WO2022241705A1 - 车辆监测方法、装置、设备及计算机可读存储介质 - Google Patents

车辆监测方法、装置、设备及计算机可读存储介质 Download PDF

Info

Publication number
WO2022241705A1
WO2022241705A1 PCT/CN2021/094788 CN2021094788W WO2022241705A1 WO 2022241705 A1 WO2022241705 A1 WO 2022241705A1 CN 2021094788 W CN2021094788 W CN 2021094788W WO 2022241705 A1 WO2022241705 A1 WO 2022241705A1
Authority
WO
WIPO (PCT)
Prior art keywords
data
operating
vehicle
sets
normal
Prior art date
Application number
PCT/CN2021/094788
Other languages
English (en)
French (fr)
Inventor
尹航一
蒋朋家
李家鑫
Original Assignee
四川金瑞麒智能科学技术有限公司
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 四川金瑞麒智能科学技术有限公司 filed Critical 四川金瑞麒智能科学技术有限公司
Priority to PCT/CN2021/094788 priority Critical patent/WO2022241705A1/zh
Priority to CN202180092815.5A priority patent/CN116963936A/zh
Publication of WO2022241705A1 publication Critical patent/WO2022241705A1/zh

Links

Images

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60C—VEHICLE TYRES; TYRE INFLATION; TYRE CHANGING; CONNECTING VALVES TO INFLATABLE ELASTIC BODIES IN GENERAL; DEVICES OR ARRANGEMENTS RELATED TO TYRES
    • B60C23/00—Devices for measuring, signalling, controlling, or distributing tyre pressure or temperature, specially adapted for mounting on vehicles; Arrangement of tyre inflating devices on vehicles, e.g. of pumps or of tanks; Tyre cooling arrangements
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60L—PROPULSION OF ELECTRICALLY-PROPELLED VEHICLES; SUPPLYING ELECTRIC POWER FOR AUXILIARY EQUIPMENT OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRODYNAMIC BRAKE SYSTEMS FOR VEHICLES IN GENERAL; MAGNETIC SUSPENSION OR LEVITATION FOR VEHICLES; MONITORING OPERATING VARIABLES OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRIC SAFETY DEVICES FOR ELECTRICALLY-PROPELLED VEHICLES
    • B60L58/00—Methods or circuit arrangements for monitoring or controlling batteries or fuel cells, specially adapted for electric vehicles
    • B60L58/10—Methods or circuit arrangements for monitoring or controlling batteries or fuel cells, specially adapted for electric vehicles for monitoring or controlling batteries

