Early warning method and device, network equipment, electric energy equipment and medium

Abstract

The invention discloses an early warning method and device, network equipment, electric energy equipment and a medium. The method comprises the following steps: performing correlation analysis on sample production data based on a preset sub-assembly sample and sample operation fault data to obtain production fault correlation data; and if the production fault correlation data is greater than a preset threshold value, establishing a mode of determining an early warning index according to sample production data corresponding to the sample operation fault data. And then, historical production data characteristics of the sub-assemblies in the production process are obtained, and if the historical production data characteristics of the sub-assemblies are matched with the early warning indexes, early warning prompt information is sent out, so that potential equipment fault risks are found in advance in the equipment sub-assemblies which do not show related abnormity, and the early warning prompt information is sent out.

Description

Technical Field

[0001] The present invention relates to the field of battery technology, and in particular to an early warning method, device, network equipment, electric energy equipment and medium. Background Art

[0002] In the manufacturing industry, when the production of equipment sub-assemblies is completed, various tests are often carried out to distinguish qualified sub-assemblies from unqualified sub-assemblies. For example, in the relevant technology of the power battery industry, in order to distinguish faulty batteries from qualified batteries, the relevant parameters of the battery need to be tested when the battery is produced to screen out faulty batteries with relevant parameters outside the standard range, and the batteries with relevant parameters within the standard range are identified as qualified batteries that can be used. However, when the qualified batteries in the relevant technology are used, some unexpected failures may still occur. When the qualified batteries in the relevant technology have a use failure, the current common treatment method is: all batteries in the same batch or all electric equipment using the batch of batteries are recycled, and the batteries in the batch are identified as faulty batteries; or, only the batteries that have failed are processed. It can be seen that in the relevant technology, for batteries or electric energy equipment that actually have potential failure risks but currently do not show relevant abnormalities, the existing technology cannot provide timely early warning. Summary of the invention

[0003] The implementation methods of this specification are intended to solve at least one of the technical problems in the related art to a certain extent. To this end, the implementation methods of this specification propose an early warning method, device, network equipment, electric energy equipment and medium.

[0004] The present specification provides a device early warning method, wherein the device includes a subcomponent, and the method includes:

[0005] Obtaining historical production data characteristics of the subassembly during the production process;

[0006] If the historical production data characteristics of the subcomponent match the warning indicator, a warning prompt message is issued;

[0007] Among them, the early warning indicator is created by obtaining production failure correlation data based on sample production data and sample operation failure data of preset sub-component samples; if the production failure correlation data is greater than a predetermined threshold, the early warning indicator is determined according to the sample production data corresponding to the sample operation failure data; wherein the number of the preset sub-component samples is multiple.

[0008] In one of the embodiments, if the historical production data characteristics of the sub-component match the warning index, a warning prompt message is issued to the sub-component, thereby effectively discovering the potential problems of the sub-component and issuing a warning thereto.

[0009] In one embodiment, the device is an electric energy device, the subassembly is a battery, and the preset subassembly sample is a preset sample battery; before the correlation analysis is performed between the sample production data based on the preset subassembly sample and the sample operation fault data, the method further includes:

[0010] Acquire sample actual production data of the preset sample battery during the production process; wherein the sample actual production data corresponds to sample expected production data;

[0011] According to the comparison result between the sample actual production data and the sample expected production data, abnormal production data is determined in the sample actual production data as the sample production data.

[0012] In one embodiment, the method of determining the abnormal production data includes at least one of the following steps:

[0013] determining the sample actual production data different from the sample expected production data as the abnormal production data;

[0014] The sample actual production data whose difference data with the sample expected production data exceeds a preset difference threshold range is determined as the abnormal production data.

[0015] In one embodiment, the correlation analysis based on the sample production data and the sample operation failure data includes:

[0016] Extracting features of the abnormal production data to obtain target production data features;

[0017] A correlation analysis is performed on the target production data features and the sample operation fault data to obtain the production fault correlation data.

[0018] In one embodiment, after extracting features from the abnormal production data, the method further includes:

[0019] Extracting features from the abnormal production data to obtain features of the abnormal production data;

[0020] The abnormal production data features that meet the first preset screening condition are screened out as the target production data features.

