • the data is shared with the vehicle/parts manufacturer, repair service provider, and other equipment (such as third-party monitoring equipment) to solve the problem from the source, improve the safety of the vehicle/parts, and ensure the safety of the vehicle.
  • the cloud-based diagnostic system is not limited to the diagnosis and management of vehicle/parts faults, and is also applicable to the diagnosis and management of faults such as ships, airplanes, trains, and drones.
  • the apparatus 1000 includes: a monitoring data receiving module 1010, a central database 1020, a data preprocessing module 1030, a feature database 1040, a fault diagnosis module 1050, and a fault level. Decision module 1060, system downgrade decision machine 1070, fault statistics module 1080.
  • the monitoring data receiving module 1010 is configured to receive monitoring data uploaded by the vehicle, wherein the monitoring data refers to working state data of a component or a function system monitored by the vehicle.
  • the monitoring data is a vehicle or a component detected by the vehicle. Or data related to the working state of the functional system;
  • the central database 1020 is configured to parse the monitoring data received by the monitoring data receiving module 1010 to obtain the parsed monitoring data; specifically, parsing the data packet uploaded by the vehicle, and inputting the parsed data into the data preprocessing module 1030;
  • the central database is further configured to: categorize and store the parsed monitoring data by using a component or a functional system from which the monitoring data comes from, thereby establishing a relatively complete database for monitoring data, and the database can be used for Later, analyze the impact on the life of the vehicle/parts after a certain fault occurs, or use it to improve the fault diagnosis system.
  • the central database 1020 includes: a temporary storage unit 1021, a periodic storage unit 1022; optionally, a temporary storage unit 1021 and a periodic storage unit 1022 are used as components (component 1, component 2, ...
  • the component performs structured classification storage management for the category;
  • the temporary storage unit 1021 is configured to temporarily store the real-time data uploaded by the vehicle;
  • the periodic storage unit 1022 is configured to store the diagnosis result output by the fault diagnosis module 1050 as a fault label.
  • the component i is used as an example to describe the structured storage management of the periodic storage unit 1022.
  • the component i has a relatively independent storage area, and the storage area is divided into a faultless data area and a fault data area, and fault data.
  • the area can be further subdivided into sensor type data, actuator type fault data or other fault data; corresponding to each type of fault, it can be further classified in detail, from fault 1 to fault n, for example, sensor fault class data can be divided into current sensor fault data. , voltage sensor fault data, temperature sensor fault data, and pressure sensor fault data.
  • periodic storage unit 1022 can periodically clean the stored data (eg, weekly, monthly, or yearly).
  • the data pre-processing module 1030 is configured to perform feature vector extraction and dimension reduction processing on the monitoring data (also referred to as raw data) input by the central database 1020, to reduce the data amount and extract valid data feature vectors, thereby shortening the fault diagnosis time. And improving the accuracy of the fault diagnosis; wherein the feature vector is a set of numbers representing the monitoring data, optionally, performing average or variance calculation on the monitoring data to obtain an average value or a variance value, corresponding to the average value of the monitoring data Or a set of variance values can be thought of as a set of numbers; optionally, a set of numbers can be represented as ⁇ A, B, C, D...Z ⁇ .
  • the feature database 1040 is configured to store the feature vector obtained after the data preprocessing module processes; further, as shown in FIG. 5, the feature database 1040 includes: a temporary storage unit 1041 and a periodic storage unit 1042.
  • the storage management of the feature database 1040 is similar to the structured classification storage management of the central database 1020, and the structured classification storage management is performed by using the component as a category and the diagnosis result output by the fault diagnosis module 1050 as a label, as shown in FIG. The description can be seen in the structured classification storage management described by the central database 1020. It should be clarified that the feature database 1040 stores the feature vector data in real time (also referred to as raw data) stored in the Huang Zongyang database.
  • the feature database 1040 is periodically cleaned to store the feature vector data; the principle of periodic cleaning is: (1) the feature vector is similar only needs to retain the latest feature vector; (2) correspondingly, the periodicity of the central database 1020 The storage unit 1022 only needs to retain real-time data (also referred to as raw data) corresponding to the feature vectors retained by the feature database 1040.
  • the fault diagnosis module 1050 is configured to perform fault diagnosis on the categorized stored feature vectors in parallel based on a machine learning algorithm.
  • the machine learning algorithm used by the fault diagnosis module 1050 is a support vector machine algorithm based on a decision-oriented acyclic graph DDAG. The specific fault diagnosis is described in the following embodiments, and details are not described herein again.
  • the fault level determination module 1060 is configured to classify the fault diagnosis results output by the fault diagnosis module 1050.
  • the level is divided into: level 1 fault (most serious), level 2 fault, level 3 fault...; optional, the grade is divided into: severe fault, medium fault, general fault.
  • the system degradation decision machine 1070 is configured to perform a decision according to the fault level determined by the fault level determination module 1060 and/or input relevant data of the fault diagnosis result, and pass the corresponding danger beyond the expected safety state or the fault affecting the safety state of the whole vehicle.
  • the warning signal informs the vehicle that the fault occurs, for example, the system degradation decision machine 1070 determines that the fault occurred in the brake system exceeds the expected safety state according to the relevant data of the brake system fault diagnosis result, and prompts the vehicle that the fault occurs through the danger warning signal; further Ground, request the vehicle to be safe to park and prompt the vehicle to repair as soon as possible.
  • the fault statistics module 1080 is configured to receive the fault diagnosis result outputted by the fault level determining module 1060, and perform partition management and statistics according to the component/system; further, the fault statistics module 1080 is specifically used for one or more of the following contents but It is not limited to the following examples: counting the probability of failure of each component, counting the probability of occurrence of each type of failure in each component, counting the probability of occurrence of a failure within each component, and counting all components.
  • the probability of occurrence of different levels of faults, etc.; optionally, the period of statistics may be any period of time, such as one year, three months, one month, n weeks, n days, etc.; optionally, the fault statistics module 1080 is used to calculate faults. The results are sent to the appropriate vehicle or component manufacturer, repair service provider, and other equipment.
  • the cloud-based vehicle fault diagnosis apparatus provided by the embodiment of the invention can perform fault diagnosis in parallel by using feature vectors extracted from monitoring data of different components or functional systems based on the support vector machine algorithm, thereby not only shortening the diagnosis time, It can also avoid the influence of different data in the data transmission process and improve the accuracy of fault diagnosis.
  • the embodiment of the invention provides a cloud-based vehicle fault diagnosis method, as shown in FIG. 7 , the specific steps of the method The sudden is:
  • S100 The vehicle uploads the monitoring data of the monitored function system/parts to the cloud diagnostic device/system; optionally, the vehicle uploads the monitoring data directly to the central database; optionally, the monitored data is monitored The data is packaged and uploaded to the central database; optionally, the vehicle directly uploads the monitored monitoring data of the functional system/components to the data preprocessing module;
  • the central database receives the monitoring data uploaded by the vehicle, and parses the monitoring data and transmits the monitoring data to the data preprocessing module.
  • the central database further performs structured classification storage management on the received or parsed monitoring data.
  • the data pre-processing module receives the parsed monitoring data transmitted by the central database, and performs fault feature extraction on the received monitoring data to obtain a feature vector and transmits the extracted feature vector to the feature database; further, the data pre- The processing module extracts the feature vector by using the wavelet packet decomposition to obtain the feature vector, and then performs the dimension reduction processing on the extracted feature vector through the kernel principal component analysis to obtain the dimension vector after the dimension reduction; wherein the wavelet packet algorithm can Multi-level frequency band division of signals in the whole frequency band, so the completeness of fault feature extraction is high; alternatively, in order to reduce the computational complexity of the fault diagnosis classifier and improve the accuracy of fault separation, a radial basis can be used -