Definitions

  • This description relates to the technical field of data monitoring, in particular to a vehicle monitoring method, device, equipment and computer-readable storage medium.
  • One aspect of the present specification provides a method for vehicle monitoring, the method comprising: acquiring the operating condition data of the components to be detected of the vehicle, the operating condition data at least including several groups of normal data sets under normal operating conditions, each The set of normal data sets includes first operating data, and the first motion data of the vehicle at the corresponding moment of the first operating data; at least one set of measured data sets of the component to be detected in actual operation is obtained, and each set of The measured data set includes second running data, and second running data of the vehicle at a time corresponding to the second running data; wherein, both the first running data and the second running data reflect the running state of the component to be detected Information, the first motion data and the second motion data both reflect the motion state information of the vehicle; based on the operating condition data and the measured data set, determine the The operating status of the component to be tested.
  • the device includes: a first acquisition module, configured to acquire operating condition data of the components to be detected of the vehicle, the operating condition data at least including several sets of normal data sets under normal operating conditions, each set of normal data
  • the set includes first operating data, and the first motion data of the vehicle at the time corresponding to the first operating data
  • the second acquisition module is used to acquire at least one set of measured data sets of the component to be detected in actual operation, each set
  • the measured data set includes second operating data, and second moving data of the vehicle at a time corresponding to the second operating data; wherein, the first operating data and the second operating data both reflect the Running state information, the first motion data and the second motion data both reflect the motion state information of the vehicle
  • a state determination module is configured to determine each group of the The running state of the component to be detected reflected by the measured data set.
  • a vehicle monitoring device including at least one storage medium and at least one processor, the at least one storage medium is used to store computer instructions; the at least one processor is used to execute the computer instructions to achieve the following: The vehicle monitoring method described in any of the above schemes.
  • Another aspect of this specification provides a computer-readable storage medium, the storage medium stores computer instructions, and when the computer reads the computer instructions in the storage medium, the computer executes the vehicle monitoring method described in any of the solutions above.
  • the possible beneficial effects of the embodiments of this specification include but are not limited to: (1) By comparing the intervals of the measured data sets of the parts to be tested with the normal data sets, the working conditions of the parts to be tested can be quickly and accurately obtained; (2) ) Use self-selected data sets under a variety of complex working conditions that are in line with the actual operating conditions as training samples, and through a specific model training process, the intelligent machine learning model trained to determine the operating status of the components to be detected, whether it is data processing The accuracy and adaptability are both high, and the monitoring effect of the expected operating status information data can be obtained, which improves the data processing efficiency and avoids human errors; (3) correspondingly screens the data sets in the preset interval to avoid external environment factor interference.
  • Fig. 1 is a schematic diagram of an application scenario of a vehicle monitoring system according to some embodiments of this specification
  • Fig. 2 is an exemplary flow chart of a vehicle monitoring method according to some embodiments of the present specification
  • Fig. 3a is an example diagram showing the vehicle speed and its corresponding motor current value under normal working conditions according to some embodiments of the present specification
  • Fig. 3b is an example diagram showing the vehicle speed and its corresponding motor current value in actual operation according to some embodiments of the present specification
  • Fig. 4 is an exemplary flowchart of a method for determining the operating state of a component to be detected according to some embodiments of the present specification
  • Fig. 5 is an exemplary flow chart of a method for determining the operating state of a component to be detected according to other embodiments of the present specification
  • Fig. 6 is a schematic structural diagram of an exemplary vehicle monitoring device according to some embodiments of the present specification.
  • system means for distinguishing different components, elements, parts, parts or assemblies of different levels.
  • the words may be replaced by other expressions if other words can achieve the same purpose.
  • Fig. 1 is a schematic diagram of an application scenario of a vehicle monitoring system according to some embodiments of the present specification.
  • Scenarios can be applied to transportation systems and traffic service systems.
  • the scene can be applied to the monitoring of vehicle data in any area.
  • the vehicle monitoring system 100 can be applied to data monitoring of intelligent vehicles in scenic spots, factories, ports, schools and other areas.
  • the vehicle monitoring system 100 can be an online service platform, including a server 110 , a network 120 , a terminal 130 , a database 140 and a vehicle 150 .
  • the server 110 may include a processing device 112 .
  • server 110 may be used to process information and/or data related to monitoring for a vehicle.
  • Server 110 may be an individual server or a group of servers.
  • the server group can be centralized or distributed (eg, server 110 can be a distributed system).
  • the server 110 may be regional or remote in some embodiments.
  • the server 110 can access information and/or materials stored in the terminal 130 , the database 140 , and the vehicle 150 through the network 120 .
  • the server 110 can be directly connected to the terminal 130, the database 140, and the vehicle 150 to access information and/or materials stored therein.
  • server 110 may execute on a cloud platform.
  • the cloud platform may include one of private cloud, public cloud, hybrid cloud, community cloud, decentralized cloud, internal cloud, etc. or any combination thereof.
  • server 110 may include processing device 112 .
  • the processing device 112 may process data and/or information related to vehicle monitoring to perform one or more of the functions described herein. For example, processing device 112 may acquire normal data sets and measured data sets. For another example, the processing device 112 may determine the running state of the component to be detected based on the normal data set and the measured data set.
  • the processing device 112 may include one or more sub-processing devices (eg, a single-core processing device or a multi-core, multi-core processing device).
  • processing device 112 may include a central processing unit (CPU), an application specific integrated circuit (ASIC), an application specific instruction processor (ASIP), a graphics processing unit (GPU), a physical processing unit (PPU), a digital signal processor ( DSP), field programmable gate array (FPGA), programmable logic circuit (PLD), controller, microcontroller unit, reduced instruction set computer (RISC), microprocessor, etc. or any combination of the above.
  • CPU central processing unit
  • ASIC application specific integrated circuit
  • ASIP application specific instruction processor
  • GPU graphics processing unit
  • PPU physical processing unit
  • DSP digital signal processor
  • FPGA field programmable gate array
  • PLD programmable logic circuit
  • controller microcontroller unit, reduced instruction set computer (RISC), microprocessor, etc. or any combination of the above.
  • Network 120 may facilitate the exchange of data and/or information.
  • one or more components in the vehicle monitoring system 100 eg, the server 110 , the terminal 130 , the database 140 , the vehicle 150
  • network 120 may be any type of wired or wireless network.
  • network 120 may include a cable network, a wired network, a fiber optic network, a telecommunications network, an intranet, the Internet, an area network (LAN), a wide area network (WAN), a wireless area network (WLAN), a metropolitan area network (MAN) , Public Switched Telephone Network (PSTN), Bluetooth network, ZigBee network, Near Field Communication (NFC) network, etc.
  • LAN area network
  • WAN wide area network
  • WLAN wireless area network
  • MAN metropolitan area network
  • PSTN Public Switched Telephone Network
  • Bluetooth network ZigBee network
  • NFC Near Field Communication
  • network 120 may include one or more network access points.
  • network 120 may include wired or wireless network access points, such as base stations and/or internetwork switching points 120-1, 120-2, ..., through which one or more components of system 100 may be connected to network 120 to exchange data and/or information.
  • Terminal 130 may be used to input and/or obtain data and/or information.
  • the terminal 130 may include a smartphone 130-1, a tablet 130-2, a laptop 130-3, and the like.
  • the terminal 130 may include a mobile terminal device or the like.
  • a mobile device may include a smart home device, a wearable device, a smart mobile device, a virtual reality device, an augmented reality device, or any combination of the above examples.
  • smart home devices may include smart lighting devices, smart electrical control devices, smart monitoring devices, smart TVs, smart cameras, walkie-talkies, etc., or any combination of the above examples.
  • wearable devices may include bracelets, footwear, glasses, helmets, watches, clothing, backpacks, smart accessories, etc. or any combination of the above examples.
  • the smart mobile device may include a mobile phone, a personal digital assistant, a game device, a navigation device, a POS machine, a laptop computer, a desktop computer, etc. or any combination thereof.
  • the user can obtain the running status of the vehicle through the terminal 130 . In some embodiments, the user can obtain the normal data set and the measured data set of the vehicle through the terminal 130 .
  • Database 140 may store data and/or instructions.
  • the database 140 may store information obtained from the terminal 130 .
  • database 140 may store information and/or instructions for execution or use by server 110 to perform the example methods described herein.
  • the database 140 may store normal data sets, measured data sets, machine learning models, and the like.
  • the database 140 may include mass memory, removable memory, volatile read-write memory (eg, random access memory RAM), read-only memory (ROM), or any combination thereof.
  • database 140 may be implemented on a cloud platform.
  • the cloud platform may include private cloud, public cloud, hybrid cloud, community cloud, decentralized cloud, internal cloud, etc. or any combination thereof.
  • the vehicle 150 may include various types of vehicles, such as bicycles, electric vehicles, motorcycles, automobiles, trucks, vans and so on.
  • the vehicle 150 may send the acquired data to one or more devices in the vehicle monitoring system 100 .
  • the vehicle 150 may send the acquired data to the server 110 through the network 120 for subsequent steps.
  • vehicle 150 may communicate with one or more devices in vehicle monitoring system 100 .
  • the vehicle 150 can communicate with the terminal 130 through the network 120 .
  • the vehicle 150 may directly communicate with the terminal 130 .
  • data for vehicles 150 may be stored in database 140 .
  • database 140 may be connected to network 120 to communicate with one or more components of system 100 (eg, server 110, terminal 130, vehicle 150, etc.).
  • One or more components of system 100 may access data or instructions stored in database 140 via network 120 .
  • the server 110 may extract normal data sets, measured data sets, and machine learning models from the database 140 and perform corresponding processing.
  • the database 140 may directly connect or communicate with one or more components in the system 100 (eg, the server 110, the terminal 130, the vehicle 150).
  • database 140 may be part of server 110 .
  • Fig. 2 is an exemplary flow chart of a vehicle monitoring method according to some embodiments of the present specification.
  • the process 200 can be executed by a processing device (such as the server 110 ), a vehicle monitoring system (such as the system 100 ) or a vehicle monitoring device (such as the device 600 ). It includes:
  • Step 210 obtain the operating condition data of the components to be tested in the vehicle, the operating condition data includes at least several sets of normal data sets under normal operating conditions, each set of normal data sets includes the first operating data, and the first operating data corresponds to The first movement data of the vehicle at a time.
  • the component to be detected is a component that needs to be detected in the vehicle.
  • the components to be detected may include but not limited to motors, engines, transmissions, electronically controlled power steering systems, automatic air conditioning systems, automatic transmission electronic control systems, and the like.
  • the operating condition data refers to the data generated by the vehicle under one or more (at least two) operating conditions.
  • the operating conditions of the vehicle include but are not limited to normal working conditions, abnormal working conditions, aging working conditions, working conditions to be repaired, etc.
  • the operating condition data may include several sets of normal data sets under normal operating conditions.
  • the operating condition data may include at least one of the following in addition to several groups of normal data sets under normal working conditions: abnormal data sets under abnormal working conditions, several groups of aging data sets under aging conditions Data sets, several groups of data sets to be repaired under the working conditions to be repaired.
  • the normal data set is a data set generated under the normal working conditions of the vehicle.
  • the normal data set may include the odometer value, battery compartment temperature, battery compartment humidity, battery capacity, sensor operating conditions, etc. of the vehicle under normal conditions.
  • the normal data set may also include a collection of data such as speed, voltage value, pulling force value, position, acceleration, etc. of the vehicle at a certain moment or within a certain period of time.
  • a set of normal data sets includes a first running data and a first motion data of the vehicle at a time corresponding to the first running data.
  • the first operating data may be one or more current values of the motor under normal working conditions
  • the first motion data may be one or more speeds of the vehicle at a time corresponding to the one or more current values.
  • the first operating data may be one or more power values of the motor under normal working conditions
  • the first motion data may be one or more accelerations of the vehicle at a time corresponding to the one or more power values of the motor.
  • the first operating data may be one or more power values of the engine under normal operating conditions
  • the first motion data may be one or more exhaust volumes of the vehicle at the time corresponding to one or more power values of the vehicle .
  • the first running data is a motor current value of 50A
  • the corresponding first motion data (speed) of the vehicle at this time is 60km/h.
  • the first running data is the motor power value is 10kw
  • the corresponding first motion data (speed) of the vehicle at this time is 110km/h.
  • the car will obtain multiple sets of normal data sets during normal driving.
  • the multiple sets of normal data sets may include multiple sets of first running data and multiple sets of first motion data during vehicle driving.
  • a dataset is a collection of data generated by a vehicle at runtime.
  • the vehicle may be any type of vehicle, and may include, for example, a miniature vehicle, a small vehicle, a medium-sized vehicle, or a large vehicle.
  • a vehicle may include a car, a truck, a passenger car, a trailer, a partial vehicle, a motorcycle, and the like.
  • the data set may include system resource data, hardware status data, hardware data, and software module operation data.
  • the data set may include lower computer sensor data, lower computer system resource data, lower computer sensor status data, upper computer sensor data, upper computer sensor status data, upper computer system resource data, upper computer business software operation data , Host computer business software resource occupation data.
  • data sets may be collected in the form of publish and subscribe topics. For example, only the corresponding data requirements need to be published during collection, so that only the required data is collected.
  • the data set may include, but is not limited to, data such as current value, voltage value, mileage, acceleration, tension, position, inertia, gasoline level, light detection and distance measurement corresponding to the running time of the vehicle.
  • the data set of the vehicle may be detected by sensors or measurement units of the vehicle.
  • the data set can be used to include cloud storage display analysis, human-computer interaction, software business, remote problem location, remote debugging, and big data analysis.
  • the way of obtaining the data set can be any feasible way.
  • the data set can be obtained by reading stored data, calling related interfaces, or other methods. For example, through network transmission or direct upload.
  • vehicles can upload data sets through the network, and the processing equipment can then obtain them.
  • the acquired data sets may be stored in database 140 .
  • the running data is the data of the vehicle components to be detected in the driving state.
  • the running data may include the current value, voltage value, and power of the motor.
  • the operating data may also include output power, output energy, torque, etc. of the engine.
  • the first operating data, the second operating data, the third operating data, the fourth operating data, and the fifth operating data are data in different states of the operating data.
  • the motion data is the data in the motion state of the vehicle.
  • the motion data may include the speed, acceleration, displacement, etc. of the vehicle when it is in motion.
  • the first motion data, the second motion data, the third motion data, the fourth motion data, and the fifth motion data are data of motion data in different states.
  • Step 220 acquiring at least one set of measured data sets of the component to be detected during actual operation, each set of measured data sets includes second operating data, and second motion data of the vehicle at a time corresponding to the second operating data; wherein, the first operating data Both the first motion data and the second motion data reflect the motion state information of the component to be detected, and the first motion data and the second motion data reflect the motion state information of the vehicle.
  • the measured data set is the data set measured by the component to be tested in actual operation.
  • the measured data set corresponds to the normal data set, and represents the actual operation of the components to be detected of the vehicle.
  • a set of measured data sets includes a second running data and a second motion data of the vehicle at a time corresponding to the second running data.
  • the second running data is a motor current value of 50A
  • the corresponding second motion data (speed) of the vehicle at this time is 50km/h.
  • the vehicle will obtain multiple sets of measured data sets during actual operation.