[0021] In one embodiment, before performing the correlation analysis based on the sample production data and the sample operation fault data, the method further includes:

[0022] The sample operation fault data features are obtained from the sample fault tracing database as the sample operation fault data.

[0023] In one embodiment, the method for constructing the sample fault tracing database includes:

[0024] Acquiring operation data characteristics of a faulty battery during operation; the faulty battery is the preset sample battery that fails during operation;

[0025] Filtering out the operation data features that meet the second preset screening condition as the sample operation fault data features;

[0026] The sample operation fault data characteristics are stored in the sample fault tracing database.

[0027] In one embodiment, the step of obtaining the operating data characteristics of the faulty battery during operation includes:

[0028] Acquiring initial operation data of the faulty battery during operation;

[0029] Performing data cleaning on the initial operation data to obtain target operation data;

[0030] Feature extraction is performed on the target operation data to obtain the operation data feature.

[0031] In one embodiment, after obtaining the target production data characteristics, the method includes:

[0032] Classified abnormal data features obtained by classifying the target production data features;

[0033] After obtaining the sample operation fault data characteristics, the method includes:

[0034] Classified fault data features obtained by classifying the sample operation fault data features;

[0035] A correlation analysis is performed on the classified abnormal data features and the classified fault data features to obtain the production fault correlation data.

[0036] In one embodiment, the analysis method of performing correlation analysis to obtain the production failure correlation data includes at least one of clustering, variance analysis, and proportion statistics.

[0037] The embodiment of this specification provides an electric energy device early warning method applied to a network device, wherein the electric energy device includes a battery, and the method includes:

[0038] Obtaining historical production data characteristics of the battery during the production process;

[0039] Matching the historical production data characteristics with the early warning index, if the match is successful, sending a corresponding early warning signal to the electric energy equipment, so that the electric energy equipment sends out early warning prompt information according to the early warning signal; the method for creating the early warning index includes:

[0040] Performing correlation analysis on sample production data and sample operation failure data of preset sample batteries to obtain production failure correlation data; the number of the preset sample batteries is multiple;

[0041] If the production fault correlation data is greater than a predetermined threshold, a warning indicator is determined according to the sample production data corresponding to the sample operation fault data.

[0042] The embodiments of this specification provide an early warning method applied to an electric energy device, wherein the electric energy device includes a battery, and the method includes:

[0043] Receive a warning signal sent by a network device, and send corresponding warning prompt information according to the warning signal; the warning signal is determined by the network device according to the matching situation between the historical production data characteristics of the battery and the warning indicator, and the creation method of the warning indicator includes:

[0044] Performing correlation analysis on sample production data and sample operation failure data of preset sample batteries to obtain production failure correlation data; the number of the preset sample batteries is multiple;

[0045] If the production fault correlation data is greater than a predetermined threshold, a warning indicator is determined according to the sample production data corresponding to the sample operation fault data.

[0046] The embodiment of this specification provides an early warning method applied to an electric device, wherein the electric device includes a battery, and the electric device stores historical production data characteristics of the battery. The method includes:

[0047] Receive early warning indicators from network devices;

[0048] Matching the historical production data characteristics with the early warning indicators, and if the match is successful, issuing corresponding early warning prompt information;

[0049] The method for creating the early warning indicator includes:

[0050] Performing correlation analysis on sample production data and sample operation failure data of preset sample batteries to obtain production failure correlation data; the number of the preset sample batteries is multiple;

[0051] If the production fault correlation data is greater than a predetermined threshold, a warning indicator is determined according to the sample production data corresponding to the sample operation fault data.

[0052] In an implementation manner of the present specification, a network device is provided, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method described in any of the above implementation manners is implemented.

[0053] An electric energy device is provided in an implementation manner of the present specification, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the early warning method described in any of the above implementation manners is implemented.

[0054] In one embodiment, the electric energy device is a vehicle, a ship or an aircraft.

[0055] In an embodiment of the present specification, a computer-readable storage medium is provided. When instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device can execute the method described in any of the above embodiments.

[0056] In an embodiment of the present specification, a computer program product is provided, including a computer program. When the computer program is executed by a processor, the method described in any of the above embodiments is implemented.