  • the kernel principal component analysis algorithm performs feature selection and dimension reduction processing on the extracted feature vector; optionally, the data preprocessing module receives real-time data directly uploaded by the vehicle;
  • the feature vector obtained through the foregoing processing may also be directly transmitted to the fault diagnosis module for fault diagnosis and positioning;
  • the feature database receives the feature vector transmitted by the data preprocessing module, and adopts a structured classification and storage management for the received feature vector data.
  • the module transmits the feature vector to the fault diagnosis module; notably, the feature database is not a necessary module, and the function of the module is to better manage the feature vector;
  • the fault diagnosis module receives the feature vector transmitted by the feature database, and performs real-time diagnosis and location of the fault based on the machine learning algorithm.
  • the machine learning algorithm is a support vector machine algorithm based on the decision-guided acyclic graph DDAG.
  • the fault diagnosis module includes one or more fault diagnosis units, and the fault diagnosis unit may be a corresponding component/function system configuration, and the input feature vector is paralleled according to the component/function system division.
  • the fault diagnosis unit of the corresponding component can diagnose and locate the corresponding components and their internal components;
  • the fault diagnosis unit of the system can diagnose and locate the entire system or a certain functional system, and can avoid misjudgment caused by the impact of the data transmission process on the diagnosis;
  • the fault diagnosis based on support vector machine needs to construct the fault classifier offline, which is the support vector machine training model.
  • the PSO algorithm is used to optimize the penalty factor parameters of the support vector machine.
  • radial basis kernel function parameters to improve the accuracy of fault diagnosis.
  • the vehicle power system or key components such as the engine, drive motor, high-voltage battery system, inverter, DCDC, OBC, automatic driving or assisted driving system, according to sensor failure, actuator failure and other faults Rack or real vehicle road test to collect data when different components have different faults;
  • the embodiment of the invention provides a cloud-based fault diagnosis method, which can diagnose faults of different components/systems in parallel by using a support vector machine algorithm, thereby shortening the diagnosis time and avoiding data.
  • the fault diagnosis module transmits the diagnosis result to the central database, the feature database, and the fault level determination module respectively; further, the central database and the feature database receive the diagnosis result, and use the diagnosis result as a label to classify and manage the corresponding data.
  • the central database and the feature database receive the diagnosis result, and use the diagnosis result as a label to classify and manage the corresponding data.
  • the S700 the fault level determining module performs level determination on the received diagnosis result, and the specific level division is described in the foregoing embodiment, and details are not described herein; further, the diagnosis result of the determination level is transmitted to the system degrading decision machine.
  • Fault statistics module the fault level determining module performs level determination on the received diagnosis result, and the specific level division is described in the foregoing embodiment, and details are not described herein; further, the diagnosis result of the determination level is transmitted to the system degrading decision machine.
  • the fault statistics module performs data statistics on the received diagnosis result.
  • the fault statistics module sends the statistical data to the manufacturer. , service providers, etc., used to improve products and / or services;
  • the S900 the system demotion decision machine receives the diagnosis result of the level determination transmitted by the fault level determination module, and sends a corresponding danger warning signal to the vehicle based on the result decision output by the fault level determination module; the internal decision of the system degrading decision machine and The control process for sending a hazard warning signal is shown in Figure 10, as follows:
  • System downgrade decision opportunity The decision is made based on the result of the fault level determination module output. If the fault that is diagnosed is determined to be a fault of the severity level by the fault level determination module, the system degradation decision opportunity directly sends a corresponding danger warning signal to the vehicle; optionally, the danger warning signal indicates that the vehicle has experienced a serious fault, or / and request emergency handling of the serious fault; optionally, the severity level fault means that if the vehicle is likely to cause loss of control, it will endanger the driver's life; for example, if the brake pedal of the vehicle fails, the braking performance of the vehicle Can not be guaranteed, can be classified as the most serious failure;
  • the system downgrade decision opportunity performs a comprehensive judgment based on the real-time data uploaded by the vehicle:
  • the vehicle system data uploaded by the vehicle will be input into the vehicle model to determine whether the vehicle is in a safe state; on the other hand, the uploaded component data will be input to the corresponding component model to determine whether the component is in an expected state.
  • the safety state further, if the vehicle or a component determines that it is in a critically dangerous state, the system degradation decision opportunity sends a corresponding danger warning signal to the vehicle.
  • the vehicle model and the component model can be constructed by mathematical formulas, or Through the intelligent network training such as auditing network; in the specific implementation, the important parameter values of the safety state of the reaction system can be calculated based on the measured input/output signals of the system (the whole vehicle or component) and the model of the system, if the parameter value exceeds the expected value
  • the safety scope, the system demotion decision opportunity determines that the vehicle exceeds the expected safety state at this time, and sends a corresponding system hazard warning signal to the vehicle, which can analyze the safety state of the vehicle in real time to ensure the safe driving of the vehicle.
  • the vehicle transmits the monitored real-time data about the battery pack current sensor to the temporary storage unit of the central database, and the data pre-processing module;
  • the data preprocessing module sends the extracted feature vector to the fault diagnosis module by performing feature extraction operation on the received data; further, sends the fault diagnosis unit to the battery; and the data preprocessing module further features
  • the vector is stored in the feature database, and the structure and storage manner of the feature database can refer to the central database; specifically, the battery-related feature vector is stored in the battery area of the feature database;
  • the battery fault diagnosis unit determines that the battery pack current sensor has failed, and sends the diagnosis result to the fault statistics module, the fault level determination module, the central database, and the feature database;
  • the central database will use the battery pack current sensor fault as the fault label, and transfer the real-time data about the battery stored in the temporary storage unit to the periodic storage unit - battery area - fault data area - sensor fault area - current sensor In the fault area; the feature database also adopts a similar method to transfer the feature vector of the temporary storage unit to the regular storage unit;
  • the fault diagnosis results will also be sent to the whole vehicle and third-party monitoring equipment (such as mobile phones);
  • the fault statistics module will calculate the probability of the battery pack current sensor failure, and send the result to the whole vehicle or parts factory, as well as the manufacturer maintenance (4S) store;
  • the fault severity level determination module determines whether the fault is a serious fault, and if it is determined to be a serious fault, sends the diagnosis result to the system degradation decision machine;
  • the demotion decision machine makes a demotion order decision based on the serious fault, and sends a corresponding hazard warning signal to the VCU of the whole vehicle, requests the power to be reduced to zero, and cuts off the high voltage; at the same time, reminds the driver that the power system failure occurs, and the side needs to be turned parking.
  • the battery pack current sensor fault diagnosis method stores the adjustment vector of the battery pack current sensor and stores the adjustment vector in the corresponding classification storage, and the battery fault diagnosis unit determines that the battery pack current sensor is faulty according to the feature vector.
  • the influence of monitoring data of other components/systems on the monitoring data of the battery pack current sensor can be avoided, and the accuracy of fault diagnosis can be improved.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Evolutionary Biology (AREA)
  • Artificial Intelligence (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Evolutionary Computation (AREA)
  • Automation & Control Theory (AREA)
  • Transportation (AREA)
  • Human Computer Interaction (AREA)
  • Mechanical Engineering (AREA)
  • Software Systems (AREA)
  • Signal Processing (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Computational Linguistics (AREA)
  • Vehicle Cleaning, Maintenance, Repair, Refitting, And Outriggers (AREA)
  • Testing And Monitoring For Control Systems (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Traffic Control Systems (AREA)