  • the multiple sets of measured data sets may include multiple sets of second running data and multiple sets of second motion data during vehicle driving.
  • the types of motion data and operating data can be determined according to the type of the component to be inspected. For example, if the component to be detected is a motor, it can be determined that the type of the running data is the motor current value or voltage value, etc., and the motion data is the vehicle speed or acceleration, etc. For another example, if the component to be detected is an engine, it may be determined that the type of the running data is the power of the engine, and the motion data is the acceleration of the vehicle.
  • the second running data of the measured data set is of the same type as the first running data of the normal data set, and the second motion data of the measured data set is of the same type as the second running data of the normal data set.
  • the first operating data in the normal data set is the current value of the motor
  • the second operating data in the measured data set is also the current value of the motor.
  • the first motion data in the normal data set is the speed of the vehicle
  • the second motion data in the measured data set is the speed of the vehicle.
  • the first operating data in the normal data set is the power value of the engine
  • the second operating data in the measured data set is also the power value of the engine.
  • the first motion data in the normal data set is the acceleration of the vehicle
  • the second motion data in the measured data set is the acceleration of the vehicle.
  • the acquisition interval of the measured data set may be determined according to the type of the component to be inspected.
  • the acquisition interval length is the time length of the interval for acquiring the measured data set.
  • the acquisition interval may be shortened for components to be detected with high importance, and the acquisition interval may be increased for components to be detected with low importance.
  • the acquisition interval of the measured data set can be set to 30 seconds, which means that the measured data set is acquired once every 30 seconds.
  • the acquisition interval of the measured data set can be set to 40 seconds, which means that the measured data set is acquired once every 40 seconds.
  • the corresponding setting of the acquisition interval of the measured data set can be more targeted to obtain data sets that meet different expected needs, and the efficiency of obtaining the measured data set can be improved.
  • Step 230 based on the operating condition data and the measured data sets, determine the running status of the component to be tested reflected by each set of measured data sets.
  • the operating status of the component to be tested reflected by each set of measured data sets can be determined.
  • the multiple sets of normal data sets include multiple sets of first operating data and first motion data.
  • the maximum and minimum values of multiple first motion data corresponding to each first running data of multiple sets of normal data sets can be analyzed and obtained.
  • the maximum and minimum values of the first motion data are compared with the second motion data corresponding to a set of measured data sets. After the comparison, the operating status of the components to be detected can be obtained. For example, taking the motor as an example, compare the current value at a certain speed of a set of measured data sets with the maximum or minimum current value at this speed of the normal data set, and if the difference between the two current values exceeds a threshold, you can It is considered that the state of the motor is abnormal.
  • the state of the motor can be considered normal.
  • the current values corresponding to 50km/h under the normal data set include 10A, 12A, and 14A.
  • the threshold is set to 0.5A, and the maximum current value of 50km/h in the normal data set is 14A.
  • the current value corresponding to 50km/h in the measured data set is 14.2A. Compared with 14A, the current value does not exceed the threshold of 0.5A, and the motor state can be considered normal.
  • the multiple first motion data corresponding to each first running data under the normal data set can be processed to obtain the average value of the multiple first motion data
  • the second motion data of the measured data set can be combined with the Average value comparison can obtain the running state of the component to be tested.
  • several groups of power values at a certain speed in the normal data set can be processed to obtain the average power value at that speed. Compare the average power value with the power value of the speed under the measured data set. If the difference between the two power values exceeds the threshold, the engine state can be considered abnormal. If the difference between the two power values does not exceed the threshold value, the engine state can be considered normal. .
  • the power values corresponding to 50km/h in the normal data set include 500kw, 600kw, and 700kw, and the average power value of 50km/h in the normal data set is 600kw.
  • the set threshold is 150kw.
  • the power value corresponding to 50km/h in the measured data set is 440kw, and the difference between 440kw and 600kw is 160kw. If it exceeds the threshold, it can be considered that the engine state is abnormal.
  • the normal data interval of the first running data corresponding to each first motion data may be determined based on several sets of normal data sets.
  • the second motion data corresponding to the second running data is compared with the normal data interval corresponding to the first motion data, and a comparison result is obtained.
  • the operating state of the component to be detected is determined. Specifically, as shown in Fig. 3a and Fig. 3b, the motor is used as the running data, and the vehicle speed is used as the motion data as an example.
  • Fig. 3a is an example diagram showing the vehicle speed and its corresponding motor current value under normal working conditions according to some embodiments of the present specification.
  • Fig. 3b is an example diagram showing the vehicle speed and its corresponding motor current value in actual operation according to some embodiments of the present specification.
  • the value of the second operating data corresponding to the vehicle speed of 10km/h is 8A, and under normal working conditions, the value range of multiple first operating data corresponding to the vehicle speed of 10km/h is 2A-11A .
  • the value of the second operating data is in the normal interval of the first operating data, so the measured data set of this group reflects that the components to be detected are normal.
  • the value of the second operating data corresponding to the vehicle speed of 30km/h is 20A, and under normal working conditions, the value range of multiple first operating data corresponding to the vehicle speed of 30km/h is 4A-18A .
  • the value of the second operating data is not in the normal interval of the first operating data, so the measured data set of this group reflects that the component to be detected is abnormal.
  • the value of the second operating data corresponding to the vehicle speed of 50km/h is 20A, and under normal working conditions, the value range of multiple first operating data corresponding to the vehicle speed of 50km/h is 10A-25A .
  • the value of the second operating data is in the normal interval of the first operating data, so the measured data set of this group reflects that the components to be detected are normal.
  • a corresponding maintenance program is started to perform maintenance on the components to be detected. For example, when the measured data set in the motor reflects that the number of abnormal parts to be detected is greater than 10, the corresponding maintenance program is started to overhaul the motor.
  • Fig. 4 is an exemplary flowchart of a method for determining the running state of a component to be detected according to some embodiments of the present specification.
  • the operating condition data in addition to several sets of normal data sets under normal operating conditions, also includes abnormal data sets under abnormal operating conditions.
  • the process 400 can be executed by a processing device (such as the server 110 ), a vehicle monitoring system (such as the system 100 ), or a vehicle monitoring device (such as the device 600 ). It includes:
  • Step 410 acquiring several sets of normal data sets of the components to be detected of the vehicle under normal working conditions.
  • Step 410 is the same as step 210, see step 210 for details, and will not be repeated here.
  • Step 420 acquiring the abnormal data sets of the components to be inspected under abnormal working conditions of the vehicle, each set of abnormal data sets includes third operating data, and third motion data of the vehicle at a time corresponding to the third operating data.
  • the abnormal data set is a data set generated by the component to be detected of the vehicle under abnormal working conditions.
  • An example is the data set generated by the component to be tested in the event of a failure.
  • the data set includes a normal data set and an abnormal data set, and the abnormal data set can be obtained by filtering the normal data set from the data set.
  • the abnormal data set includes third operating data and third motion data of the vehicle at a time corresponding to the third operating data.
  • the third running data reflects the running state information of the components to be monitored of the vehicle
  • the third motion data reflects the motion state information of the vehicle.
  • the running data and motion data please refer to the related description of FIG. 2 .
  • the component to be detected can be any feasible component in the vehicle
  • the first external data and the second external data can be any feasible data type that can reflect the vehicle operating environment information
  • the first operating data, the second operating data, and the third operating data can be It is any feasible data type that can reflect the running status information of the components to be monitored of the vehicle.
  • the first motion data, the second motion data, and the third motion data may be any feasible data type that can reflect the motion state information of the vehicle.
  • the component to be detected can include a motor; the first external data can at least include a road gradient value and/or a vehicle load, for example, the first external data can be a road gradient value and a vehicle load, and the second external data can at least include Road gradient value and/or vehicle load, for example, the second external data can only be road gradient value or vehicle load;
  • the running data may at least include the current value of the motor; the first movement data may at least include the vehicle speed, the second movement data may at least include the vehicle speed, and the third movement data may at least include the vehicle speed. Since the selected road gradient value, vehicle load, motor current value, and vehicle speed can better reflect the operating state of the vehicle to be detected, it is more representative, so that the calculation of the operating state of the to-be-detected parts can be performed more accurately and efficiently monitor.
  • Step 430 using the normal data set as a positive sample and the abnormal data set as a negative sample, and using the positive sample and the negative sample to train the initial first machine learning model to obtain the trained first machine learning model.
  • the initial first machine learning model may be a classifier.
  • logistic regression models, support vector machines, random forests, or other classification models, etc. may be used as classifiers.
  • the initial training samples of the first machine learning model are normal data set and abnormal data set. Positive samples are normal data sets and labels are normal. Negative samples are non-normal datasets, and the labels are non-normal.
  • the data set may also include external data.
  • the external data may include first external data and second external data. Both the first external data and the second external data can reflect the vehicle operating environment information, and the external data may interfere with the final determination of the state of the component to be detected to a certain extent. It is understandable that when the external data are different, the data results of the normal data set and the measured data set will also have corresponding errors.
  • the external data may include road grade values and/or vehicle load.
  • the slope of the road indicates the angle of the driving road to the horizontal ground.
  • the slope of the road may be 3 degrees, 5 degrees, or the like.
  • the vehicle load indicates the weight of items carried on the vehicle, for example, the vehicle load may be 1 ton, 3 tons, and so on.
  • the data set may also include external data of the vehicle at a time corresponding to the running data. For example, when the current value of the motor is 10A, the slope of the road at the corresponding moment is 5 degrees.
  • the normal data set includes first external data of the vehicle at a time corresponding to the first operating data.
  • the normal data set corresponding to the first external data within the first preset interval may be used as a positive sample.
  • the first external data is the slope value of the road
  • the first preset interval is 0° to 5°
  • the normal data set between 0° and 5° is used as a positive sample.
  • the abnormal data set includes second external data of the vehicle at a time corresponding to the second operating data.
  • the normal data set corresponding to the second external data within the second preset interval may be used as a positive sample.
  • the second external data is a road slope value
  • the second preset interval is 0° to 5°
  • the abnormal data set between 0° and 5° is used as a negative sample.
  • the first preset interval may be the same as or different from the second preset interval. It can be understood that filtering out the data sets within the preset interval can effectively compare the interference of the external environment on the data sets.
  • the initial first machine learning model can be trained using a commonly used method, such as a gradient descent method.
  • the initial machine learning model may also be trained using the Levenberg-Marquardt algorithm (LM algorithm, also called the attenuated least squares method) to obtain the first machine learning model after training.
  • LM algorithm Levenberg-Marquardt algorithm
  • the LM algorithm is more suitable for small and medium-sized networks that identify vehicle fault points, such as BP neural network models.
  • Step 440 based on the measured data set, using the trained first machine learning model, to determine whether the operation state of the component to be detected reflected by each set of measured data set is normal or abnormal.
  • a set of measured data sets are input into the trained first machine learning model, and the running status of the component to be tested reflected by the measured data sets can be obtained.
  • the component to be detected in the measured data set may be a motor
  • the running data may be a current value of the motor.
  • the motion data may be the speed at which the vehicle is traveling.
  • the current value of the motor is 10A
  • the driving speed of the vehicle is 30km/h, which can be input into the first machine school model after training, and the final output result is normal, indicating that the operating state of the component to be detected is normal.
  • the current value of the motor is 20A
  • the driving speed of the vehicle is 40km/h, which can be input into the first machine school model after training to obtain the final output result as abnormal, indicating that the operating state of the component to be detected is abnormal.
  • Fig. 5 is an exemplary flow chart of a method for determining an operating state of a component to be detected according to other embodiments of the present specification.
  • the operating condition data may also include several sets of aging data sets under the aging condition and several sets of data sets to be maintained under the condition to be maintained.
  • the process 500 can be executed by a processing device (such as the server 110 ), a vehicle monitoring system (such as the system 100 ), or a vehicle monitoring device (such as the device 600 ). It includes:
  • Step 510 obtaining several groups of aging data sets of the parts to be detected under aging conditions, each group of aging data sets includes the fourth operation data, and the fourth motion data of the vehicle at the time corresponding to the fourth operation data;
  • Several sets of data sets to be maintained under maintenance conditions each set of data sets to be maintained includes the fifth operating data, and the fifth motion data of the vehicle at the time corresponding to the fifth operating data; wherein, the fourth operating data and the fifth operating data are both Reflecting the running state information of the component to be detected, the fourth motion data and the fifth motion data both reflect the motion state information of the vehicle.
  • each set of aging data sets includes fourth operating data, and the fourth operating data corresponds to fourth motion data of the vehicle at a time.
  • each group of data sets to be repaired includes fifth operating data and fifth motion data of the vehicle at a time corresponding to the fifth operating data.
  • the fourth running data and the fifth running data respectively correspond to the first running data of the normal data set
  • the fourth motion data and the fifth motion data respectively correspond to the first motion data of the normal data set. It represents data in different operating states.
  • At least one or a combination of visual sensors, motor Hall sensors, inertial sensors, and GPS positioning devices can be used to collect operating condition data and/or the measured data set.
  • the visual sensor can be used to obtain imaging noise detection data, so as to judge the operating status of the vehicle glass parts in front of the visual sensor through the imaging noise detection data, such as the aging of the vehicle glass;
  • the combination of the motor Hall sensor and the GPS positioning device can be used to obtain speed data, pulse abnormal data, noise data and/or frequency data, so as to judge the running state of the motor Hall sensor through the speed data, pulse abnormal data, noise data and/or frequency data, for example, according to the motor Hall sensor Whether the speed output by the Hall sensor matches the real speed, for example, judge whether the pulse is abnormal according to the noise data and frequency data, etc., so as to judge the health status of the Hall sensor of the motor;
  • the inertial sensor can be used to obtain vehicle positioning data, speed And/or acceleration data, so as to provide a data basis for further analysis and determination of the running
  • the expected specific operating condition data and/or measured data sets can be obtained more efficiently, thereby providing a favorable data basis for subsequent data analysis and calculation.
  • Step 520 using the normal data set, the aging data set and the data set to be repaired as training samples to train the initial second machine learning model to obtain the trained second machine learning model.
  • the initial training samples of the second machine learning model are normal data set, aging data set and data set to be repaired.
  • the preset algorithm trains the initial machine learning model.
  • the introduction of the model, the training method, and the specific algorithm used are similar to those in Figure 4. For details, refer to the relevant description in Figure 4, and will not be repeated here.
  • the loss function is optimized by adjusting the parameters of the initial second machine learning model (for example, parameters such as learning rate, number of iterations, batch size, etc.), and when the loss function satisfies the preset condition, the training ends and a well-trained Second machine learning model.
  • the parameters of the initial second machine learning model for example, parameters such as learning rate, number of iterations, batch size, etc.