[0057] In an embodiment of the present specification, a battery warning device is provided, the device comprising:

[0058] A production feature acquisition module is used to obtain actual production data features of the battery to be evaluated during the production process;

[0059] The early warning prompt issuing module is used to issue an early warning prompt message when the actual production data characteristics of the battery to be evaluated match the early warning index; wherein the method for creating the early warning index includes:

[0060] Performing correlation analysis on sample production data and sample operation failure data of preset sample batteries to obtain production failure correlation data; the number of the preset sample batteries is multiple;

[0061] If the production fault correlation data is greater than a predetermined threshold, a warning indicator is determined according to the sample production data corresponding to the sample operation fault data.

[0062] In an embodiment of the present specification, a vehicle warning device is provided, wherein the vehicle includes a power battery, and the device includes:

[0063] A production feature acquisition module, used to acquire historical production data features during the production process of the power battery;

[0064] The warning prompt issuing module is used to match the historical production data characteristics with the production warning indicators. If the match is successful, a corresponding warning signal is issued to the vehicle, so that the vehicle issues a warning prompt information according to the warning signal; the creation method of the warning indicator includes:

[0065] Performing correlation analysis on sample production data and sample operation failure data of preset sample batteries to obtain production failure correlation data; the number of the preset sample batteries is multiple;

[0066] If the production fault correlation data is greater than a predetermined threshold, a warning indicator is determined according to the sample production data corresponding to the sample operation fault data.

[0067] In the implementation manner of the above specification, production failure correlation data is obtained by performing correlation analysis based on sample production data and sample operation failure data of preset sub-component samples; if the production failure correlation data is greater than a predetermined threshold, a warning indicator is created according to the sample production data corresponding to the sample operation failure data. Then, the historical production data characteristics of the sub-component in the production process are obtained, and if the historical production data characteristics of the sub-component match the warning indicators, a warning prompt message is issued. In the event that some equipment sub-components fail, the equipment sub-components that do not currently show relevant abnormalities but have potential failure risks can be determined by matching the warning indicators with the historical production data characteristics of the equipment sub-components, and a warning prompt message is issued. Furthermore, the created warning indicators can also be used to guide the production process of the equipment sub-components to improve the quality of the equipment sub-components. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] Figure 1a A schematic diagram of an application scenario of a method for creating a battery production early warning indicator provided in an embodiment of this specification;

[0069] Figure 1b A schematic diagram of the process flow of the device early warning method provided in the implementation mode of this specification;

[0070] Figure 2 A schematic diagram of a process for determining sample production data provided for an implementation of this specification;

[0071] Figure 3 A schematic diagram of a process for determining production fault correlation data provided in an embodiment of this specification;

[0072] Figure 4a A schematic diagram of a process for determining target production data characteristics provided in an embodiment of this specification;

[0073] Figure 4b A schematic diagram of battery voltage and current provided for an embodiment of this specification;

[0074] Figure 4c A schematic diagram of battery temperature and current provided for an embodiment of this specification;

[0075] Figure 4d A schematic diagram of battery capacity and current provided for an embodiment of this specification;

[0076] Figure 5a A schematic diagram of a process for determining operating data characteristics provided in an embodiment of this specification;

[0077] Figure 5b A schematic diagram of battery cell voltage provided for an embodiment of this specification;

[0078] Figure 5c A schematic diagram of battery cell voltage provided for an embodiment of this specification;

[0079] Figure 6 A schematic diagram of a process for determining production fault correlation data provided in an embodiment of this specification;

[0080] Figure 7 A pie chart showing the statistical proportion of fault type data features and abnormal battery production data features provided in the implementation scheme of this specification;

[0081] Figure 8 A flow chart of a method for determining production early warning indicators provided in the implementation mode of this specification;

[0082] Fig. 9 A schematic diagram of the process flow of the vehicle warning method provided in the embodiment of this specification;

[0083] Fig.10 A schematic diagram of the process flow of the vehicle warning method provided in the embodiment of this specification;

[0084] Fig.11 A flowchart of a method for creating a battery production early warning indicator provided in an embodiment of this specification;

[0085] Fig.12 A schematic diagram of a device for creating a battery production early warning indicator provided in an embodiment of this specification;