Abstract

本发明实施例提供了一种基于云的车辆故障诊断方法，接收车辆上传的监测数据；从所述监测数据中提取所述监测数据的特征向量；以所述监测数据所来自的所述车辆的零部件或功能系统为标签，将所述监测数据的特征向量分类存储；基于支持向量机算法，对分类存储的所述特征向量并行地进行故障诊断。本发明技术方案能够通过基于支持向量机算法，将从不同零部件或功能系统的监测数据所提取的特征向量并行地进行故障诊断,不仅可以缩短诊断时间，还可以避免数据传递过程中的不同数据的彼此影响,提高故障诊断的准确性。

Description

一种基于云的车辆故障诊断方法、装置及其系统 技术领域
本发明涉及故障诊断领域，尤其涉及一种基于云的车辆故障诊断方法、装置及其系统。
背景技术
汽车已经成为人们普遍选择的日常出行交通工具，随着经济、科学技术的发展，电动汽车越来越普及，智能化、互联化汽车开始出现，汽车的设计与生产业越来越多的采用电子技术、自动化技术和计算机技术，一方面使得汽车的自动化程度越来越高，另一方面也对汽车的维修和监测提出了更高的要求，由于电脑控制系统的应用，汽车的结构也变得越来越复杂，增加了汽车的故障诊断难度。
汽车故障诊断关系到车辆及其驾驶人员的安全，是保障汽车正常行驶的必要措施。现有的故障诊断技术主要为定性分析法(例如基于专家系统的故障诊断、定量分析法(例如基于解析模型的故障诊断、基于数据驱动的故障诊断)；其中，基于数据驱动的故障诊断使用较多的是基于机器学习算法的故障诊断技术。定性分析故障诊断只能实现离线诊断，并且对较小故障的诊断准确率较低；基于解析模型的故障诊断以及基于机器学习算法的故障诊断都可以准确地诊断系统发生的较小故障；其中，基于解析模型的故障诊断实时性较好，但是对于较为复杂较大的非线性系统其解析模型很难构建；基于机器学习算法的故障诊断技术可以实现多种故障的诊断管理，但是其计算量较大，计算时间较长，现有的单片机并不能满足其在线使用。随着云计算的发展及成熟，其可以帮助解决基于机器学习算法故障诊断计算量大以及耗时长的问题。
现有技术提供了一种基于云计算的汽车故障检测系统，该系统基于云计算可以解决单片机计算能力不足或检测设备昂贵、不方便安装于汽车上等次要问题，但是该系统逻辑简单，并没有针对如何准确诊断故障提供具体技术方案，并且故障诊断的准确率低，难以保证车辆的安全。
发明内容
本发明实施例提供了一种基于云的车辆故障诊断方法、系统及其装置，可以提高故障诊断的准确度，减少诊断时间。
第一方面提供了一种基于云的车辆故障诊断方法，该方法包括：接收车辆上传的监测数据,其中，监测数据是指车辆通过监测设备监测到的其零部件或功能系统的工作状态的数据，零部件是指组成车辆的配件，例如：制动器、变速器、压缩机、胎压监测器、水泵等；功能系统是指多个部件组成的用于实现一定功能的整体，例如：电池管理系统、制动安全系统、动力系统等；从监测数据中提取该监测数据的特征向量,提取的特征向量是表征所述监测数据的一组数；例如，特征向量是通过对监测数据进行平均或方差计算得到的对应于监测数据原始数据的平均值或方差值的集合,可选的，一组数表示为{A，B，C，D…Z}；以监测数据所来自的车辆的零部件或功能系统 为标签，将监测数据的特征向量分类存储；基于支持向量机算法，对分类存储的特征向量并行地进行故障诊断。
结合第一方面，在第一方面的第一种可能实现的方式中，在从监测数据中提取该监测数据的特征向量之前，还包括：将接收到的监测数据进行解析，得到解析后的监测数据；以监测数据所来自的车辆的零部件或功能系统为标签，将解析后的监测数据分类存储；其中，针对解析后的监测数据的分类存储的标签与针对特征向量的分类存储的标签相对应；从所述监测数据中提取所述监测数据的特征向量具体包括：从所述解析后的监测数据中提取该解析后的监测数据的特征向量。
结合第一方面的第一种可能实现的方式，在第一方面的第二种可能实现的方式中，该方法还包括：定期删除与最近提取的特征向量所表征的监测数据来自的零部件或功能系统相同的之前存储的特征向量。
第二方面提供了一种基于云的车辆故障诊断装置，包括：监测数据接收模块、数据预处理模块、特征数据库、故障诊断模块；所述监测数据接收模块用于接收车辆上传的监测数据,所述监测数据为所述车辆监测到的零部件或功能系统的工作状态数据；所述数据预处理模块用于从所述监测数据接收模块接收到的监测数据中提取所述监测数据的特征向量,所述特征向量为表征所述监测数据的一组数；所述特征数据库用于以所述监测数据所来自的所述车辆的零部件或功能系统为标签，将所述数据预处理模块提取的特征向量分类存储；所述故障诊断模块用于基于支持向量机算法，对所述特征数据库分类存储的所述特征向量并行地进行故障诊断。
结合第二方面，在第二方面的第一种可能实现的方式中，该装置还包括：中央数据库；所述中央数据库用于：将所述监测数据接收模块接收到的监测数据进行解析，得到解析后的监测数据；以所述监测数据所来自的所述车辆的零部件或功能系统为标签，将所述解析后的监测数据分类存储；其中，针对所述解析后的监测数据的分类存储的标签与针对所述特征向量的分类存储的标签相对应；所述数据预处理模块具体用于：从所述中央数据库解析后的监测数据中提取所述解析后的监测数据的特征向量。
结合第二方面或第二方面的第一种可能的实现方式，在第二方面的第二种可能的实现方式中，特征数据库还用于：定期删除与最近提取的特征向量所表征的监测数据来自的零部件或功能系统相同的的之前存储的特征向量。
结合第二方面的第二种可能的实现方式，在第二方面的第三种可能的实现方式中，中央数据库删除与特征数据库删除的特征向量相对应的监测数据。
第三方面提供了一种基于云的车辆故障诊断系统，该系统包括：第二方面或第二方面的第一种实现方式或第二方面的第二中实现方式中的装置、车辆；车辆将其监测到的数据上传至第二方面或第二方面的第一种实现方式或第二方面的第二中实现方式中的装置；第二方面或第二方面的第一种实现方式或第二方面的第二中实现方式中的装置根据接收到的数据进行故障诊断。
本发明实施例通过基于支持向量机算法，将来自不同零部件或功能系统的监测数据的特征向量分类存储，并且将分类存储的特征向量并行地进行故障诊断,不仅可以缩短诊断时间，还可以避免数据传递过程中的不同数据的彼此影响，提高故障诊断的准确性。
附图说明
图1：本发明实施例给出的一种基于云的故障诊断系统示意图；
图2：本发明实施例给出的一种云端诊断装置示意图；
图3：本发明实施例给出的一种中央数据库示意图；
图4：本发明实施例给出的一种中央数据库的定期存储单元示意图；
图5：本发明实施例给出的一种特征数据库示意图；
图6：本发明实施例给出的一种特征数据库的定期存储单元的示意图；
图7：本发明实施例给出的一种基于云的故障诊断方法的流程图；
图8：本发明实施例给出的一种并行计算的示意图；
图9：本发明实施例给出的一种离线训练以及测试故障分类模型方法的流程图；
图10：本发明实施例给出的一种系统降级决策机内部决策方法的流程图；
图11：本发明实施例给出的一种电池包电流传感器故障诊断方法的流程图。
具体实施方式
下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
本发明实施例可能涉及的术语缩略语及定义如下：
ADAS  Advanced Driver Assistant System  高级驾驶辅助系统
VCU   Vehicle Control Unit              整车控制单元
OBC   On Board Charger                  车载充电机
SVM   Support Vector Machine            支持向量机
GA    Genetic Algorithm                 遗传算法
PSO   Particle Swarm Optimization       粒子群优化
DDAG  Decision Directed Acyclic Graph   决策导向无环图
本发明实施例提供了一种基于云的车辆故障诊断系统，可用于实时地或/和在线诊断车辆故障、统计车辆功能系统/零部件故障数据等，如图1所示，该系统包括以下几个组成部分：云端诊断装置1000、车辆2000、整车或零部件制造商3000、维修服务商4000、其他设备5000。