  • the training samples collect normal data sets, aging data sets and data sets to be repaired in the whole life cycle of the components to be tested. Since the data in the full life cycle can fully reflect the operating status of the components to be tested, the training The data richness and representativeness of the sample are high, and the initial machine learning model is trained with the preset algorithm by using this training sample until the expected convergence is reached. After that, high accuracy, high adaptability and high convergence are obtained. Post-training machine learning model.
  • the LM algorithm may also be used to train the initial machine learning model to obtain the trained second machine learning model.
  • the BP neural network model is trained using the labeled training sample data of the full life cycle of multiple electric motors until it converges stably (the recognition rate is no longer improved), and basically meets expectations (may be different from the original label to a certain extent). in and out, but may be more reasonable).
  • the process of utilizing the LM algorithm to train the BP neural network model can be performed as the following steps:
  • Step 2 Calculate network output and error index function E(x(k));
  • Step 3 calculating the Jacobian matrix J(x);
  • Step 4. Calculate delta_x and E(x(k)) respectively;
  • Step 5 If E(x(k)) ⁇ e, go to step 7; otherwise, calculate x(k+1) and calculate the error index function E(x(k+1)) as the weight and threshold;
  • Step 7 the training ends.
  • the values of all training-related parameters can be set according to the specific training process, and are not specifically limited here.
  • gradient vectors and Jacobian matrices can be used to train the BP neural network model with the LM algorithm.
  • the specific training process of the BP neural network model using the above-mentioned LM algorithm can obtain a trained machine learning model with expected convergence and high recognition rate, which is especially suitable for the application scenario of vehicle complex fault point identification.
  • Step 530 based on the measured data set, using the trained second machine learning model to determine whether the operating status of the component to be detected reflected by each set of measured data set is normal, aging or to be repaired.
  • the trained second machine learning model may be used to determine the operating status of the component to be inspected reflected by each set of measured data sets. For example, by inputting a set of measured data sets into the trained second machine learning model, the running status of the component to be detected reflected by the measured data set can be obtained.
  • the component to be detected may be a motor
  • the running data may be a current value of the motor.
  • the motion data may be the speed at which the vehicle is traveling.
  • the motor current value of 10A and the vehicle's driving speed of 10km/h can be input into the trained first machine school model to obtain the final output result as aging, indicating that the operating state of the component to be detected is aging.
  • the current value of the motor is 30A
  • the driving speed of the vehicle is 10km/h
  • the current, acceleration, speed, slope value and body weight information of the vehicle in the real-time operation scene are recorded and uploaded to the cloud, and the date when the vehicle is put into operation is used as the starting point to record the time and generate the actual measurement in days.
  • the data set is input into the machine learning model after training, and the neural network output of the day is recorded.
  • the operator is provided with guiding suggestions for vehicle maintenance.
  • the imaging noise detection information data of the visual sensor can be used to judge the aging situation of the glass in front of the visual sensor; Whether the real speed matches) and the motor Hall sensor pulse (judging the abnormal pulse through noise and frequency), and finally judge the health status of the Hall sensor.
  • Fig. 6 is a schematic structural diagram of a vehicle monitoring device according to some embodiments of the present specification.
  • the vehicle monitoring device 600 may include a first acquisition module 610 , a second acquisition module 620 , and a state determination module 630 . These modules may also be implemented as an application program or a set of instructions read and executed by a processing engine. Furthermore, a module may be any combination of hardware circuits and applications/instructions. For example, a module may be part of a processor when a processing engine or processor executes an application/set of instructions.
  • the first acquisition module 610 is configured to acquire the operating condition data of the components to be detected of the vehicle, the operating condition data includes at least several groups of normal data sets under normal operating conditions, and each group of normal data sets includes the first operating data, and The first running data corresponds to the first motion data of the vehicle at the moment;
  • the second acquisition module 620 is used to acquire at least one set of measured data sets of the component to be detected in actual operation, each set of measured data sets includes the second running data, and the second set of measured data sets
  • the second running data corresponds to the second motion data of the vehicle at the moment; wherein, both the first running data and the second running data reflect the running state information of the component to be detected, and the first moving data and the second moving data both reflect the moving state information of the vehicle;
  • the state determination module 630 is configured to determine the operating state of the components to be inspected reflected in each set of measured data sets based on the operating condition data and the measured data sets.
  • the second acquisition module 620 is configured to: determine the acquisition interval of the measured data sets according to the type of the component to be inspected; acquire each group of the measured data sets at intervals of the acquisition interval.
  • the state determination module 630 is used to: determine the normal data interval of the first running data corresponding to each first exercise data based on several sets of normal data sets; determine the first normal data interval based on the measured data set and the normal data interval When the motion data is the same as the second motion data, the second motion data corresponding to the second operation data is compared with the normal data interval corresponding to the first motion data, and a comparison result is obtained; based on the comparison result, the operation of the component to be detected is determined state.
  • the operating condition data also includes abnormal data sets under abnormal working conditions
  • the state determination module 630 is used to: acquire abnormal data sets of components to be detected under abnormal working conditions, each group of abnormal data sets
  • the data set includes the third running data, and the third motion data of the vehicle at the time corresponding to the third running data; wherein, the third running data reflects the running state information of the component to be detected, and the third motion data reflects the motion state information of the vehicle; the normal
  • the data set is used as a positive sample
  • the abnormal data set is used as a negative sample
  • the initial first machine learning model is trained using the positive and negative samples to obtain the first machine learning model after training; based on the measured data set, the first machine learning model after training is used.
  • the machine learning model determines whether the operating status of the component to be detected reflected by each set of measured data sets is normal or abnormal.
  • each set of normal data sets also includes the first external data of the vehicle at the time corresponding to the first operating data
  • each set of abnormal data sets also includes the second external data of the vehicle at the time corresponding to the third operating data, the first external data
  • the data and the second external data both reflect the operating environment information of the vehicle; the normal data set is used as a positive sample, and the abnormal data set is used as a negative sample, including: the normal data corresponding to the first external data located in the first preset interval set as a positive sample; the abnormal data set corresponding to the second external data within the second preset interval as a negative sample.
  • the component to be detected includes a motor
  • each of the first external data and the second external data includes at least a road gradient value and/or a vehicle load
  • Each of the data includes at least a current value of the motor
  • each of the first motion data, the second motion data and the third motion data includes at least the driving speed of the vehicle.
  • the operating condition data also includes several groups of aging data sets under the aging condition and several groups of data sets to be maintained under the to-be-maintained condition
  • the state determination module 630 is used to: acquire the Several groups of aging data sets under certain conditions, each group of aging data sets includes the fourth operation data, and the fourth motion data of the vehicle at the corresponding time of the fourth operation data; obtain several groups of data to be repaired under the working condition of the component to be detected
  • Each set of data sets to be maintained includes the fifth running data and the fifth running data of the vehicle at the corresponding moment of the fifth running data; wherein, the fourth running data and the fifth running data both reflect the running status information of the components to be detected, and the fifth running data Both the four motion data and the fifth motion data reflect the motion state information of the vehicle;
  • the normal data set, the aging data set and the data set to be repaired are used as training samples to train the initial second machine learning model to obtain the second machine learning model after training ; Based on the measured data set, using the trained second machine learning model to determine
  • At least one or a combination of visual sensors, motor Hall sensors, inertial sensors, and GPS positioning devices can be used to collect operating condition data and/or the measured data set.
  • the visual sensor can be used to obtain imaging noise detection data, so as to judge the operating status of the vehicle glass parts in front of the visual sensor through the imaging noise detection data, such as the aging of the vehicle glass;
  • the combination of the motor Hall sensor and the GPS positioning device can be used to obtain speed data, pulse abnormal data, noise data and/or frequency data, so as to judge the running state of the motor Hall sensor through the speed data, pulse abnormal data, noise data and/or frequency data, for example, according to the motor Hall sensor Whether the speed output by the Hall sensor matches the real speed, for example, judge whether the pulse is abnormal according to the noise data and frequency data, etc., so as to judge the health status of the Hall sensor of the motor;
  • the inertial sensor can be used to obtain vehicle positioning data, speed And/or acceleration data, so as to provide a data basis for further analysis and determination of the running
  • training the initial machine learning model includes: using the normal data set, the aging data set and the data set to be repaired as training samples, using a preset The algorithm trains the initial machine learning model.
  • the LM algorithm may also be used to train the initial machine learning model to obtain the trained second machine learning model.
  • the BP neural network model is trained using the labeled training sample data of the full life cycle of multiple electric motors until it converges stably (the recognition rate is no longer improved), and basically meets expectations (may be different from the original label to a certain extent). out, but may be more reasonable).
  • the process of using the LM algorithm to train the BP neural network model can be performed as the following steps:
  • Step 2 Calculate network output and error index function E(x(k));
  • Step 3 calculating the Jacobian matrix J(x);
  • Step 4 calculate delta_x and E(x(k)) respectively;
  • Step 5 If E(x(k)) ⁇ e, go to step 7; otherwise, calculate x(k+1) and calculate the error index function E(x(k+1)) as the weight and threshold;
  • Step 7 the training ends.
  • the values of all training-related parameters can be set according to the specific training process, and are not specifically limited here.
  • gradient vectors and Jacobian matrices can be used to train the BP neural network model with the LM algorithm.
  • Some embodiments of this specification also provide a vehicle monitoring device, including at least one storage medium and at least one processor, at least one storage medium is used to store computer instructions; at least one processor is used to execute computer instructions to achieve any of the above The method described in the examples.
  • Some embodiments of this specification also provide a computer-readable storage medium, the storage medium stores computer instructions, and after the computer reads the computer instructions in the storage medium, the computer executes the method described in any one of the above-mentioned embodiments.
  • the illustrated system and its modules can be implemented in various ways.
  • the system and its modules may be implemented by hardware, software, or a combination of software and hardware.
  • the hardware part can be implemented by using dedicated logic;
  • the software part can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or specially designed hardware.
  • an appropriate instruction execution system such as a microprocessor or specially designed hardware.
  • processor control code for example on a carrier medium such as a magnetic disk, CD or DVD-ROM, such as a read-only memory (firmware ) or on a data carrier such as an optical or electronic signal carrier.
  • the system and its modules of the present application can not only be implemented by hardware circuits such as very large scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, etc., or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc. , can also be realized by software executed by various types of processors, for example, and can also be realized by a combination of the above-mentioned hardware circuits and software (for example, firmware).
  • the above description of the vehicle monitoring device 600 and its modules is only for convenience of description, and does not limit this specification to the scope of the illustrated embodiments. It can be understood that for those skilled in the art, after understanding the principle of the system, it is possible to combine various modules arbitrarily, or form a subsystem to connect with other modules without departing from this principle.
  • the first acquisition module 610, the second acquisition module 620, and the state determination module 630 may share one storage module, and each module may also have its own storage module. Such deformations are within the protection scope of the present application.
  • vehicle monitoring device, equipment, and computer-readable storage medium provided in the embodiments of this specification belong to the same inventive concept as the embodiments of the vehicle monitoring method.
  • the specific embodiment process can be found in the method embodiments, and will not be repeated here. repeat.
  • the possible beneficial effects of the embodiments of this specification include but are not limited to: (1) By comparing the intervals of the measured data sets of the parts to be tested with the normal data sets, the working conditions of the parts to be tested can be quickly and accurately obtained; (2) ) Use self-selected data sets under a variety of complex working conditions that are in line with the actual operating conditions as training samples, and through a specific model training process, the intelligent machine learning model trained to determine the operating status of the components to be detected, whether it is data processing The accuracy and adaptability are both high, and the monitoring effect of the expected operating status information data can be obtained, the data processing efficiency is improved, and human errors are avoided. (3) The data sets in the preset range are screened accordingly to avoid the interference of external environmental factors. It should be noted that different embodiments may have different beneficial effects. In different embodiments, the possible beneficial effects may be any one or a combination of the above, or any other possible beneficial effects.
  • aspects of this specification can be illustrated and described by several patentable categories or situations, including any new and useful process, machine, product or combination of substances, or any combination of them Any new and useful improvements.
  • various aspects of this specification may be entirely executed by hardware, may be entirely executed by software (including firmware, resident software, microcode, etc.), or may be executed by a combination of hardware and software.
  • the above hardware or software may be referred to as “block”, “module”, “engine”, “unit”, “component” or “system”.
  • aspects of this specification may be embodied as a computer product comprising computer readable program code on one or more computer readable media.
  • a computer storage medium may contain a propagated data signal embodying a computer program code, for example, in baseband or as part of a carrier wave.
  • the propagated signal may have various manifestations, including electromagnetic form, optical form, etc., or a suitable combination.
  • a computer storage medium may be any computer-readable medium, other than a computer-readable storage medium, that can be used to communicate, propagate, or transfer a program for use by being coupled to an instruction execution system, apparatus, or device.
  • Program code residing on a computer storage medium may be transmitted over any suitable medium, including radio, electrical cable, fiber optic cable, RF, or the like, or combinations of any of the foregoing.
  • the computer program codes required for the operation of each part of this manual can be written in any one or more programming languages, including object-oriented programming languages such as Java, Scala, Smalltalk, Eiffel, JADE, Emerald, C++, C#, VB.NET, Python etc., conventional procedural programming languages such as C language, Visual Basic, Fortran2003, Perl, COBOL2002, PHP, ABAP, dynamic programming languages such as Python, Ruby and Groovy, or other programming languages.
  • the program code may run entirely on the user's computer, or as a stand-alone software package, or run partly on the user's computer and partly on a remote computer, or entirely on the remote computer or processing device.
  • the remote computer can be connected to the user computer through any form of network, such as a local area network (LAN) or wide area network (WAN), or to an external computer (such as through the Internet), or in a cloud computing environment, or as a service Use software as a service (SaaS).
  • LAN local area network
  • WAN wide area network
  • SaaS service Use software as a service
  • numbers describing the quantity of components and attributes are used. It should be understood that such numbers used in the description of the embodiments use the modifiers "about”, “approximately” or “substantially” in some examples. grooming. Unless otherwise stated, “about”, “approximately” or “substantially” indicates that the stated figure allows for a variation of ⁇ 20%. Accordingly, in some embodiments, the numerical parameters used in the specification and claims are approximations that can vary depending upon the desired characteristics of individual embodiments. In some embodiments, numerical parameters should take into account the specified significant digits and adopt the general digit reservation method. Although the numerical ranges and parameters used in some embodiments of this specification to confirm the breadth of the range are approximations, in specific embodiments, such numerical values are set as precisely as practicable.

Landscapes

  • Engineering & Computer Science (AREA)
  • Mechanical Engineering (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Sustainable Development (AREA)
  • Sustainable Energy (AREA)
  • Power Engineering (AREA)
  • Transportation (AREA)
  • Testing And Monitoring For Control Systems (AREA)

Abstract

本说明书实施例提供了一种车辆监测方法、装置、设备及计算机可读介质。所述方法包括：获取车辆的待检测部件的运营工况数据，所述运营工况数据至少包括在正常工况下的若干组正常数据集，每组所述正常数据集包括第一运行数据，以及所述第一运行数据对应时刻车辆的第一运动数据；获取所述待检测部件在实际运行中的至少一组实测数据集，每组所述实测数据集包括第二运行数据，以及所述第二运行数据对应时刻所述车辆的第二运动数据；其中，所述第一运行数据和第二运行数据均反映所述待检测部件的运行状态信息，所述第一运动数据和第二运动数据均反映所述车辆的运动状态信息；基于所述运营工况数据和所述实测数据集，确定每组所述实测数据集所反映的所述待检测部件的运行状态。

Description

车辆监测方法、装置、设备及计算机可读存储介质 技术领域
本说明书涉及数据监测技术领域，特别涉及一种车辆监测方法、装置、设备及计算机可读存储介质。
背景技术
随着信息技术的发展，交通运输中车辆越来越多。车辆所产生的数据体量也越来越大，大量数据混杂在一起，想要通过人为监测车辆产生的数据并从中判断车辆的运行信息也越来越困难。
因此，希望提供一种能够对车辆数据进行自动监测的车辆监测方案。
发明内容
本说明书一个方面提供一种车辆监测的方法，所述方法包括：获取车辆的待检测部件的运营工况数据，所述运营工况数据至少包括在正常工况下的若干组正常数据集，每组所述正常数据集包括第一运行数据，以及所述第一运行数据对应时刻车辆的第一运动数据；获取所述待检测部件在实际运行中的至少一组实测数据集，每组所述实测数据集包括第二运行数据，以及所述第二运行数据对应时刻所述车辆的第二运动数据；其中，所述第一运行数据和第二运行数据均反映所述待检测部件的运行状态信息，所述第一运动数据和第二运动数据均反映所述车辆的运动状态信息；基于所述运营工况数据和所述实测数据集，确定每组所述实测数据集所反映的所述待检测部件的运行状态。
本说明书另一个方面提供一种车辆监测装置。所述装置包括：第一获取模块，用于获取车辆的待检测部件的运营工况数据，所述运营工况数据至少包括在正常工况下的若干组正常数据集，每组所述正常数据集包括第一运行数据，以及所述第一运行数据对应时刻车辆的第一运动数据；第二获取模块，用于获取所述待检测部件在实际运行中的至少一组实测数据集，每组所述实测数据集包括第二运行数据，以及所述第二运行数据对应时刻所述车辆的第二运动数据；其中，所述第一运行数据和第二运行数据均反映所述待检测部件的运行状态信息，所述第一运动数据和第二运动数据均反映所述车辆的运动状态信息；状态确定模块，用于基于所述运营工况数据和所述实测数据集，确定每组所述实测数据集所反映的所述待检测部件的运行状态。
本说明书另一个方面提供一种车辆监测设备，包括至少一个存储介质和至少一个处理器，所述至少一个存储介质用于存储计算机指令；所述至少一个处理器用于执行 所述计算机指令以实现如上述任一方案所述的车辆监测方法。
本说明书另一个方面提供计算机可读存储介质，所述存储介质存储计算机指令，当计算机读取存储介质中的计算机指令后，计算机执行上述任一方案所述的车辆监测方法。
本说明书实施例可能带来的有益效果包括但不限于：(1)通过将待检测部件的实测数据集和正常数据集的区间进行对比，可以快速准确地获得待检测部件的工况；(2)通过自选取的符合实际运行状况的多种复杂工况下的数据集作为训练样本，通过特定模型训练过程，训练得到的智能化机器学习模型来确定待检测部件的运行状态，无论是数据处理精确度还是适应性都较高，能够获得预期运行状态信息数据的监测效果，提高了数据处理效率，避免了人为误差；(3)将处于预设区间的数据集进行相应筛选，避免了外部环境因素的干扰。
附图说明
本说明书将以示例性实施例的方式进一步说明，这些示例性实施例将通过附图进行详细描述。这些实施例并非限制性的，在这些实施例中，相同的编号表示相同的结构，其中：
图1是根据本说明书一些实施例所示的车辆监测系统的应用场景示意图；
图2是根据本说明书一些实施例所示的车辆监测方法的示例性流程图；
图3a是根据本说明书的一些实施例所示的正常工况下汽车速度与其对应的电机电流值的示例图；
图3b是根据本说明书的一些实施例所示的实际运行下汽车速度与其对应的电机电流值的示例图；
图4是根据本说明书一些实施例所示的确定待检测部件运行状态的方法的示例性流程图；
图5是根据本说明书另一些实施例所示的确定待检测部件运行状态的方法的示例性流程图；
图6是根据本说明书的一些实施例所示的示例性车辆监测装置的结构示意图。
具体实施方式
为了更清楚地说明本说明书实施例的技术方案，下面将对实施例描述中所需要使用的附图作简单的介绍。显而易见地，下面描述中的附图仅仅是本说明书的一些示例 或实施例，对于本领域的普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图将本说明书应用于其它类似情景。除非从语言环境中显而易见或另做说明，图中相同标号代表相同结构或操作。
应当理解，本说明书中所使用的“系统”、“装置”、“单元”和/或“模组”是用于区分不同级别的不同组件、元件、部件、部分或装配的一种方法。然而，如果其他词语可实现相同的目的，则可通过其他表达来替换所述词语。
如本说明书和权利要求书中所示，除非上下文明确提示例外情形，“一”、“一个”、“一种”和/或“该”等词并非特指单数，也可包括复数。一般说来，术语“包括”与“包含”仅提示包括已明确标识的步骤和元素，而这些步骤和元素不构成一个排它性的罗列，方法或者设备也可能包含其它的步骤或元素。
本说明书中使用了流程图用来说明根据本说明书的实施例的系统所执行的操作。应当理解的是，前面或后面操作不一定按照顺序来精确地执行。相反，可以按照倒序或同时处理各个步骤。同时，也可以将其他操作添加到这些过程中，或从这些过程移除某一步或数步操作。
图1是根据本说明书一些实施例所示的车辆监测系统的应用场景示意图。
场景可以应用于运输系统、交通服务系统。例如，场景可以应用于任意区域的车辆数据的监测。在一些实施例中，车辆监测系统100可以应用于景区、厂区、港口、学校等地区的智能车辆的数据监控。
车辆监测系统100可以是一个线上服务平台，包括服务器110、网络120、终端130、数据库140以及车辆150。该服务器110可以包含处理设备112。
在一些实施例中，服务器110可以用于处理与为车辆监测相关的信息和/或数据。服务器110可以是独立的服务器或者服务器组。该服务器组可以是集中式的或者分布式的(如：服务器110可以是分布系统)。在一些实施例中该服务器110可以是区域的或者远程的。例如，服务器110可通过网络120访问存储于终端130、数据库140、车辆150中的信息和/或资料。在一些实施例中，服务器110可直接与终端130、数据库140、车辆150连接以访问存储于其中的信息和/或资料。在一些实施例中，服务器110可在云平台上执行。例如，该云平台可包括私有云、公共云、混合云、社区云、分散式云、内部云等中的一种或其任意组合。
在一些实施例中，服务器110可包含处理设备112。该处理设备112可处理与车辆监测相关的数据和/或信息以执行一个或多个本申请中描述的功能。例如，处理设 备112可以获取正常数据集和实测数据集。又例如，处理设备112可以基于正常数据集和实测数据集确定待检测部件的运行状态。在一些实施例中，处理设备112可包含一个或多个子处理设备(例如，单芯处理设备或多核多芯处理设备)。仅仅作为范例，处理设备112可包含中央处理器(CPU)、专用集成电路(ASIC)、专用指令处理器(ASIP)、图形处理器(GPU)、物理处理器(PPU)、数字信号处理器(DSP)、现场可编程门阵列(FPGA)、可编辑逻辑电路(PLD)、控制器、微控制器单元、精简指令集电脑(RISC)、微处理器等或以上任意组合。
网络120可促进数据和/或信息的交换。在一些实施例中，车辆监测系统100中的一个或多个组件(例如，服务器110、终端130、数据库140、车辆150)可通过网络120发送数据和/或信息给其他组件。在一些实施例中，网络120可以是任意类型的有线或无线网络。例如，网络120可包括缆线网络、有线网络、光纤网络、电信网络、内部网络、网际网络、区域网络(LAN)、广域网络(WAN)、无线区域网络(WLAN)、都会区域网络(MAN)、公共电话交换网络(PSTN)、蓝牙网络、ZigBee网络、近场通讯(NFC)网络等或以上任意组合。在一些实施例中，网络120可以包括一个或多个网络进出点。例如，网络120可包含有线或无线网络进出点，如基站和/或网际网络交换点120-1、120-2、…，通过这些进出点，系统100的一个或多个组件可连接到网络120上以交换数据和/或信息。
终端130可以用于输入和/或获取数据和/或信息。在一些实施例中，终端130可以包括智能手机130-1、平板电脑130-2、膝上型电脑130-3等。在一些实施例中，终端130可以包括移动终端设备等。例如，移动设备可以包括智能家居设备、可穿戴设备、智能移动设备、虚拟现实设备、增强现实设备等或上述举例的任意组合。在一些实施例中，智能家居设备可以包括智能照明设备、智能电器的控制设备、智能监测设备、智能电视、智能摄影机、对讲机等或上述举例的任意组合。在一些实施例中，可穿戴设备可以包括手环、鞋袜、眼镜、头盔、手表、衣物、背包、智能配饰等或上述举例的任意组合。在一些实施例中，智能移动设备可以包括移动手机、个人数字助理、游戏设备、导航设备、POS机、膝上型电脑、台式电脑等或上述举例的任意组合。
在一些实施例中，用户可以通过终端130获取车辆的运行状态。在一些实施例中，用户可以通过终端130获取车辆的正常数据集和实测数据集。
数据库140可存储资料和/或指令。在一些实施例中，数据库140可存储从终端130获取的资料。在一些实施例中，数据库140可存储供服务器110执行或使用的信息 和/或指令，以执行本申请中描述的示例性方法。在一些实施例中，数据库140可以存储正常数据集、实测数据集、机器学习模型等。在一些实施例中，数据库140可包括大容量存储器、可移动存储器、挥发性读写存储器(例如，随机存取存储器RAM)、只读存储器(ROM)等或以上任意组合。在一些实施例中，数据库140可在云平台上实现。例如，该云平台可包括私有云、公共云、混合云、社区云、分散式云、内部云等或以上任意组合。
车辆150可以包括多种类型的交通工具，例如，单车、电动车、摩托车、汽车、卡车、货车等。在一些实施例中，车辆150可以将获取的数据发送至车辆监测系统100中的一个或多个设备中。例如，车辆150可以将获取的数据通过网络120发送至服务器110进行后续步骤。在一些实施例中，车辆150可以与车辆监测系统100中的一个或多个设备进行通信。例如，车辆150可以通过网络120与终端130进行通信。又例如，车辆150可以直接与终端130进行通信。在一些实施例中，车辆150的数据可以存储在数据库140中。
在一些实施例中，数据库140可与网络120连接以与系统100的一个或多个组件(例如，服务器110、终端130、车辆150等)通讯。系统100的一个或多个组件可通过网络120访问存储于数据库140中的资料或指令。例如，服务器110可以从数据库140中提取正常数据集、实测数据集以及机器学习模型等并进行相应处理。在一些实施例中，数据库140可直接与系统100中的一个或多个组件(如，服务器110、终端130、车辆150)连接或通讯。在一些实施例中，数据库140可以是服务器110的一部分。
图2是根据本说明书一些实施例所示的车辆监测方法的示例性流程图。
如图2所示，流程200可以通过处理设备(如服务器110)、车辆监测系统(如系统100)或车辆监测装置(如装置600)执行。其包括：
步骤210，获取车辆的待检测部件的运营工况数据，运营工况数据至少包括在正常工况下的若干组正常数据集，每组正常数据集包括第一运行数据，以及第一运行数据对应时刻车辆的第一运动数据。
待检测部件为车辆中需要进行检测的部件。在一些实施例中，待检测部件可以包括但不限于电机、发动机、变速器、电控动力转向系统、自动空调系统、自动变速器电控系统等。
运营工况数据是指车辆在一种或多种(至少两种)运营工况下车辆产生的数据。例如，车辆的运营工况包括但不限于正常工况、非正常工况、老化工况、待维修工况等 等。在一些实施例中，运营工况数据可以包括正常工况下的若干组正常数据集。在一些实施例中，运营工况数据可以包括除正常工况下的若干组正常数据集之外的以下至少一种：非正常工况下的非正常数据集、老化工况下的若干组老化数据集、待维修工况下的若干组待维修数据集。
正常数据集为车辆正常工况下产生的数据集，例如，正常数据集可以包括正常状态下车辆的里程表数值、电池仓温度、电池仓湿度、电池容量、传感器运行情况等。又例如，正常数据集还可以包括车辆在某一时刻或某一时间段下的速度、电压值、拉力值、位置、加速度等数据的集合。
一组正常数据集包括一个第一运行数据、以及该第一运行数据对应时刻车辆的一个第一运动数据。在一些实施例中，第一运行数据可以为在正常工况下电机的一个或多个电流值，第一运动数据可以为车辆一个或多个电流值对应时刻的一个或多个速度。在一些实施例中，第一运行数据可以为在正常工况下电机的一个或多个功率值，第一运动数据可以为电机一个或多个功率值对应时刻车辆的一个或多个加速度。在一些实施例中，第一运行数据可以为在正常工况下发动机的一个或多个功率值，第一运动数据可以为车辆一个或多个功率值对应时刻车辆的一个或多个排气量。例如，在车辆在正常工况下行驶至第10分钟时，第一运行数据为电机电流值50A，车辆此时对应的第一运动数据(速度)为60km/h。又例如，车辆在正常工况下在行驶至第30分钟时，第一运行数据为电动机功率值为10kw，车辆此时对应的第一运动数据(速度)为110km/h。
在一些实施例中，汽车在正常工况行驶中会得到多组正常数据集。多组正常数据集可以包括车辆行驶过程中多组第一运行数据和多组第一运动数据。
数据集为车辆在运行时产生的数据的集合。车辆可以为任意类型的车辆，例如可以包括微型车、小型车、中型车、大型车。又例如，车辆可以包括轿车、载货汽车、客车、挂车、非完整车辆、摩托车等。
在一些实施例中，数据集可以包括系统资源数据、硬件状态数据、硬件数据、软件模块运行数据。在一些实施例中，数据集可以包括下位机传感器数据、下位机系统资源数据、下位机传感器状况数据、上位机传感器数据、上位机传感器状况数据、上位机系统资源数据、上位机业务软件运行数据、上位机业务软件资源占用数据。
在一些实施例中，数据集可以采用话题的发布和订阅形式来收集。例如，采集时只需发布相应的数据需求，从而只采集所需的数据。
在一些实施例中，数据集可以包括但不限于车辆运行时刻对应的电流值、电压 值、里程数、加速度、拉力、位置、惯性、汽油液位、光检测和测距等数据。
在一些实施例中，车辆的数据集可以由车辆的传感器或测量单元探测得到。在一些实施例中，数据集可以用于包括云端存储展示分析、人机交互、软件业务、远程问题定位、远程调试、大数据分析。
获取数据集的方式可以为任意可行的方式。在一些实施例中，可以通过读取存储的数据、调用相关接口或其他方式获取数据集。例如，通过网络传输或者直接上传等方式。又例如，车辆可以通过网络上传数据集，处理设备再进行获取。在一些实施例中，获取的数据集可以存储在数据库140中。
运行数据为车辆待检测部件在行驶状态下的数据。运行数据可以包括电机的电流值、电压值、功率。运行数据还可以包括发动机的输出功率、输出能量，扭矩等。关于第一运行数据、第二运行数据、第三运行数据、第四运行数据、第五运行数据为运行数据在不同状态下的数据。