其中，云端诊断装置1000通过无线通信技术与车辆2000、整车或零部件制造商3000、维修服务商4000、其他设备5000实现数据交互；可选的，该系统不对无线通信技术作任何限制，可以是任何协议下的无线通信技术的一种或多种。云端诊断装置1000可以基于车辆2000上传的监测数据实时诊断与定位车辆发生的故障，并且管理、统计故障数据；进一步地，可以将处理或统计后的故障数据发送给对应的车辆2000、整车或零部件制造商3000、维修服务商4000或其他设备5000；比如，零部件制造商A想知道其生产的零部件A故障情况，云端诊断装置1000可以将统计获得的零部件A的故障数据发送给零部件 制造商A，该零部件A的故障数据包括但不限于零部件A发生故障的车辆数量、某一车辆的零部件A发生故障的次数等；再比如，车辆B想知道其实时运行的故障情况，云端诊断装置1000可以将根据车辆B实时上传的监测到的数据获得车辆B的故障数据发送给车辆B，该车辆B的故障数据包括但不限于车辆B的安全系数、某一零部件的故障提示等。
车辆2000用于指示与云端诊断装置1000通过无线通信技术互联的一辆或多辆汽车，并不特指行驶中的某一车辆；车辆2000配置有监测感应装置用于监测车辆运行数据或是零部件的运行数据，可以根据故障诊断需求或指令设置将监测到的相关数据上传至云端诊断装置1000，由云端诊断装置1000进一步地处理车辆2000上传的监测数据。
整车或零部件制造商3000、维修服务商4000、其他设备5000不是该系统的必要的组成部分，它们基于各自的需求从云端诊断装置1000获得/接收与故障相关的数据，用于分析某一故障发生的概率、频率、对整车/功能系统/零部件的影响等。
本发明实施例通过将车辆故障诊断环节移到云端故障诊断装置，既可以解决单一车辆的单片机计算能力的局限性，提高故障诊断的准确率；又可以基于云端实现统一管理众多车辆的众多故障，并将获得数据与整车/零部件制造商、维修服务商、其他设备(例如第三方监控设备)共享，用于从源头解决故障，提高车辆/零部件的安全性，保障车辆行驶安全。需要指出的是：该基于云端诊断系统不限于车辆/零部件故障的诊断和管理，同样适用于船、飞机、火车、无人机等故障的诊断和管理。
本发明实施例提供了一种云端诊断装置，如图2所示，该装置1000包括：监测数据接收模块1010、中央数据库1020、数据预处理模块1030、特征数据库1040、故障诊断模块1050、故障等级判定模块1060、系统降级决策机1070、故障统计模块1080。
监测数据接收模块1010用于接收车辆上传的监测数据，其中，监测数据是指车辆监测到的零部件或功能系统的工作状态数据，可选的，监测数据是车辆监测到的整车或零部件或功能系统的工作状态的相关数据；
中央数据库1020用于将监测数据接收模块1010接收到的监测数据进行解析，得到解析后的监测数据；，具体的为解析车辆上传的数据包,并且将解析后的数据输入数据预处理模块1030；可选的，中央数据库还用于：以监测数据来自的零部件或功能系统为标签，将解析后的监测数据分类存储，进而能够建立一个比较完备的有关监测数据的数据库，该数据库可以用于后期分析某一故障出现后对整车/零部件寿命的影响、或者用于完善故障诊断系统。
进一步地，如图3所示，中央数据库1020包括：临时存储单元1021、定期储存单元1022；可选的，临时存储单元1021和定期存储单元1022以零部件(零部件1，零部件2，…，零部件n，系统)为类别进行结构化分类存储管理；临时存储单元1021用于临时存储车辆上传的实时数据；定期存储单元1022用于存储以故障诊断模块1050输出的诊断结果作为故障标签，从临时存储单元1021存储的实时数据中转移过来的数据。如图4所示，以零部件i为例描述定期存储单元1022结构化分类存储管理，零部件i有一个相对独立的存储区域，该存储区域分为无故障数据区域和故障数据区域，故障数据区域又可以细分为传感器类数据、执行器类故障数据或其它故障数据；对应每类故障，可以进一步地详细分类，从故障1到故障n，例如传感器故障类数据可以分为电流传感器故障数据、电压传感器故障数据、温度传感器故障数据、以及压力传感器故障数据等。可选的，为了节省存储 空间，定期存储单元1022可以对所存储的数据进行定期清洗(如每周、每月或每年)。
数据预处理模块1030用于对中央数据库1020输入的监测数据(也称之为原始数据)进行特征向量的提取和降维处理，以缩小数据量和提取有效的数据特征向量，可以缩短故障诊断时间和提高故障诊断的准确率；其中，特征向量为表征所述监测数据的一组数，可选的，对监测数据进行平均或方差计算，得到平均值或方差值，对应监测数据的平均值或方差值的集合可以看做是一组数，；可选的，一组数可以表示为{A，B，C，D…Z}。
特征数据库1040用于存储数据预处理模块处理后获得的特征向量；进一步地，如图5所示，特征数据库1040包括：临时存储单元1041、定期储存单元1042。特征数据库1040的存储管理与中央数据库1020的结构化分类存储管理类似，也是以零部件为类别和以故障诊断模块1050输出的诊断结果为标签进行结构化分类存储管理，如图6所示，具体描述可以参见中央数据库1020描述的结构化分类存储管理。需要明确的是，特征数据库1040存储到特征向量数据与黄宗洋数据库存储的实时数据(也称之为原始数据)是意义对应的。可选的，定期对特征数据库1040存储到额特征向量数据进行清洗；定期清洗的原则是：(1)特征向量相似的只需要保留最新的特征向量；(2)对应地，中央数据库1020的定期存储单元1022只需要保留与特征数据库1040保留的特征向量相对应的实时数据(也称之为原始数据)。
故障诊断模块1050用于基于机器学习算法对分类存储的特征向量并行的进行故障诊断。可选的，故障诊断模块1050所使用到的机器学习算法为基于决策导向无环图DDAG的支持向量机算法。具体的故障诊断在下面实施例进行描述，在此不再赘述。
故障等级判定模块1060用于将故障诊断模块1050输出的故障诊断结果划分等级。可选的，等级划分为：1级故障(最严重)，2级故障，3级故障…；可选的，等级划分为：严重故障、中等故障、一般故障。
系统降级决策机1070用于根据故障等级判定模块1060判定的故障等级或/和将故障诊断结果的相关数据输入模型后进行决策，将超出预期安全状态或影响整车安全状态的故障通过相应的危险警示信号通知发生该故障的车辆，例如：系统降级决策机1070根据刹车系统故障诊断结果的相关数据认为刹车系统发生的故障已超出预期安全状态，则通过危险警示信号提示发生该故障的车辆；进一步地，请求车辆保证安全的情况下停车以及提示车辆尽快进行维修。
故障统计模块1080用于接收经故障等级判定模块1060输出的故障诊断结果，并且根据零部件/系统进行分区管理、统计；进一步地，故障统计模块1080具体用于下面内容的一种或多种但又不限于下面的例举：统计每个零部件发生故障的概率、统计每个零部件中的每类故障发生的概率、统计每个零部件内的某个故障发生的概率、统计所有零部件不同等级故障发生的概率等；可选的，统计的周期可以任意时间段，例如一年、三月、一月、n周、n日等；可选的，故障统计模块1080用于将故障统计的结果发送给相应的整车或零部件制造商、维修服务商、其他设备。
本发明实施例提供的基于云的车辆故障诊断装置，能够通过基于支持向量机算法，将从不同零部件或功能系统的监测数据所提取的特征向量并行地进行故障诊断,不仅可以缩短诊断时间，还可以避免数据传递过程中的不同数据的彼此影响,提高故障诊断的准确性。
本发明实施例提供了一种基于云的车辆故障诊断方法，如图7所示，该方法的具体步 骤为：
S100：车辆将监测到的功能系统/零部件的监测数据上传至云端诊断装置/系统；可选的，车辆将监测数据直接上传至中央数据库；可选的，以数据包的形式将所监测到的数据打包上传至中央数据库；可选的，车辆将监测到的功能系统/零部件的监测数据直接上传至数据预处理模块；
S200：中央数据库接收到车辆上传的监测数据，并且将该监测数据进行解析后传输给数据预处理模块；可选的，中央数据库还将接收到的或解析后的监测数据进行结构化分类存储管理，具体可以参照上述实施例的描述，在此不再赘述；
S300：数据预处理模块接收中央数据库传输的解析后的监测数据，并且对所接收到的监测数据进行故障特征提取以获得特征向量及将提取后的特征向量传输给特征数据库；进一步地，数据预处理模块通过小波包分解对接收到的数据进行故障特征提取得到特征向量，再通过核主元分析对提取的特征向量进行降维处理，以得到降维后的特征向量；其中，小波包算法可以在全频带对信号进行多层次的频带划分，因此对于故障特征提取的完备度较高；可选的，为了降低故障诊断分类器的计算复杂度，提高故障分离的精度，可以使用径向基-核主元分析算法对所提取的特征向量进行特征选择，降维处理；可选的，数据预处理模块接收车辆直接上传的实时数据；
可选的，经过上述处理获得的特征向量也可以直接传输给故障诊断模块进行故障诊断与定位；
S400：特征数据库接收数据预处理模块传输的特征向量，并且对接收到的特征向量数据采取结构化分类存储管理，具体具体可以参照上述实施例的描述，在此不再赘述；进一步地，特征数据模块将特征向量传输给故障诊断模块；值得注意的，特征数据库不是必要的模块，该模块的作用是为了更好地管理特征向量；
S500：故障诊断模块接收特征数据库传输的特征向量，基于机器学习算法进行故障的实时诊断与定位；可选的，机器学习算法为基于决策向导无环图DDAG的支持向量机算法。如图8所示，可选的，故障诊断模块包含有一个或多个故障诊断单元，故障诊断单元可以是对应零部件/功能系统配置，对输入的特征向量根据零部件/功能系统划分进行并行计算以分别获得诊断结果，并行计算可以缩短故障诊断的时间和提高故障诊断的准确率；其中，对应零部件的故障诊断单元可以对相应的零部件及其内部元件进行故障诊断和定位；对应功能系统的故障诊断单元可以对整个系统或则某一个功能系统进行故障诊断和定位，能够避免数据传递过程对诊断影响而导致的误判；
进一步地，基于支持向量机的故障诊断需要离线构造故障分类器，也就是所说的支持向量机训练模型；在训练支持向量机模型时，使用粒子群优化PSO算法优化支持向量机的惩罚因子参数和径向基核函数参数，以提高故障诊断的准确率。下面提供了离线训练以及测试故障分类模型的方法，如图9所示，具体步骤如下：
(1)将车辆动力系统或关键部件如发动机、驱动电机、高压电池系统、逆变器、DCDC、OBC、自动驾驶或辅助驾驶系统，按照传感器故障、执行器故障以及其它故障，通过仿真、台架或实车路试采集不同部件发生不同故障时的数据；
(2)对采集到的数据使用小波包和核主元算法进行数据的特征提取和降维处理，并最终输出不同的特征向量；
(3)使用支持向量机训练不同故障的特征向量，并使用PSO优化支持向量机的结构参数-惩罚因子参数和径向基核函数参数；
(4)设置支持向量机的结构参数的初始值以及搜索范围、PSO的参数值；
(5)随机给定一组粒子的速度和位置；
(6)训练支持向量机分类模型，并计算每个粒子的适应度函数值；
(7)更新粒子位置的个体和全局最优值；
(8)更新粒子的速度和位置；
(9)判断是否达到最大迭代次数；若达到最大迭代次数，则获得最优的支持向量机结构参数-惩罚因子参数和径向基核函数参数值；
(10)获得最优的支持向量机分类模型；
(11)使用测试数据去测试获得的支持向量机分类模型，并输出故障诊断准确率，以判定分类模型的准确度；并将测试后的支持向量机分类模型用于在线实时故障诊断。
本发明实施例提供了的一种基于云的故障诊断方法，该方法通过基于支持向量机算法，将不同零部件/系统的特征向量并行地进行故障诊断,不仅可以缩短诊断时间，还可以避免数据传递的影响以提高故障诊断的准确性。
可选的，S600：故障诊断模块将诊断结果分别传输给中央数据库、特征数据库和故障等级判定模块；进一步地，中央数据库和特征数据库接收诊断结果用于以诊断结果为标签对相应数据进行分类管理，具体的分类管理参见前面实施例，在此不再赘述；
可选的，S700：故障等级判定模块对接收到的诊断结果进行等级判定，具体等级划分参见上面实施例描述，在此不再赘述；进一步地，将判定等级的诊断结果传输给系统降级决策机、故障统计模块；
可选的，S800：故障统计模块对接收到的诊断结果进行数据统计，具体的统计工作参考上面实施例的描述，在此不再赘述；进一步地，故障统计模块将统计的数据发送给制造商、服务商等，用于完善产品和/或服务；
可选的，S900：系统降级决策机接收故障等级判定模块传输的经过等级判定的诊断结果，并且基于故障等级判定模块输出的结果决策，向车辆发送相应危险警示信号；系统降级决策机内部决策及发送危险警示信号的控制过程如图10所示，具体如下：
(1)系统降级决策机会根据故障等级判定模块输出的结果，进行决策。如果被诊断出的故障在故障等级判定模块判定为严重等级的故障，此时系统降级决策机会直接发送相应的危险警示信号给车辆；可选的，危险警示信号是提示车辆发生了严重故障，或者/和请求紧急处理该严重故障；可选的，严重等级故障是指如果发生很可能导致车辆失去控制，危害到驾驶员的生命安全；例如，车辆制动踏板出现故障，则车俩制动性能不能保证，可归为最严重故障；
(2)如果所发生的故障在故障等级判定模块判定为非严重等级的故障，系统降级决策机会根据车辆上传的实时数据进行综合判断：
一方面，车辆上传的整车系统数据会输入到整车模型，判定整车是否处于安全状态；另一方面，上传的零部件数据会输入到相应零部件模型，判定部件是否处于一种预期的安全状态；进一步地，如果整车或某零部件判定其处于临界危险状态，系统降级决策机会发送相应的危险警示信号给车辆。其中，整车模型、零部件模型可以通过数学公式构建，或 者通过审计网络等智能算法训练；在具体实现中，可以基于系统(整车或部件)的测量输入/输出信号以及系统的模型计算出反应系统安全状态的重要参数值，如果参数值超出预期的安全范围，系统降级决策机会判定此时车辆超出预期的安全状态，会发送相应的系统危险警示信号给车辆，可实时分析车辆的安全状态，保证车辆安全行驶。
下面以电池包电流传感器故障为例，详细描述故障诊断与定位的过程，如图11所示，具体实现过程如下：
(1)车辆将监测到的关于电池包电流传感器的实时数据传输给中央数据库的临时存储单元，以及数据预处理模块；
(2)数据预处理模块通过对接收到的数据进行特征提取操作，将提取后的特征向量发送给故障诊断模块；进一步地，发送给电池故障诊断单元；并且，数据预处理模块还会将特征向量存储到特征数据库，特征数据库的结构及存储方式可以参考中央数据库；具体地，将电池相关的特征向量存储到特征数据库的电池区域；
(3)电池故障诊断单元判定电池包电流传感器发生故障，并将此诊断结果发送到故障统计模块、故障等级判定模块、中央数据库、特征数据库；
(4)此时，中央数据库会以电池包电流传感器故障为故障标签，将临时存储单元的存储的关于电池的实时数据转移到定期存储单元-电池区域-故障数据区域-传感器故障区域-电流传感器故障区域中；特征数据库的也采取类似的方法将临时存储单元的特征向量转移至定期存储单元；
(5)同时，故障诊断结果也会发送给整车以及第三方的监控设备(如手机)；
(6)同时，故障统计模块会统计电池包电流传感器故障发生的概率，并将结果发送给整车或零部件厂，以及厂商维修(4S)店；
(7)同时，故障严重等级判定模块中会判定此故障是否为严重故障，如果判定为严重故障则将此诊断结果发送给系统降级决策机；
(8)降级决策机基于该严重故障作出降级命令决策，并发送相应的危险警示信号给整车的VCU，请求降功率至零，并切断高压；同时，提醒驾驶员发生动力系统故障，需靠边停车。
本发明实施例提供的电池包电流传感器故障诊断方法，通过将电池包电流传感器的监测数据提取后调整向量存储到相应的分类存储，由电池故障诊断单元根据特征向量判定电池包电流传感器发生故障，可以避免其他零部件/系统的监测数据对电池包电流传感器监测数据的影响，提高故障诊断的准确性。
最后应说明的是：以上各实施例仅用以说明本发明的技术方案，而非对其限制；尽管参照前述各实施例对本发明进行了详细的说明，本领域的普通技术人员应当理解：其依然可以对前述各实施例所记载的技术方案进行修改，或者对其中部分或者全部技术特征进行等同替换；而这些修改或者替换，并不使相应技术方案的本质脱离本发明各实施例技术方案的范围。