运动数据为车辆运动状态下的数据。例如，运动数据可以包括车辆运动时的速度、加速度、排气量等。关于第一运动数据、第二运动数据、第三运动数据、第四运动数据、第五运动数据为运动数据在不同状态下的数据。
步骤220，获取待检测部件在实际运行中的至少一组实测数据集，每组实测数据集包括第二运行数据，以及第二运行数据对应时刻车辆的第二运动数据；其中，第一运行数据和第二运行数据均反映待检测部件的运行状态信息，第一运动数据和第二运动数据均反映车辆的运动状态信息。
实测数据集为待检测部件在实际运行中测量得到的数据集。实测数据集与正常数据集对应，表示车辆的待检测部件在实际运行的情况。
一组实测数据集包括一个第二运行数据以及该第二运行数据对应时刻车辆的一个第二运动数据。例如，在车辆在实际运行中行驶至第20分钟时，第二运行数据为电机电流值50A，车辆此时对应的第二运动数据(速度)为50km/h。在一些实施例中，汽车在实际运行中会得到多组实测数据集。多组实测数据集可以包括车辆行驶过程中多组第二运行数据和多组第二运动数据。
在一些实施例中，可以根据待检测部件的类型确定运动数据和运行数据的类型。例如，待检测部件为电机，可以确定运行数据的类型为电机电流值或电压值等、运动数据为汽车速度或加速度等。又例如，待检测部件为发动机，可以确定运行数据的类型为发动机的功率、运动数据为汽车加速度。在一些实施例中，实测数据集的第二运行数据 与正常数据集的第一运行数据种类相同，实测数据集的第二运动数据与正常数据集的第二运行数据种类相同。例如，正常数据集的第一运行数据为电机的电流值，则实测数据集的第二运行数据也为电机的电流值。正常数据集的第一运动数据为车辆的速度，则实测数据集的第二运动数据为车辆的速度。又例如，正常数据集的第一运行数据为发动机的功率值，则实测数据集的第二运行数据也为发动机的功率值。正常数据集的第一运动数据为车辆的加速度，则实测数据集的第二运动数据为车辆的加速度。
在一些实施例中，可以根据待检测部件的类型确定实测数据集的获取间隔时长。获取间隔时长为获取实测数据集间隔的时间长度。例如，对于重要程度高的待检测部件可以缩短获取间隔时长，对于重要程度低的待检测部件可以增加获取间隔时长。例如待检测部件为发动机，可以设置实测数据集的获取间隔时长为30秒，表示间隔30秒获取一次实测数据集。又例如，待检测部件为电机，可以设置实测数据集的获取间隔时长为40秒，表示间隔40秒获取一次实测数据集。根据不同类型待检测部件的特点进行实测数据集获取间隔时长的相应设定，能够更加针对性地获取符合不同预期需求的数据集，提高实测数据集获取的效率。
步骤230，基于运营工况数据和实测数据集，确定每组实测数据集所反映的待检测部件的运行状态。
在一些实施例中，可以基于若干组正常数据集，确定每组实测数据集反映的待检测部件的运行状态。多组正常数据集包括多组第一运行数据和第一运动数据。
在一些实施例中，可以分析得到多组正常数据集的每个第一运行数据所对应的多个第一运动数据的最大值和最小值，当第二运行数据与第一运行数据相同时，该第一运动数据的最大值和最小值与一组实测数据集所对应的第二运动数据进行比较。比较后可以得到待检测部件的运行状态。例如，以电机为例，将一组实测数据集的某一速度下的电流值与正常数据集的该速度下的最大或最小电流值进行比较，如果两者的电流值相差超过阈值，则可以认为电机状态异常，如果两者电流值相差未超过阈值，则可以认为电机状态正常。具体的，正常数据集下50km/h对应的电流值包括10A、12A、14A。设定阈值为0.5A，正常数据集50km/h最大的电流值为14A。实测数据集下50km/h对应的电流值为14.2A，与14A相比未超过0.5A的阈值，则可以认为电机状态正常。
在一些实施例中，可以将正常数据集下的每个第一运行数据对应的多个第一运动数据处理得到多个第一运动数据的平均值，将实测数据集的第二运动数据与该平均值比较可以得到待检测部件的运行状态。以发动机为例，可以将正常数据集某一速度下的 若干组功率值处理得到该速度下的平均功率值。将该平均功率值与实测数据集下该速度的功率值作比较，如果两者功率值相差超过阈值，则可以认为发动机状态异常，如果两者功率值相差未超过阈值，则可以认为发动机状态正常。具体的，正常数据集下50km/h对应的功率值包括500kw、600kw、700kw，则正常数据集的50km/h的平均功率值为600kw。设定阈值为150kw。实测数据集下50km/h对应的功率值为440kw，440kw与600kw相比相差160kw，超过了阈值，则可以认为发动机状态异常。
在一些实施例中，可以基于若干组正常数据集，确定每个第一运动数据所对应的第一运行数据的正常数据区间。当第一运动数据与所述第二运动数据相同时，将该第二运动数据对应第二运行数据与该第一运动数据对应的所述正常数据区间进行比较，并获得比较结果。基于比较结果，确定待检测部件的运行状态。具体如图3a和图3b所示，以电机作为运行数据、车辆速度为运动数据作为示例。通过将待检测部件的实测数据集和正常数据集的区间进行匹配对比，可以快速准确地获得待检测部件的运行状态，尤其适用于监测状况不太复杂的工况。
图3a是根据本说明书的一些实施例所示的正常工况下汽车速度与其对应的电机电流值的示例图。图3b是根据本说明书的一些实施例所示的实际运行下汽车速度与其对应的电机电流值的示例图。
在实际运行下，汽车速度在10km/h对应的第二运行数据的值为8A,在正常工况下，汽车速度在10km/h对应的多个第一运行数据的值的区间为2A-11A。在速度都为10km/h时，第二运行数据的值在第一运行数据的正常区间中，故本组实测数据集反映待检测部件正常。
在实际运行下，汽车速度在30km/h对应的第二运行数据的值为20A,在正常工况下,汽车速度在30km/h对应的多个第一运行数据的值的区间为4A-18A。在速度都为30km/h时，第二运行数据的值不在第一运行数据的正常区间中，故本组实测数据集反映待检测部件不正常。
在实际运行下，汽车速度在50km/h对应的第二运行数据的值为20A,在正常工况下,汽车速度在50km/h对应的多个第一运行数据的值的区间为10A-25A。在速度都为50km/h时，第二运行数据的值在第一运行数据的正常区间中，故本组实测数据集反映待检测部件正常。
在一些实施例中，当实测数据集反映待检测部件不正常的数量大于阈值时，则启动相应的检修程序对待检测部件进行检修。例如当电机中实测数据集反映待检测部件 不正常的数量大于10时，则启动相应的检修程序对电机进行检修。
图4是根据本说明书一些实施例所示的确定待检测部件运行状态的方法的示例性流程图。
在一些实施例中，除了正常工况下的若干组正常数据集，运营工况数据还包括非正常工况下的非正常数据集。
如图4所示，流程400可以通过处理设备(如服务器110)、车辆监测系统(如系统100)或车辆监测装置(如装置600)执行。其包括：
步骤410，获取车辆的待检测部件在正常工况下的若干组正常数据集。
关于步骤410与步骤210相同，详细内容参见步骤210，此处不再赘述。
步骤420，获取车辆的待检测部件在非正常工况下的非正常数据集，每组非正常数据集包括第三运行数据，以及第三运行数据对应时刻车辆的第三运动数据。
在一些实施例中，非正常数据集为车辆的待检测部件在非正常工况下产生的数据集。例如待检测部件在故障情况下产生的数据集。在一些实施例中，数据集包括正常数据集和非正常数据集，非正常数据集可以由数据集将正常数据集筛选后得到。
在一些实施例中，非正常数据集包括第三运行数据以及第三运行数据对应时刻车辆的第三运动数据。其中，第三运行数据反映车辆的待监测部件的运行状态信息，第三运动数据反映车辆的运动状态信息。关于运行数据和运动数据的解释可以参见图2相关描述。
待检测部件可以是车辆中任意可行的部件，第一外部数据和第二外部数据可以是能够反映车辆运行环境信息的任意可行数据类型，第一运行数据、第二运行数据、第三运行数据可以是能够反映车辆的待监测部件的运行状态信息的任意可行数据类型。第一运动数据、第二运动数据、第三运动数据可以是能够反映车辆的运动状态信息的任意可行数据类型。
在一些实施例中，待检测部件可以包括电机；第一外部数据可以至少包括道路坡度值和/或车辆载重，例如第一外部数据可以是道路坡度值和车辆载重，第二外部数据可以至少包括道路坡度值和/或车辆载重，例如第二外部数据可以仅是道路坡度值或车辆载重；第一运行数据可以至少包括电机的电流值，第二运行数据可以至少包括电机的电流值，第三运行数据可以至少包括电机的电流值；第一运动数据可以至少包括车辆的行驶速度，第二运动数据可以至少包括车辆的行驶速度，第三运动数据可以至少包括车辆的行驶速度。由于选取的道路坡度值和车辆载重、电机的电流值、车辆的行驶速度 较能反映车辆待检测部件的运行状态，较具有代表性，从而能够更准确高效地进行待检测部件的运行状态的计算监测。
步骤430，将正常数据集作为正样本，将非正常数据集作为负样本，使用正样本和负样本对初始第一机器学习模型进行训练，以获得训练后第一机器学习模型。
在一些实施例中，初始的第一机器学习模型可以为分类器。在一些实施例中，可以使用逻辑回归模型、支持向量机、随机森林或其它分类模型等作为分类器。
初始的第一机器学习模型训练样本为正常数据集和非正常数据集。正样本为正常数据集，标签为正常。负样本为非正常数据集，标签为非正常。
在一些实施例中，数据集还可以包括外部数据。外部数据可以包括第一外部数据和第二外部数据。第一外部数据和第二外部数据均可反映车辆运行环境信息，外部数据在一定程度下可能会干扰最后对待检测部件状态的确定。可以理解的是，外部数据不同时，正常数据集和实测数据集的数据结果也会有相应的误差。
在一些实施例中，外部数据可以包括道路坡度值和/或车辆载重。道路的坡度表示行驶道路与水平地面的角度。例如道路的坡度可以为3度、5度等。车辆载重表示车辆上承载物品的重量，例如车辆载重可以为1吨、3吨等。
在一些实施例中，数据集还可以包括运行数据对应时刻车辆的外部数据。例如电机在电流值为10A时，对应时刻道路的坡度为5度。
在一些实施例中，正常数据集包括第一运行数据对应时刻车辆的第一外部数据。可以将位于第一预设区间内的第一外部数据对应的正常数据集作为正样本。例如，第一外部数据为道路坡度值，第一预设区间为0度到5度，将位于0度到5度的正常数据集作为正样本。
在一些实施例中，非正常数据集包括第二运行数据对应时刻车辆的第二外部数据。可以将位于第二预设区间内的第二外部数据对应的正常数据集作为正样本。例如，第二外部数据为道路坡度值，第二预设区间为0度到5度，将位于0度到5度的非正常数据集作为负样本。
在一些实施例中，第一预设区间可以与第二预设区间相同或不同，可以理解的是，将处于预设区间内的数据集筛选出来可以有效比较外部环境对数据集的干扰。
在一些实施例中，可以采用常用的方法进行训练初始第一机器学习模型，例如梯度下降法。在一些实施例中，还可以使用Levenberg-Marquardt算法(LM算法，也称衰减的最小平方法)对初始机器学习模型进行训练，获得训练后第一机器学习模型。LM 算法比较适用于针对车辆故障点进行甄别的中小型网络，例如BP神经网络模型等。
步骤440，基于实测数据集，使用训练后第一机器学习模型，确定每组实测数据集所反映的待检测部件的运行状态为正常或非正常。
在一些实施例中，将一组实测数据集输入到训练后的第一机器学习模型中，可以得到实测数据集反映的待检测部件的运行状态。例如，实测数据集中待检测部件可以为电机、运行数据可以为电机的电流值。运动数据可以为车辆行驶的速度。可以将电机的电流值10A，车辆的行驶速度为30km/h输入到训练后的第一机器学校模型中得到最后的输出结果为正常，表示待检测部件的运行状态为正常。又例如，可以将电机的电流值20A，车辆的行驶速度为40km/h输入到训练后的第一机器学校模型中得到最后的输出结果为非正常，表示待检测部件的运行状态为非正常。
图5是根据本说明书的另一些实施例所示的确定待检测部件运行状态的方法的示例性流程图。
在一些实施例中，运营工况数据还可以包括老化工况下的若干组老化数据集、待维修工况下的若干组待维修数据集。
如图5所示，流程500可以通过处理设备(如服务器110)、车辆监测系统(如系统100)或车辆监测装置(如装置600)执行。其包括：
步骤510，获取待检测部件在老化工况下的若干组老化数据集，每组老化数据集包括第四运行数据，以及第四运行数据对应时刻车辆的第四运动数据；获取待检测部件在待维修工况下的若干组待维修数据集，每组待维修数据集包括第五运行数据，以及第五运行数据对应时刻车辆的第五运动数据；其中，第四运行数据和第五运行数据均反映待检测部件的运行状态信息，第四运动数据和第五运动数据均反映车辆的运动状态信息。
在一些实施例中，获取车辆的待检测部件在老化工况下的若干组老化数据集，每组老化数据集包括第四运行数据，以及第四运行数据对应时刻车辆的第四运动数据。
获取车辆的待检测部件在待维修工况下的若干组待维修数据集，每组待维修数据集包括第五运行数据，以及第五运行数据对应时刻车辆的第五运动数据。
在一些实施例中，第四运行数据、第五运行数据分别对应着正常数据集第一运行数据，第四运动数据、第五运动数据分别对应着正常数据集的第一运动数据。其表示不同运行状态下的数据。
在一些实施例中，可以利用视觉传感器、电机霍尔传感器、惯性传感器、GPS定 位装置中的至少一种或几种结合，进行运营工况数据和/或所述实测数据集的采集工作。具体地，视觉传感器可以用于获取成像噪声检测数据，以便通过成像噪声检测数据判断视觉传感器前的车辆玻璃部件的运行状态，如车辆玻璃老化情况等；电机霍尔传感器与GPS定位装置的结合设置，可以用于获取速度数据、脉冲异常数据、噪声数据和/或频率数据，以便通过速度数据、脉冲异常数据、噪声数据和/或频率数据判断电机霍尔传感器的运行状态，例如，根据电机霍尔传感器所输出的速度和真实速度是否匹配，再例如，根据噪声数据和频率数据判断脉冲有无异常等，以便判断电机霍尔传感器的健康状态；惯性传感器可以用于获取车辆的定位数据、速度和/或加速度数据，从而为进一步分析确定车辆或车辆某一待检测部件的运行状态提供数据依据。
通过针对性地采用上述数据采集装置或几种的结合，能够更高效地获取期望的特定运营工况数据和/或实测数据集，进而为后续数据分析和计算提供有利数据基础。
步骤520，将正常数据集、老化数据集和待维修数据集作为训练样本，训练初始第二机器学习模型，以获得训练后第二机器学习模型。
初始第二机器学习模型训练样本为正常数据集、老化数据集和待维修数据集，标签分别为正常、老化和待维修，将正常数据集、老化数据集和待维修数据集作为训练样本，使用预设算法对初始机器学习模型进行训练。关于模型介绍、训练方法、具体采用算法等与图4类似，具体参见图4相关描述，此处不再赘述。
在一些实施例中，通过调整初始第二机器学习模型的参数(例如，学习率、迭代次数、批次大小等参数)优化损失函数，当损失函数满足预设条件时，训练结束，得到训练好第二机器学习模型。
在一些实施例中，训练样本采集待检测部件全生命周期内的正常数据集、老化数据集和待维修数据集，由于全生命周期内的数据较能全面反映待检测部件的运行状态，使得训练样本的数据丰富度和代表性较高，采用此种训练样本使用预设算法对初始机器学习模型进行训练，直至达到预期收敛训练结束，自此获得高精确度、高适应性以及高收敛性的训练后机器学习模型。
在一些实施例中，还可以使用LM算法对初始机器学习模型进行训练，获取训练后第二机器学习模型。在一些实施例中，使用多个电机的全生命周期的标注训练样本数据，训练BP神经网络模型，直到其稳定收敛(识别率不再提升)，并且基本合乎预期(可能会和原标注有一定出入，但是可能更为合理)。
在一些实施例中，利用LM算法训练BP神经网络模型的过程，可以执行为以 下步骤：
步骤1、给出训练误差允许值e、系数a,b以及初始化权值和阈值向量x(0)，令k＝0，a＝a0；
步骤2、计算网络输出及误差指标函数E(x(k))；
步骤3、计算Jacobian矩阵J(x)；
步骤4、分别计算delta_x和E(x(k))；
步骤5、若E(x(k))<e，转至步骤7；否则，计算x(k+1)并为权值和阈值，计算误差指标函数E(x(k+1))；