Claims (16)

  1. 一种基于云的车辆故障诊断方法，其特征在于，包括：
    接收车辆上传的监测数据,所述监测数据为所述车辆监测到的零部件或功能系统的工作状态数据；
    从所述监测数据中提取所述监测数据的特征向量,所述特征向量为表征所述监测数据的一组数；
    以所述监测数据所来自的所述车辆的零部件或功能系统为标签，将所述监测数据的特征向量分类存储；
    基于支持向量机算法，对分类存储的所述特征向量并行地进行故障诊断。
  2. 如权利要求1所述的方法，其特征在于，在所述从所述监测数据中提取所述监测数据的特征向量之前，还包括：
    将所述监测数据进行解析，得到解析后的监测数据；
    以所述监测数据所来自的所述车辆的零部件或功能系统为标签，将所述解析后的监测数据分类存储；其中，针对所述解析后的监测数据的分类存储的标签与针对所述特征向量的分类存储的标签相对应；
    所述从所述监测数据中提取所述监测数据的特征向量具体包括：
    从所述解析后的监测数据中提取所述解析后的监测数据的特征向量。
  3. 如权利要求1或2所述的方法，其特征在于，所述基于支持向量机算法，对分类存储的所述特征向量并行地进行故障诊断包括：
    基于决策导向无环图DDAG构造故障分类器，对分类存储的所述特征向量并行地计算诊断结果，所述诊断结果至少包括故障发生的位置。
  4. 如权利要求1所述的方法，其特征在于，所述从所述监测数据中提取所述监测数据的特征向量包括：
    通过小波包分解从所述监测数据中提取所述监测数据的特征向量。
  5. 如权利要求4所述的方法，其特征在于，在所述以所述监测数据所来自的所述车辆的零部件或功能系统为标签,将所述监测数据的特征向量分类存储之前，还包括：
    通过核主元分析对所述特征向量进行降维处理，得到降维处理后的特征向量；
    所述以所述监测数据所来自的所述车辆的零部件或功能系统为标签,将所述监测数据的特征向量分类存储具体包括：
    以所述监测数据所来自的所述车辆的零部件或功能系统为标签,将所述降维处理后的特征向量分类存储；
    所述基于支持向量机算法，对分类存储的所述特征向量并行地进行故障诊断具体包括：
    基于支持向量机算法，对分类存储的所述降维处理后特征向量并行地进行故障诊断。
  6. 如权利要求3所述的方法，其特征在于，还包括：
    判定所述诊断结果所指示故障的严重等级，所述严重等级划分为：严重故障、中等故障、一般故障。
  7. 如权利要求6所述的方法，其特征在于，在所述判定所述诊断结果所指示故障的严重等级之后，还包括：
    如果所述严重等级为严重故障，则向所述车辆发送危险警示信号，所述危险警示信号用于提示所述车辆正在发生危及车辆正常行驶的严重故障。
  8. 一种基于云的车辆故障诊断装置，其特征在于，包括：监测数据接收模块、数据预处理模块、特征数据库、故障诊断模块；
    所述监测数据接收模块用于接收车辆上传的监测数据,所述监测数据为所述车辆监测到的零部件或功能系统的工作状态数据；
    所述数据预处理模块用于从所述监测数据接收模块接收到的监测数据中提取所述监测数据的特征向量,所述特征向量为表征所述监测数据的一组数；
    所述特征数据库用于以所述监测数据所来自的所述车辆的零部件或功能系统为标签，将所述数据预处理模块提取的特征向量分类存储；
    所述故障诊断模块用于基于支持向量机算法，对所述特征数据库分类存储的所述特征向量并行地进行故障诊断。
  9. 如权利要求8所述的装置，其特征在于，还包括：中央数据库；
    所述中央数据库用于：将所述监测数据接收模块接收到的监测数据进行解析，得到解析后的监测数据；
    以所述监测数据所来自的所述车辆的零部件或功能系统为标签，将所述解析后的监测数据分类存储；其中，针对所述解析后的监测数据的分类存储的标签与针对所述特征向量的分类存储的标签相对应；
    所述数据预处理模块具体用于：从所述中央数据库解析后的监测数据中提取所述解析后的监测数据的特征向量。
  10. 如权利要求8或9所述的装置，其特征在于，所述故障诊断模块具体用于：
    基于决策导向无环图DDAG构造故障分类器，对所述特征数据库分类存储的特征向量并行地计算诊断结果，所述诊断结果至少包括故障发生的位置。
  11. 如权利要求8所述的装置，其特征在于，所述数据预处理模块具体用于：
    通过小波包分解从所述监测数据中提取所述特征向量。
  12. 如权利要求11所述的装置，其特征在于，所述数据预处理模块具体还用于：
    通过核主元分析对提取的所述特征向量进行降维处理，得到降维处理后的特征向量；
    所述特征数据库具体用于：以所述监测数据所来自的所述车辆的零部件或功能系统为标签,将所述数据预处理模块降维处理后的特征向量分类存储；
    所述故障诊断模块具体用于：基于支持向量机算法，对所述特征数据库分类存储的降维处理后特征向量并行地进行故障诊断。
  13. 如权利要求10所述的装置，其特征在于，还包括：故障等级判定模块；
    所述故障等级判定模块用于判定所述故障诊断模块输出的诊断结果所指示故障的严重等级，所述严重等级划分为：严重故障、中等故障、一般故障。
  14. 如权利要求13所述的装置，其特征在于，还包括：系统降级决策机；
    所述系统降级决策机用于如果所述故障等级判定模块判定的严重等级为严重故 障，则向所述车辆发送危险警示信号，所述危险警示信号用于提示所述车辆正在发生可能危及车辆正常行驶的严重故障。
  15. 一种基于云的车辆故障诊断系统，其特征在于，包括：权利要求8-14任选一所述的装置、车辆；
    所述车辆将监测数据上传至所述权利要求8-14任选一所述的装置；
    所述权利要求8-14任选一所述的装置根据所述监测数据进行故障诊断。
  16. 如权利要求15所述的系统，其特征在于，还包括：
    如果故障诊断出为严重故障，则向所述实时运行车辆发送危险警示信号，所述危险警示信号用于提示所述车辆正在发生可能危及车辆正常行驶的严重故障；
    所述实时运行车辆根据所述危险警示信号作出故障应对措施。
PCT/CN2017/097124 2017-01-13 2017-08-11 一种基于云的车辆故障诊断方法、装置及其系统 WO2018129917A1 (zh)

Priority Applications (6)

Application Number Priority Date Filing Date Title
EP22177467.2A EP4119919A1 (en) 2017-01-13 2017-08-11 Cloud-based vehicle fault diagnosis method, apparatus, and system
EP17890940.4A EP3564647B1 (en) 2017-01-13 2017-08-11 Cloud-based vehicle fault diagnosis method, device and system thereof
JP2019538380A JP6830540B2 (ja) 2017-01-13 2017-08-11 クラウドベースの車両故障診断方法、装置およびシステム
KR1020197023618A KR102263337B1 (ko) 2017-01-13 2017-08-11 클라우드 기반 차량 고장 진단 방법, 장치 및 시스템
US16/510,289 US11468715B2 (en) 2017-01-13 2019-07-12 Cloud-based vehicle fault diagnosis method, apparatus, and system
US17/939,438 US12086165B2 (en) 2017-01-13 2022-09-07 Cloud-based vehicle fault diagnosis method, apparatus, and system

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201710025101.8A CN108303264B (zh) 2017-01-13 2017-01-13 一种基于云的车辆故障诊断方法、装置及其系统
CN201710025101.8 2017-01-13

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US16/510,289 Continuation US11468715B2 (en) 2017-01-13 2019-07-12 Cloud-based vehicle fault diagnosis method, apparatus, and system

Publications (1)

Publication Number Publication Date
WO2018129917A1 true WO2018129917A1 (zh) 2018-07-19

Family

ID=62839111

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/097124 WO2018129917A1 (zh) 2017-01-13 2017-08-11 一种基于云的车辆故障诊断方法、装置及其系统

Country Status (6)

Country Link
US (2) US11468715B2 (zh)
EP (2) EP3564647B1 (zh)
JP (1) JP6830540B2 (zh)
KR (1) KR102263337B1 (zh)
CN (1) CN108303264B (zh)
WO (1) WO2018129917A1 (zh)