步骤6、若E(x(k+1))<E(x(k))，则令k＝k+1，a＝a/b，回到步骤2；否则这次更新权值和阈值，令x(k+1)＝x(k)，a＝ab，并回到步骤4。
步骤7、训练结束。
其中，训练误差允许值e、系数a,b以及初始化权值和阈值向量x等所有训练涉及参数的取值，可以根据具体训练过程进行相应设定，在此不作特别限定。
在一些实施例中，可以采用梯度向量和雅各布矩阵，进行LM算法对BP神经网络模型的训练过程。
利用上述LM算法对BP神经网络模型的具体训练过程，能够获取预期收敛性的高识别率的训练后机器学习模型，尤其适用于车辆复杂故障点甄别的应用场景。
步骤530，基于实测数据集，使用训练后第二机器学习模型，确定每组实测数据集所反映的待检测部件的运行状态是正常、老化或待维修。
在一些实施例中，可以使用训练后的第二机器学习模型确定每组实测数据集反应的待检测部件的运行状态。例如，将一组实测数据集输入到训练后的第二机器学习模型中，可以得到实测数据集反映的待检测部件的运行状态。在一些实施例中，待检测部件可以为电机、运行数据可以为电机的电流值。运动数据可以为车辆行驶的速度。例如，可以将电机的电流值10A，车辆的行驶速度为10km/h输入到训练后的第一机器学校模型中得到最后的输出结果为老化，表示待检测部件的运行状态为老化。又例如，可以将电机的电流值30A，车辆的行驶速度为10km/h输入到训练后的第一机器学校模型中得到最后的输出结果为待维修，表示待检测部件的运行状态为待维修。
在一些实施例中，记录实时运营场景中车辆的电流、加速度、速度、坡度值及车身承重信息等上传至云端，并以该车投入运营的日期为起点，记录时间，以天为单位生成实测数据集，并输入到训练后机器学习模型中，记录当天的神经网络输出，最后根 据神经网络的多日输出结果，为运营厂商提供车辆维护的指导性建议。示例性地，可以针对视觉传感器的成像噪声检测信息数据，判断视觉传感器前的玻璃老化情况；还可以针对电机霍尔传感器，通过高精度组合导航定位数据(判断电机霍尔传感器所输出的速度和真实速度是否匹配)和电机霍尔传感器脉冲(通过噪声和频率等判断脉冲异常)，最终来判断霍尔传感器的健康状态。
图6是根据本说明书的一些实施例所示的车辆监测装置的结构示意图。
在一些实施例中，车辆监测装置600可以包括第一获取模块610、第二获取模块620、状态确定模块630。这些模块也可以作为应用程序或一组由处理引擎读取和执行的指令实现。此外，模块可以是硬件电路和应用/指令的任何组合。例如，当处理引擎或处理器执行应用程序/一组指令时，模块可以是处理器的一部分。
第一获取模块610，用于获取车辆的待检测部件的运营工况数据，运营工况数据至少包括在正常工况下的若干组正常数据集，每组正常数据集包括第一运行数据，以及第一运行数据对应时刻车辆的第一运动数据；第二获取模块620，用于获取待检测部件在实际运行中的至少一组实测数据集，每组实测数据集包括第二运行数据，以及第二运行数据对应时刻车辆的第二运动数据；其中，第一运行数据和第二运行数据均反映待检测部件的运行状态信息，第一运动数据和第二运动数据均反映车辆的运动状态信息；状态确定模块630，用于基于运营工况数据和实测数据集，确定每组实测数据集所反映的待检测部件的运行状态。
在一些实施例中，第二获取模块620用于：根据待检测部件的类型，确定实测数据集的获取间隔时长；间隔获取间隔时长获取每组实测数据集。
在一些实施例中，状态确定模块630用于：基于若干组正常数据集，确定每第一运动数据所对应的第一运行数据的正常数据区间；基于实测数据集和正常数据区间，确定第一运动数据与第二运动数据相同时，将该第二运动数据对应第二运行数据与该第一运动数据对应的正常数据区间进行比较，并获得比较结果；基于比较结果，确定待检测部件的运行状态。
在一些实施例中，运营工况数据还包括非正常工况下的非正常数据集，状态确定模块630用于：获取待检测部件在非正常工况下的非正常数据集，每组非正常数据集包括第三运行数据，以及第三运行数据对应时刻车辆的第三运动数据；其中，第三运行数据反映待检测部件的运行状态信息，第三运动数据反映车辆的运动状态信息；将正常数据集作为正样本，将非正常数据集作为负样本，使用正样本和负样本对初始第一机器 学习模型进行训练，获得训练后第一机器学习模型；基于实测数据集，使用训练后第一机器学习模型，确定每组实测数据集所反映的待检测部件的运行状态为正常或非正常。
在一些实施例中，每组正常数据集还包括第一运行数据对应时刻车辆的第一外部数据，每组非正常数据集还包括第三运行数据对应时刻车辆的第二外部数据，第一外部数据和第二外部数据均反映车辆的运行环境信息；将正常数据集作为正样本，将非正常数据集作为负样本，包括：将位于第一预设区间内的第一外部数据对应的正常数据集作为正样本；将位于第二预设区间内的第二外部数据对应的非正常数据集作为负样本。
在一些实施例中，待检测部件包括电机，第一外部数据和第二外部数据中的每个至少包括道路坡度值和/或车辆载重，第一运行数据、第二运行数据和第三运行数据中的每个至少包括电机的电流值，第一运动数据、第二运动数据和第三运动数据中的每个至少包括车辆的行驶速度。
在一些实施例中，运营工况数据还包括老化工况下的若干组老化数据集、待维修工况下的若干组待维修数据集，状态确定模块630用于：获取待检测部件在老化工况下的若干组老化数据集，每组老化数据集包括第四运行数据，以及第四运行数据对应时刻车辆的第四运动数据；获取待检测部件在待维修工况下的若干组待维修数据集，每组待维修数据集包括第五运行数据，以及第五运行数据对应时刻车辆的第五运动数据；其中，第四运行数据和第五运行数据均反映待检测部件的运行状态信息，第四运动数据和第五运动数据均反映车辆的运动状态信息；将正常数据集、老化数据集和待维修数据集作为训练样本，训练初始第二机器学习模型，以获得训练后第二机器学习模型；基于实测数据集，使用训练后第二机器学习模型，确定每组实测数据集所反映的待检测部件的运行状态是正常、老化或待维修。
在一些实施例中，可以利用视觉传感器、电机霍尔传感器、惯性传感器、GPS定位装置中的至少一种或几种结合，进行运营工况数据和/或所述实测数据集的采集工作。具体地，视觉传感器可以用于获取成像噪声检测数据，以便通过成像噪声检测数据判断视觉传感器前的车辆玻璃部件的运行状态，如车辆玻璃老化情况等；电机霍尔传感器与GPS定位装置的结合设置，可以用于获取速度数据、脉冲异常数据、噪声数据和/或频率数据，以便通过速度数据、脉冲异常数据、噪声数据和/或频率数据判断电机霍尔传感器的运行状态，例如，根据电机霍尔传感器所输出的速度和真实速度是否匹配，再例如，根据噪声数据和频率数据判断脉冲有无异常等，以便判断电机霍尔传感器的健康状态；惯性传感器可以用于获取车辆的定位数据、速度和/或加速度数据，从而为进一步分析 确定车辆或车辆某一待检测部件的运行状态提供数据依据。
在一些实施例中，将正常数据集、老化数据集和待维修数据集作为训练样本，训练初始机器学习模型包括：将正常数据集、老化数据集和待维修数据集作为训练样本，使用预设算法对初始机器学习模型进行训练。
在一些实施例中，还可以使用LM算法对初始机器学习模型进行训练，获取训练后第二机器学习模型。在一些实施例中，使用多个电机的全生命周期的标注训练样本数据，训练BP神经网络模型，直到其稳定收敛(识别率不再提升)，并且基本合乎预期(可能会和原标注有一定出入，但是可能更为合理)。
在一些实施例中，利用LM算法训练BP神经网络模型的过程，可以执行为以下步骤：
步骤1、给出训练误差允许值e、系数a,b以及初始化权值和阈值向量x(0)，令k＝0，a＝a0；
步骤2、计算网络输出及误差指标函数E(x(k))；
步骤3、计算Jacobian矩阵J(x)；
步骤4、分别计算delta_x和E(x(k))；
步骤5、若E(x(k))<e，转至步骤7；否则，计算x(k+1)并为权值和阈值，计算误差指标函数E(x(k+1))；
步骤6、若E(x(k+1))<E(x(k))，则令k＝k+1，a＝a/b，回到步骤2；否则这次更新权值和阈值，令x(k+1)＝x(k)，a＝ab，并回到步骤4。
步骤7、训练结束。
其中，训练误差允许值e、系数a,b以及初始化权值和阈值向量x等所有训练涉及参数的取值，可以根据具体训练过程进行相应设定，在此不作特别限定。
在一些实施例中，可以采用梯度向量和雅各布矩阵，进行LM算法对BP神经网络模型的训练过程。
本说明书的一些实施例还提供了一种车辆监测设备，包括至少一个存储介质和至少一个处理器，至少一个存储介质用于存储计算机指令；至少一个处理器用于执行计算机指令以实现如上述任一实施例所述的方法。
本说明书的一些实施例还提供了一种计算机可读存储介质，存储介质存储计算机指令，当计算机读取存储介质中的计算机指令后，计算机执行如上述任一实施例所述的方法。
应当理解，所示的系统及其模块可以利用各种方式来实现。例如，在一些实施例中，系统及其模块可以通过硬件、软件或者软件和硬件的结合来实现。其中，硬件部分可以利用专用逻辑来实现；软件部分则可以存储在存储器中，由适当的指令执行系统，例如微处理器或者专用设计硬件来执行。本领域技术人员可以理解上述的方法和系统可以使用计算机可执行指令和/或包含在处理器控制代码中来实现，例如在诸如磁盘、CD或DVD-ROM的载体介质、诸如只读存储器(固件)的可编程的存储器或者诸如光学或电子信号载体的数据载体上提供了这样的代码。本申请的系统及其模块不仅可以有诸如超大规模集成电路或门阵列、诸如逻辑芯片、晶体管等的半导体、或者诸如现场可编程门阵列、可编程逻辑设备等的可编程硬件设备的硬件电路实现，也可以用例如由各种类型的处理器所执行的软件实现，还可以由上述硬件电路和软件的结合(例如，固件)来实现。
需要注意的是，以上对于车辆监测装置600及其模块的描述，仅为描述方便，并不能把本说明书限制在所举实施例范围之内。可以理解，对于本领域的技术人员来说，在了解该系统的原理后，可能在不背离这一原理的情况下，对各个模块进行任意组合，或者构成子系统与其他模块连接。例如，第一获取模块610、第二获取模块620、状态确定模块630、可以共用一个存储模块，各个模块也可以分别具有各自的存储模块。诸如此类的变形，均在本申请的保护范围之内。
另外，需要说明的是，本说明书实施例提供的车辆监测装置、设备、计算机可读存储介质与车辆监测方法实施例属于同一发明构思，具体实施例过程可详见方法实施例，在此不再赘述。
本说明书实施例可能带来的有益效果包括但不限于：(1)通过将待检测部件的实测数据集和正常数据集的区间进行对比，可以快速准确地获得待检测部件的工况；(2)通过自选取的符合实际运行状况的多种复杂工况下的数据集作为训练样本，通过特定模型训练过程，训练得到的智能化机器学习模型来确定待检测部件的运行状态，无论是数据处理精确度还是适应性都较高，能够获得预期运行状态信息数据的监测效果，提高了数据处理效率，避免了人为误差。(3)将处于预设区间的数据集进行相应筛选，避免了外部环境因素的干扰。需要说明的是，不同实施例可能产生的有益效果不同，在不同的实施例里，可能产生的有益效果可以是以上任意一种或几种的组合，也可以是其他任何可能获得的有益效果。
上文已对基本概念做了描述，显然，对于本领域技术人员来说，上述详细披露 仅仅作为示例，而并不构成对本说明书的限定。虽然此处并没有明确说明，本领域技术人员可能会对本说明书进行各种修改、改进和修正。该类修改、改进和修正在本说明书中被建议，所以该类修改、改进、修正仍属于本说明书示范实施例的精神和范围。
同时，本说明书使用了特定词语来描述本说明书的实施例。如“一个实施例”、“一实施例”、和/或“一些实施例”意指与本说明书至少一个实施例相关的某一特征、结构或特点。因此，应强调并注意的是，本说明书中在不同位置两次或多次提及的“一实施例”或“一个实施例”或“一个替代性实施例”并不一定是指同一实施例。此外，本说明书的一个或多个实施例中的某些特征、结构或特点可以进行适当的组合。
此外，本领域技术人员可以理解，本说明书的各方面可以通过若干具有可专利性的种类或情况进行说明和描述，包括任何新的和有用的工序、机器、产品或物质的组合，或对他们的任何新的和有用的改进。相应地，本说明书的各个方面可以完全由硬件执行、可以完全由软件(包括固件、常驻软件、微码等)执行、也可以由硬件和软件组合执行。以上硬件或软件均可被称为“数据块”、“模块”、“引擎”、“单元”、“组件”或“系统”。此外，本说明书的各方面可能表现为位于一个或多个计算机可读介质中的计算机产品，该产品包括计算机可读程序编码。
计算机存储介质可能包含一个内含有计算机程序编码的传播数据信号，例如在基带上或作为载波的一部分。该传播信号可能有多种表现形式，包括电磁形式、光形式等，或合适的组合形式。计算机存储介质可以是除计算机可读存储介质之外的任何计算机可读介质，该介质可以通过连接至一个指令执行系统、装置或设备以实现通讯、传播或传输供使用的程序。位于计算机存储介质上的程序编码可以通过任何合适的介质进行传播，包括无线电、电缆、光纤电缆、RF、或类似介质，或任何上述介质的组合。
本说明书各部分操作所需的计算机程序编码可以用任意一种或多种程序语言编写，包括面向对象编程语言如Java、Scala、Smalltalk、Eiffel、JADE、Emerald、C++、C#、VB.NET、Python等，常规程序化编程语言如C语言、Visual Basic、Fortran2003、Perl、COBOL2002、PHP、ABAP，动态编程语言如Python、Ruby和Groovy，或其他编程语言等。该程序编码可以完全在用户计算机上运行、或作为独立的软件包在用户计算机上运行、或部分在用户计算机上运行部分在远程计算机运行、或完全在远程计算机或处理设备上运行。在后种情况下，远程计算机可以通过任何网络形式与用户计算机连接，比如局域网(LAN)或广域网(WAN)，或连接至外部计算机(例如通过因特网)，或在云计算环境中，或作为服务使用如软件即服务(SaaS)。
此外，除非权利要求中明确说明，本说明书所述处理元素和序列的顺序、数字字母的使用、或其他名称的使用，并非用于限定本说明书流程和方法的顺序。尽管上述披露中通过各种示例讨论了一些目前认为有用的发明实施例，但应当理解的是，该类细节仅起到说明的目的，附加的权利要求并不仅限于披露的实施例，相反，权利要求旨在覆盖所有符合本说明书实施例实质和范围的修正和等价组合。例如，虽然以上所描述的系统组件可以通过硬件设备实现，但是也可以只通过软件的解决方案得以实现，如在现有的处理设备或移动设备上安装所描述的系统。
同理，应当注意的是，为了简化本说明书披露的表述，从而帮助对一个或多个发明实施例的理解，前文对本说明书实施例的描述中，有时会将多种特征归并至一个实施例、附图或对其的描述中。但是，这种披露方法并不意味着本说明书对象所需要的特征比权利要求中提及的特征多。实际上，实施例的特征要少于上述披露的单个实施例的全部特征。
一些实施例中使用了描述成分、属性数量的数字，应当理解的是，此类用于实施例描述的数字，在一些示例中使用了修饰词“大约”、“近似”或“大体上”来修饰。除非另外说明，“大约”、“近似”或“大体上”表明所述数字允许有±20％的变化。相应地，在一些实施例中，说明书和权利要求中使用的数值参数均为近似值，该近似值根据个别实施例所需特点可以发生改变。在一些实施例中，数值参数应考虑规定的有效数位并采用一般位数保留的方法。尽管本说明书一些实施例中用于确认其范围广度的数值域和参数为近似值，在具体实施例中，此类数值的设定在可行范围内尽可能精确。
针对本说明书引用的每个专利、专利申请、专利申请公开物和其他材料，如文章、书籍、说明书、出版物、文档等，特此将其全部内容并入本说明书作为参考。与本说明书内容不一致或产生冲突的申请历史文件除外，对本说明书权利要求最广范围有限制的文件(当前或之后附加于本说明书中的)也除外。需要说明的是，如果本说明书附属材料中的描述、定义、和/或术语的使用与本说明书所述内容有不一致或冲突的地方，以本说明书的描述、定义和/或术语的使用为准。
最后，应当理解的是，本说明书中所述实施例仅用以说明本说明书实施例的原则。其他的变形也可能属于本说明书的范围。因此，作为示例而非限制，本说明书实施例的替代配置可视为与本说明书的教导一致。相应地，本说明书的实施例不仅限于本说明书明确介绍和描述的实施例。