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109241849A (zh) * 2018-08-07 2019-01-18 浙江大学 面向智能电厂汽轮机主机的经验模态和小波分解及特征选择的故障诊断方法
CN111459816A (zh) * 2020-03-31 2020-07-28 北京百度网讯科技有限公司 故障注入测试方法、装置、系统及存储介质
CN111873978A (zh) * 2020-07-10 2020-11-03 中车齐齐哈尔车辆有限公司 列车管压力的监测方法和系统
CN112327190A (zh) * 2020-10-14 2021-02-05 北方工业大学 一种储能电池健康状态辨识方法
JP2021044231A (ja) * 2019-09-11 2021-03-18 コリア インスティテュート オブ エナジー リサーチＫｏｒｅａ Ｉｎｓｔｉｔｕｔｅ Ｏｆ Ｅｎｅｒｇｙ Ｒｅｓｅａｒｃｈ 燃料電池システムの多段階故障診断方法及び装置
CN113536658A (zh) * 2021-05-21 2021-10-22 西北工业大学 基于stm32嵌入式处理器的机电设备轻量化故障诊断方法
CN115373369A (zh) * 2022-08-24 2022-11-22 中国第一汽车股份有限公司 车辆故障诊断系统及方法
CN115810227A (zh) * 2022-12-01 2023-03-17 深圳瑞为智能科技有限公司 一种车载设备异常快速识别方法及系统

Families Citing this family (40)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP6992889B2 (ja) * 2018-05-14 2022-01-13 富士通株式会社 判定ルール取得装置、判定ルール取得方法および判定ルール取得プログラム
CN112041819B (zh) * 2018-06-29 2022-08-05 罗伯特·博世有限公司 用于监测和识别电驱动系统中的传感器故障的方法
CN110794802B (zh) * 2018-08-01 2023-05-23 上海汽车集团股份有限公司 一种故障诊断处理方法及装置
CN109241915B (zh) * 2018-09-11 2022-03-25 浙江大学 基于振动信号平稳非平稳性判别及特征甄别的智能电厂泵机故障诊断方法
CN109353376A (zh) * 2018-10-24 2019-02-19 西安英特迈思信息科技有限公司 轨道车辆监测系统及其监测方法
US20200224601A1 (en) * 2019-01-15 2020-07-16 GM Global Technology Operations LLC Method of diagnosing a propulsion system of a vehicle, and a system therefor
CN111624973A (zh) * 2019-02-28 2020-09-04 北京新能源汽车股份有限公司 一种故障信号分级测试方法、装置及汽车
US10922906B2 (en) * 2019-03-28 2021-02-16 GM Global Technology Operations LLC Monitoring and diagnosing vehicle system problems using machine learning classifiers
CN111846095B (zh) * 2019-05-14 2022-05-17 北京骑胜科技有限公司 一种故障检测设备、电助力车及故障检测方法
CN111915026B (zh) * 2019-06-10 2024-08-02 中车大同电力机车有限公司 故障处理方法、装置、电子设备及存储介质
US12061971B2 (en) 2019-08-12 2024-08-13 Micron Technology, Inc. Predictive maintenance of automotive engines
CN110351391A (zh) * 2019-08-16 2019-10-18 深圳市道通科技股份有限公司 一种汽车诊断云平台系统、业务实现方法
US11702086B2 (en) * 2019-08-21 2023-07-18 Micron Technology, Inc. Intelligent recording of errant vehicle behaviors
US11262739B2 (en) * 2019-09-05 2022-03-01 Hitachi Energy Switzerland Ag Artificial intelligence/machine learning driven assessment system for a community of electrical equipment users
CN112987711B (zh) * 2019-11-30 2022-08-09 华为技术有限公司 自动驾驶规控算法优化方法及仿真测试装置
CN111105522B (zh) * 2019-12-10 2022-02-22 河南嘉晨智能控制股份有限公司 一种车辆健康预测系统及方法
KR102268053B1 (ko) * 2019-12-16 2021-06-22 주식회사 브이웨이 클라우드 서버 및 로보틱 프로세스 자동화 기술 기반의 차량용 지능형 안전분석 시스템
CN111157898A (zh) * 2020-01-07 2020-05-15 清华大学深圳国际研究生院 一种新能源车辆在线电池故障检测分析方法及装置
CN113555591A (zh) * 2020-04-24 2021-10-26 罗伯特·博世有限公司 燃料电池诊断装置以及相应的诊断方法、车辆和车辆系统
JP7491746B2 (ja) * 2020-06-10 2024-05-28 株式会社日立製作所 分散システム
CN111678709B (zh) * 2020-06-16 2021-04-30 山东大学 基于云平台的电动汽车分布式网络化测试系统及方法
CN111813074B (zh) * 2020-06-16 2021-09-03 中国人民解放军国防科技大学 基于故障噪声比及特征提取的微小故障诊断方法及系统
CN111968726B (zh) * 2020-08-17 2024-10-01 北京赛迈特锐医疗科技有限公司 序贯型ai诊断模型临床应用调度管理系统及其方法
CN112100239A (zh) * 2020-09-11 2020-12-18 深圳市道通科技股份有限公司 车辆检测设备画像生成方法、装置、服务器及可读存储介质
CN112165528B (zh) * 2020-09-30 2022-03-11 重庆长安汽车股份有限公司 车辆事件及其事件文件数据的管理方法、系统及存储介质
US20220111836A1 (en) * 2020-10-09 2022-04-14 Nec Laboratories America, Inc. Modular network based knowledge sharing for multiple entities
CN112379659A (zh) * 2020-11-12 2021-02-19 西安石油大学 一种石油钻机故障预测系统
KR20220108250A (ko) * 2021-01-25 2022-08-03 현대자동차주식회사 환경차의 고장 제어 시스템
KR102286272B1 (ko) 2021-02-02 2021-08-06 주식회사 파트리지시스템즈 이벤트 검출 장치 및 그 방법
CN113156913B (zh) * 2021-02-05 2024-08-06 深圳大雷汽车检测股份有限公司 一种abs故障诊断系统及方法
CN112987687B (zh) * 2021-02-08 2022-05-24 南京航空航天大学 一种云-端融合的智能线控底盘健康监测系统及方法
CN113110401B (zh) * 2021-05-21 2022-08-05 广东美房智高机器人有限公司 一种智能生成机器人故障解决方案的方法
CN113392936B (zh) * 2021-07-09 2022-09-02 四川英创力电子科技股份有限公司 一种基于机器学习的烤箱故障诊断方法
CN113723177A (zh) * 2021-07-19 2021-11-30 江铃汽车股份有限公司 汽车故障诊断方法、系统、终端设备及可读存储介质
JP7484847B2 (ja) 2021-08-23 2024-05-16 トヨタ自動車株式会社 自動運転車制御装置
CN113867322A (zh) * 2021-10-25 2021-12-31 武汉极目智能技术有限公司 高级辅助驾驶系统安装质量的在线监控方法及监控系统
KR102526059B1 (ko) 2021-11-25 2023-04-27 주식회사 파트리지시스템즈 데이터 취득 조절을 위한 시스템
CN114710479B (zh) * 2022-03-31 2023-12-01 三一电动车科技有限公司 故障诊断方法、装置及系统
KR102560856B1 (ko) 2022-06-22 2023-07-28 케이지모빌리티 주식회사 차량 전체수명주기 동안의 건강상태 진단방법
KR20240131186A (ko) 2023-02-23 2024-08-30 에이치엘만도 주식회사 차량상태 진단제어 시스템 및 방법

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2001296915A (ja) * 2000-04-13 2001-10-26 Yazaki Corp 遠隔自己診断システム
CN103455026A (zh) * 2013-08-23 2013-12-18 王绍兰 一种车辆故障诊断和预警方法及装置
CN104133467A (zh) * 2014-07-30 2014-11-05 浪潮集团有限公司 一种基于云计算的obds远程故障诊断恢复系统
CN105137963A (zh) * 2015-09-09 2015-12-09 山东丽驰新能源汽车有限公司 车辆远程监控与故障诊断方法及装置
CN105867351A (zh) * 2016-04-29 2016-08-17 大连楼兰科技股份有限公司 车辆故障码实时采集与历史数据分析诊断的方法及装置
CN106054867A (zh) * 2016-07-28 2016-10-26 重庆峰创科技有限公司 一种基于物联网与云计算的新能源汽车故障诊断装置
CN106327344A (zh) * 2016-08-28 2017-01-11 华南理工大学 一种基于车联网的车辆故障在线检测预警装置和方法

Family Cites Families (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2005309077A (ja) 2004-04-21 2005-11-04 Fuji Xerox Co Ltd 故障診断方法および故障診断装置、並びに搬送装置および画像形成装置、並びにプログラムおよび記憶媒体
JP2006315813A (ja) * 2005-05-13 2006-11-24 Murata Mach Ltd 移動体の診断システム
JP2007161044A (ja) 2005-12-13 2007-06-28 Toyota Motor Corp 車両故障診断装置及び方法
CN100470417C (zh) * 2006-12-22 2009-03-18 浙江大学 工业生产过程小样本条件下的故障诊断系统及方法
US8301333B2 (en) * 2010-03-24 2012-10-30 GM Global Technology Operations LLC Event-driven fault diagnosis framework for automotive systems
CN102752360B (zh) 2012-03-01 2015-07-08 浙江吉利汽车研究院有限公司 一种基于云计算的汽车故障检测系统
JP6003810B2 (ja) * 2013-06-04 2016-10-05 株式会社デンソー 車両用基準値生成装置
KR101498368B1 (ko) * 2013-06-05 2015-03-05 주식회사 웨이브엠 위험물 차량 관제 시스템 및 그 방법
US20160035152A1 (en) * 2013-12-31 2016-02-04 Agnik, Llc Vehicle data mining based on vehicle onboard analysis and cloud-based distributed data stream mining algorithm
CN104021238A (zh) 2014-03-25 2014-09-03 重庆邮电大学 一种铅酸动力电池系统故障诊断方法
CN105095566B (zh) 2015-06-29 2019-06-04 南京航空航天大学 一种基于小波分析和svm的逆变器故障诊断方法
CN105510051A (zh) 2016-01-17 2016-04-20 蒋和平 基于物联网与云计算的新能源汽车故障诊断系统
US10728101B2 (en) * 2016-05-01 2020-07-28 Argus Cyber Security Ltd. In-vehicle network anomaly detection
WO2018017467A1 (en) * 2016-07-18 2018-01-25 NantOmics, Inc. Distributed machine learning systems, apparatus, and methods
JP6105141B1 (ja) * 2016-09-30 2017-03-29 株式会社日立パワーソリューションズ プリプロセッサおよび診断装置

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2001296915A (ja) * 2000-04-13 2001-10-26 Yazaki Corp 遠隔自己診断システム
CN103455026A (zh) * 2013-08-23 2013-12-18 王绍兰 一种车辆故障诊断和预警方法及装置
CN104133467A (zh) * 2014-07-30 2014-11-05 浪潮集团有限公司 一种基于云计算的obds远程故障诊断恢复系统
CN105137963A (zh) * 2015-09-09 2015-12-09 山东丽驰新能源汽车有限公司 车辆远程监控与故障诊断方法及装置
CN105867351A (zh) * 2016-04-29 2016-08-17 大连楼兰科技股份有限公司 车辆故障码实时采集与历史数据分析诊断的方法及装置
CN106054867A (zh) * 2016-07-28 2016-10-26 重庆峰创科技有限公司 一种基于物联网与云计算的新能源汽车故障诊断装置
CN106327344A (zh) * 2016-08-28 2017-01-11 华南理工大学 一种基于车联网的车辆故障在线检测预警装置和方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3564647A4 *

Cited By (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109241849B (zh) * 2018-08-07 2021-02-09 浙江大学 面向智能电厂汽轮机主机的特征分解选择与故障诊断方法
CN109241849A (zh) * 2018-08-07 2019-01-18 浙江大学 面向智能电厂汽轮机主机的经验模态和小波分解及特征选择的故障诊断方法
JP2021044231A (ja) * 2019-09-11 2021-03-18 コリア インスティテュート オブ エナジー リサーチＫｏｒｅａ Ｉｎｓｔｉｔｕｔｅ Ｏｆ Ｅｎｅｒｇｙ Ｒｅｓｅａｒｃｈ 燃料電池システムの多段階故障診断方法及び装置
JP7008098B2 (ja) 2019-09-11 2022-01-25 コリア インスティテュート オブ エナジー リサーチ 燃料電池システムの多段階故障診断方法及び装置
CN111459816A (zh) * 2020-03-31 2020-07-28 北京百度网讯科技有限公司 故障注入测试方法、装置、系统及存储介质
CN111459816B (zh) * 2020-03-31 2023-11-14 北京百度网讯科技有限公司 故障注入测试方法、装置、系统及存储介质
CN111873978A (zh) * 2020-07-10 2020-11-03 中车齐齐哈尔车辆有限公司 列车管压力的监测方法和系统
CN112327190B (zh) * 2020-10-14 2023-06-20 北方工业大学 一种储能电池健康状态辨识方法
CN112327190A (zh) * 2020-10-14 2021-02-05 北方工业大学 一种储能电池健康状态辨识方法
CN113536658A (zh) * 2021-05-21 2021-10-22 西北工业大学 基于stm32嵌入式处理器的机电设备轻量化故障诊断方法
CN115373369A (zh) * 2022-08-24 2022-11-22 中国第一汽车股份有限公司 车辆故障诊断系统及方法
CN115373369B (zh) * 2022-08-24 2024-05-03 中国第一汽车股份有限公司 车辆故障诊断系统及方法
CN115810227A (zh) * 2022-12-01 2023-03-17 深圳瑞为智能科技有限公司 一种车载设备异常快速识别方法及系统

Also Published As

Publication number Publication date
KR20190107080A (ko) 2019-09-18
EP3564647A1 (en) 2019-11-06
US20230018604A1 (en) 2023-01-19
EP4119919A1 (en) 2023-01-18
JP2020507748A (ja) 2020-03-12
EP3564647A4 (en) 2019-11-13
JP6830540B2 (ja) 2021-02-17
US11468715B2 (en) 2022-10-11
US20190333291A1 (en) 2019-10-31
US12086165B2 (en) 2024-09-10
EP3564647B1 (en) 2022-06-29
CN108303264A (zh) 2018-07-20
KR102263337B1 (ko) 2021-06-09
CN108303264B (zh) 2020-03-20

Similar Documents

Publication Publication Date Title
WO2018129917A1 (zh) 一种基于云的车辆故障诊断方法、装置及其系统
US11954651B2 (en) Sensor-based digital twin system for vehicular analysis
US11468215B2 (en) Digital twin for vehicle risk evaluation
CN113119937B (zh) 基于数字孪生的智能线控制动系统及其预测控制方法
CN106406296A (zh) 一种基于车载和云端的列车故障诊断系统及方法
KR102141988B1 (ko) 자율주행 대중버스에 대한 실시간 버스 신뢰성 통합지표 모니터링 시스템 및 방법
CN111506048B (zh) 车辆故障预警方法及相关设备
CN107274700B (zh) 一种车路协同环境下多源信息采集方法及装置
WO2024087447A1 (zh) 无人驾驶矿用车辆制动性能检测方法、装置、电子设备、存储介质及计算机程序产品
CN110895414B (zh) 用于确定并监视额外燃料消耗的原因的方法和系统
CN111381165A (zh) 一种车辆动力电池监控方法、装置和平台
CN104460622A (zh) 一种车辆远程监控系统
CN116432397A (zh) 一种基于数据模型的轨道交通故障预警方法
Gubanov et al. Architecture of a system for diagnosing and predicting the technical condition of a robotic vehicle
EP3613975B1 (en) Method and system for determining a cause of extra-fuel consumption
WO2024168454A1 (zh) 车辆底盘健康管理方法、装置与系统
RU2798250C2 (ru) Способ и система для определения причины чрезмерного расхода топлива
CN111145381B (zh) 一种电动车辆的安全状态评估方法以及电动车辆
CN118033429A (zh) 一种蓄电池的电压故障预测方法、装置、设备及存储介质
CN117994979A (zh) 多目标融合的数据处理方法及装置
CN118331239A (zh) 基于云计算的新能源汽车远程诊断方法及系统
CN118306216A (zh) 一种基于大数据分析的新能源汽车配电监管系统
CN117520953A (zh) 怠速异常识别及模型训练方法、装置、车辆及介质
CN118182256A (zh) 车辆能耗管理方法、装置、设备及介质
CN112861930A (zh) 基于机器学习的车辆故障分析及预测方法和装置

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: 17890940

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2019538380

Country of ref document: JP

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 2017890940

Country of ref document: EP

Effective date: 20190729

ENP Entry into the national phase

Ref document number: 20197023618

Country of ref document: KR

Kind code of ref document: A