Claims (18)

  1. 一种车辆监测方法，其特征在于，包括：
    获取车辆的待检测部件的运营工况数据，所述运营工况数据至少包括在正常工况下的若干组正常数据集，每组所述正常数据集包括第一运行数据，以及所述第一运行数据对应时刻车辆的第一运动数据；
    获取所述待检测部件在实际运行中的至少一组实测数据集，每组所述实测数据集包括第二运行数据，以及所述第二运行数据对应时刻所述车辆的第二运动数据；其中，所述第一运行数据和第二运行数据均反映所述待检测部件的运行状态信息，所述第一运动数据和第二运动数据均反映所述车辆的运动状态信息；
    基于所述运营工况数据和所述实测数据集，确定每组所述实测数据集所反映的所述待检测部件的运行状态。
  2. 如权利要求1所述的方法，其特征在于，所述基于所述运营工况数据和所述实测数据集，确定每组所述实测数据集所反映的所述待检测部件的运行状态，包括：
    基于所述若干组正常数据集，确定每个所述第一运动数据所对应的所述第一运行数据的正常数据区间；
    基于所述实测数据集和所述正常数据区间，确定所述第一运动数据与所述第二运动数据相同时，将该第二运动数据对应第二运行数据与该第一运动数据对应的所述正常数据区间进行比较，并获得比较结果；
    基于所述比较结果，确定所述待检测部件的运行状态。
  3. 如权利要求1所述的方法，其特征在于，所述运营工况数据还包括非正常工况下的非正常数据集，基于所述运营工况数据和所述实测数据集，确定每组所述实测数据集所反映的所述待检测部件的运行状态，包括：
    获取所述待检测部件在所述非正常工况下的所述非正常数据集，每组所述非正常数据集包括第三运行数据，以及所述第三运行数据对应时刻所述车辆的第三运动数据；其中，所述第三运行数据反映所述待检测部件的运行状态信息，所述第三运动数据反映所述车辆的运动状态信息；
    将所述正常数据集作为正样本，将所述非正常数据集作为负样本，使用所述正样本和所述负样本对初始第一机器学习模型进行训练，获得训练后第一机器学习模型；
    基于所述实测数据集，使用所述训练后第一机器学习模型，确定每组所述实测数据 集所反映的所述待检测部件的运行状态为正常或非正常。
  4. 如权利要求3所述的方法，其特征在于，每组所述正常数据集还包括所述第一运行数据对应时刻所述车辆的第一外部数据，每组所述非正常数据集还包括所述第三运行数据对应时刻车辆的第二外部数据，所述第一外部数据和第二外部数据均反映所述车辆的运行环境信息；
    所述将所述正常数据集作为正样本，将所述非正常数据集作为负样本，包括：
    将位于第一预设区间内的所述第一外部数据对应的所述正常数据集作为正样本；
    将位于第二预设区间内的所述第二外部数据对应的所述非正常数据集作为负样本。
  5. 如权利要求4所述的方法，其特征在于，所述待检测部件包括电机，所述第一外部数据和所述第二外部数据中的每个至少包括道路坡度值和/或车辆载重，所述第一运行数据、所述第二运行数据和所述第三运行数据中的每个至少包括所述电机的电流值，所述第一运动数据、所述第二运动数据和所述第三运动数据中的每个至少包括所述车辆的行驶速度。
  6. 如权利要求1所述的方法，其特征在于，所述运营工况数据还包括老化工况下的若干组老化数据集、待维修工况下的若干组待维修数据集，基于所述运营工况数据和所述实测数据集，确定每组所述实测数据集所反映的所述待检测部件的运行状态，包括：
    获取所述待检测部件在所述老化工况下的所述若干组老化数据集，每组所述老化数据集包括第四运行数据，以及所述第四运行数据对应时刻所述车辆的第四运动数据；
    获取所述待检测部件在所述待维修工况下的所述若干组待维修数据集，每组所述待维修数据集包括第五运行数据，以及所述第五运行数据对应时刻所述车辆的第五运动数据；其中，所述第四运行数据和所述第五运行数据均反映所述待检测部件的运行状态信息，所述第四运动数据和所述第五运动数据均反映所述车辆的运动状态信息；
    将所述正常数据集、所述老化数据集和所述待维修数据集作为训练样本，训练初始第二机器学习模型，以获得训练后第二机器学习模型；
    基于所述实测数据集，使用所述训练后第二机器学习模型，确定每组所述实测数据集所反映的所述待检测部件的运行状态是正常、老化或待维修。
  7. 如权利要求6所述的方法，其特征在于，将所述正常数据集、所述老化数据集和所述待维修数据集作为训练样本，训练初始机器学习模型包括：
    将所述正常数据集、所述老化数据集和所述待维修数据集作为训练样本，使用预设算法对所述初始机器学习模型进行训练。
  8. 如权利要求1所述的方法，其特征在于，所述获取所述待检测部件在实际运行中的至少一组实测数据集，包括：
    根据所述待检测部件的类型，确定所述实测数据集的获取间隔时长；
    间隔所述获取间隔时长获取每组所述实测数据集。
  9. 一种车辆监测装置，其特征在于，所述装置包括：
    第一获取模块，用于获取车辆的待检测部件的运营工况数据，所述运营工况数据至少包括在正常工况下的若干组正常数据集，每组所述正常数据集包括第一运行数据，以及所述第一运行数据对应时刻车辆的第一运动数据；
    第二获取模块，用于获取所述待检测部件在实际运行中的至少一组实测数据集，每组所述实测数据集包括第二运行数据，以及所述第二运行数据对应时刻所述车辆的第二运动数据；其中，所述第一运行数据和第二运行数据均反映所述待检测部件的运行状态信息，所述第一运动数据和第二运动数据均反映所述车辆的运动状态信息；
    状态确定模块，用于基于所述运营工况数据和所述实测数据集，确定每组所述实测数据集所反映的所述待检测部件的运行状态。
  10. 如权利要求9所述的装置，其特征在于，所述状态确定模块用于：
    基于所述若干组正常数据集，确定每个所述第一运动数据所对应的所述第一运行数据的正常数据区间；
    基于所述实测数据集和所述正常数据区间，确定所述第一运动数据与所述第二运动数据相同时，将该第二运动数据对应第二运行数据与该第一运动数据对应的所述正常数据区间进行比较，并获得比较结果；
    基于所述比较结果，确定所述待检测部件的运行状态。
  11. 如权利要求9所述的装置，其特征在于，所述运营工况数据还包括非正常工况 下的非正常数据集，所述状态确定模块用于：
    获取所述待检测部件在所述非正常工况下的所述非正常数据集，每组所述非正常数据集包括第三运行数据，以及所述第三运行数据对应时刻所述车辆的第三运动数据；其中，所述第三运行数据反映所述待检测部件的运行状态信息，所述第三运动数据反映所述车辆的运动状态信息；
    将所述正常数据集作为正样本，将所述非正常数据集作为负样本，使用所述正样本和所述负样本对初始第一机器学习模型进行训练，获得训练后第一机器学习模型；
    基于所述实测数据集，使用所述训练后第一机器学习模型，确定每组所述实测数据集所反映的所述待检测部件的运行状态为正常或非正常。
  12. 如权利要求11所述的装置，其特征在于，每组所述正常数据集还包括所述第一运行数据对应时刻所述车辆的第一外部数据，每组所述非正常数据集还包括所述第三运行数据对应时刻车辆的第二外部数据，所述第一外部数据和第二外部数据均反映所述车辆的运行环境信息；
    所述将所述正常数据集作为正样本，将所述非正常数据集作为负样本，包括：
    将位于第一预设区间内的所述第一外部数据对应的所述正常数据集作为正样本；
    将位于第二预设区间内的所述第二外部数据对应的所述非正常数据集作为负样本。
  13. 如权利要求12所述的装置，其特征在于，所述待检测部件包括电机，所述第一外部数据和所述第二外部数据中的每个至少包括道路坡度值和/或车辆载重，所述第一运行数据、所述第二运行数据和所述第三运行数据中的每个至少包括所述电机的电流值，所述第一运动数据、所述第二运动数据和所述第三运动数据中的每个至少包括所述车辆的行驶速度。
  14. 如权利要求9所述的装置，其特征在于，所述运营工况数据还包括老化工况下的若干组老化数据集、待维修工况下的若干组待维修数据集，所述状态确定模块用于：
    获取所述待检测部件在所述老化工况下的所述若干组老化数据集，每组所述老化数据集包括第四运行数据，以及所述第四运行数据对应时刻所述车辆的第四运动数据；
    获取所述待检测部件在所述待维修工况下的所述若干组待维修数据集，每组所述待维修数据集包括第五运行数据，以及所述第五运行数据对应时刻所述车辆的第五运动数 据；其中，所述第四运行数据和所述第五运行数据均反映所述待检测部件的运行状态信息，所述第四运动数据和所述第五运动数据均反映所述车辆的运动状态信息；
    将所述正常数据集、所述老化数据集和所述待维修数据集作为训练样本，训练初始第二机器学习模型，以获得训练后第二机器学习模型；
    基于所述实测数据集，使用所述训练后第二机器学习模型，确定每组所述实测数据集所反映的所述待检测部件的运行状态是正常、老化或待维修。
  15. 如权利要求14所述的装置，其特征在于，将所述正常数据集、所述老化数据集和所述待维修数据集作为训练样本，训练初始机器学习模型包括：
    将所述正常数据集、所述老化数据集和所述待维修数据集作为训练样本，使用预设算法对所述初始机器学习模型进行训练。
  16. 如权利要求9所述的装置，其特征在于，所述第二获取模块用于：
    根据所述待检测部件的类型，确定所述实测数据集的获取间隔时长；
    间隔所述获取间隔时长获取每组所述实测数据集。
  17. 一种车辆监测设备，包括至少一个存储介质和至少一个处理器，所述至少一个存储介质用于存储计算机指令；所述至少一个处理器用于执行所述计算机指令以实现如权利要求1～8任一项所述的方法。
  18. 一种计算机可读存储介质，其特征在于，所述存储介质存储计算机指令，当计算机读取存储介质中的计算机指令后，计算机执行如权利要求1～8任一项所述的方法。
PCT/CN2021/094788 2021-05-20 2021-05-20 车辆监测方法、装置、设备及计算机可读存储介质 WO2022241705A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
PCT/CN2021/094788 WO2022241705A1 (zh) 2021-05-20 2021-05-20 车辆监测方法、装置、设备及计算机可读存储介质
CN202180092815.5A CN116963936A (zh) 2021-05-20 2021-05-20 车辆监测方法、装置、设备及计算机可读存储介质

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2021/094788 WO2022241705A1 (zh) 2021-05-20 2021-05-20 车辆监测方法、装置、设备及计算机可读存储介质

Publications (1)

Publication Number Publication Date
WO2022241705A1 true WO2022241705A1 (zh) 2022-11-24

Family

ID=84140068

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/094788 WO2022241705A1 (zh) 2021-05-20 2021-05-20 车辆监测方法、装置、设备及计算机可读存储介质

Country Status (2)

Country Link
CN (1) CN116963936A (zh)
WO (1) WO2022241705A1 (zh)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116778308A (zh) * 2023-08-23 2023-09-19 北京阿帕科蓝科技有限公司 对象识别方法、装置、计算机设备和存储介质
CN116861370A (zh) * 2023-09-05 2023-10-10 山东千颐科技有限公司 用于井下防爆胶轮车的运动数据处理方法
CN118260675A (zh) * 2024-05-29 2024-06-28 江西五十铃汽车有限公司 一种dpf系统强相关工况分析方法
WO2024150089A1 (en) 2023-01-13 2024-07-18 Bsq Holding S.R.L. System for monitoring and managing a company fleet of transport vehicles

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117622177B (zh) * 2024-01-23 2024-05-14 青岛创新奇智科技集团股份有限公司 一种基于工业大模型的车辆数据处理方法及装置

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5734319A (en) * 1995-03-02 1998-03-31 Sumitomo Rubber Industries, Ltd. Method of determining the inflation pressure of a tire on a moving vehicle
CN1529663A (zh) * 2001-02-26 2004-09-15 �����Զ�����ʽ���� 用于估计轮胎状态的装置和用于判定轮胎的异常状态的装置
JP2010012934A (ja) * 2008-07-03 2010-01-21 Yokohama Rubber Co Ltd:The 車両の運転状態監視方法及びその装置
CN106218329A (zh) * 2016-07-26 2016-12-14 成都之达科技有限公司 基于无线传感器的车辆参数采集方法
CN110135273B (zh) * 2019-04-19 2020-05-22 中铁第一勘察设计院集团有限公司 接触网视频图像云端智能监测与故障识别方法
CN111582338A (zh) * 2020-04-22 2020-08-25 国网江苏省电力有限公司苏州供电分公司 一种电动汽车充电设施运行状态评估方法和系统
JP2020196354A (ja) * 2019-06-03 2020-12-10 横浜ゴム株式会社 タイヤ故障予知システム、タイヤ故障予知方法

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5734319A (en) * 1995-03-02 1998-03-31 Sumitomo Rubber Industries, Ltd. Method of determining the inflation pressure of a tire on a moving vehicle
CN1529663A (zh) * 2001-02-26 2004-09-15 �����Զ�����ʽ���� 用于估计轮胎状态的装置和用于判定轮胎的异常状态的装置
JP2010012934A (ja) * 2008-07-03 2010-01-21 Yokohama Rubber Co Ltd:The 車両の運転状態監視方法及びその装置
CN106218329A (zh) * 2016-07-26 2016-12-14 成都之达科技有限公司 基于无线传感器的车辆参数采集方法
CN110135273B (zh) * 2019-04-19 2020-05-22 中铁第一勘察设计院集团有限公司 接触网视频图像云端智能监测与故障识别方法
JP2020196354A (ja) * 2019-06-03 2020-12-10 横浜ゴム株式会社 タイヤ故障予知システム、タイヤ故障予知方法
CN111582338A (zh) * 2020-04-22 2020-08-25 国网江苏省电力有限公司苏州供电分公司 一种电动汽车充电设施运行状态评估方法和系统

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2024150089A1 (en) 2023-01-13 2024-07-18 Bsq Holding S.R.L. System for monitoring and managing a company fleet of transport vehicles
CN116778308A (zh) * 2023-08-23 2023-09-19 北京阿帕科蓝科技有限公司 对象识别方法、装置、计算机设备和存储介质
CN116778308B (zh) * 2023-08-23 2024-09-13 北京阿帕科蓝科技有限公司 对象识别方法、装置、计算机设备和存储介质
CN116861370A (zh) * 2023-09-05 2023-10-10 山东千颐科技有限公司 用于井下防爆胶轮车的运动数据处理方法
CN116861370B (zh) * 2023-09-05 2023-12-01 山东千颐科技有限公司 用于井下防爆胶轮车的运动数据处理方法
CN118260675A (zh) * 2024-05-29 2024-06-28 江西五十铃汽车有限公司 一种dpf系统强相关工况分析方法

Also Published As

Publication number Publication date
CN116963936A (zh) 2023-10-27

Similar Documents

Publication Publication Date Title
WO2022241705A1 (zh) 车辆监测方法、装置、设备及计算机可读存储介质
WO2020015526A1 (zh) 驾驶行为评分方法及装置
CN110091751B (zh) 基于深度学习的电动汽车续航里程预测方法、设备及介质
CN111862585B (zh) 用于交通预测的系统和方法
CN110914837B (zh) 确定车头朝向的方法和系统
CN109791731B (zh) 一种预估到达时间的方法和系统
CN104246516B (zh) 一种确定车辆加速度的方法及装置
CN109429520B (zh) 用于检查作弊服务订单的方法、系统、设备及可读介质
US20200064230A1 (en) System for the assessment of an object
CN112345956A (zh) 一种电池组荷电状态检测方法和装置
WO2021088216A1 (zh) 电池组荷电状态数据检测方法、装置、介质、终端及汽车
CN109720207B (zh) 车辆能耗分析方法、装置及计算机可读介质
CN114940132B (zh) 电动车续航里程预测方法、测试方法和系统
CN113642242A (zh) 一种基于数字孪生的智能公交车交通仿真平台
CN113269132A (zh) 一种基于yolov4优化算法的车辆检测方法及系统
CN109154523A (zh) 一种车辆载重的测量方法、装置、终端设备及存储介质
CN116011995A (zh) 车辆保养信息推送方法、装置、电子设备及存储介质
Yue et al. Estimation of the vehicle's centre of gravity based on a braking model
CN110225446B (zh) 一种识别驾驶行为的系统、方法、装置和存储介质
TWI697859B (zh) 運輸工具價值評估的系統和方法
CN106651133B (zh) 一种使用手机实现基于adas的用户驾驶行为评分方法
Sik et al. Comparing OBD and CAN Sampling on the go with the SensorHUB Framework
US11270525B2 (en) Automated vehicle occupancy detection
CN117169744A (zh) 一种电动汽车的剩余电量估计方法、系统、设备及介质
CN109919293B (zh) 一种危险驾驶判定方法及装置

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: 21940161

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 202180092815.5

Country of ref document: CN

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21940161

Country of ref document: EP

Kind code of ref document: A1