[0086] Fig.13 A schematic diagram of a battery warning device provided in an embodiment of this specification;

[0087] Fig.14 A schematic diagram of a vehicle warning device provided in an embodiment of this specification;

[0088] Fig.15 A schematic diagram of a vehicle warning device provided in an embodiment of this specification;

[0089] Fig.16 A schematic diagram of a battery warning device provided in an embodiment of this specification;

[0090] Fig.17 A schematic diagram of a vehicle warning device provided in an embodiment of this specification. DETAILED DESCRIPTION

[0091] The embodiments of the present invention are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary, such as taking the manufacturing process of the power battery industry as an example for analysis, and taking the production early warning indicator as an example to exemplify the indicator creation process, which is intended to be used to explain the present invention and cannot be understood as limiting the present invention.

[0092] In the manufacturing industry, when a product is produced, various tests are often performed to distinguish qualified products from unqualified products. However, even when a qualified product is shipped out of the factory after the production test, there may still be some potential failure risks. For these products with potential risks, it is impossible to predict them in advance in the prior art. For ease of understanding, the following text takes a battery as an example to illustrate the implementation of the present invention.

[0093] Under the urgent need of energy conservation and emission reduction in the whole society, the development of environmentally friendly or energy-saving new energy vehicles has become a consensus in the automotive industry. As the main direction of the development of new energy vehicles, the safety and stability of electric vehicles have always been a hot topic of concern to the whole society. Frequent power battery safety accidents have caused losses in people's lives and property. Therefore, it is necessary to provide strong guarantees for the safety and stability of power batteries through comprehensive and strict process optimization methods and accurate and effective fault warning methods.

[0094] In the related art, the battery production process is monitored and certain abnormal production situations are warned to prevent power battery failures in advance. However, the monitoring method used in the battery production process in the related art does not detect abnormal production situations related to battery operation failures.

[0095] The operating conditions of the electrochemical process involved in the battery production process are smooth and single, the battery is in a quasi-steady state, and the various parameters of the battery are determined at this stage. Therefore, the initial quality of the battery depends on the battery production process, and the initial quality of the battery is related to the operational failures generated during the operation after loading. After further analysis by the inventor, since the initial quality of the battery depends on the battery production process, and the battery production process generates actual production data, then, based on the actual production data, the abnormal characteristics shown by the actual production data are indeed the root cause of the operational failures generated during the operation after loading.

[0096] Based on this, the implementation method of this specification provides a method for creating early warning indicators. Specifically, first, obtain sample production data and sample operation failure data of the sample battery, perform correlation analysis based on the obtained sample production data and sample operation failure data, and obtain production failure correlation data. It should be noted that the number of sample batteries is multiple, and the sample operation failure data can be of multiple types, production failure correlation data. Secondly, compare the production failure correlation data with a predetermined threshold value. If the production failure correlation data is greater than the predetermined threshold value, the sample operation failure data corresponding to the production failure correlation data can be determined, and then the early warning indicators are determined based on the sample production data corresponding to the sample operation failure data. It should be noted that the sample operation failure data in the implementation method of this application refers to the data obtained from the failures caused by the products that are shipped as qualified products during the normal operation process in the later stage.

[0097] The implementation method of this specification also provides a scenario example to exemplarily illustrate the method of creating early warning indicators. First, obtain the target production data features used to characterize the deviation of the actual production data of the battery from the expected production data of the battery during the production process of the battery, and the operation fault data features related to the fault type during the operation of the faulty battery; then, use the target production data features and the operation fault data features to perform correlation analysis to obtain production fault correlation data that describes the degree of correlation between the target production data features and the fault type; finally, mine the early warning indicators related to the fault type in the target production data features based on the production fault correlation data. By determining the early warning indicators that are more relevant to the fault type, when the production data features corresponding to the early warning indicators appear in the battery production process, early warning prompts can be issued accurately and timely for abnormal production situations related to battery operation failures; or, when the battery has been tested and qualified but has not yet been loaded and used, it is still in the battery warehouse storage stage or transportation stage, etc., as long as the production data features corresponding to the early warning indicators are found in the production process of a certain battery, an early warning prompt information can be issued to the battery. It should be understood that for batteries with early warning functions, the battery itself can issue a prompt message; and for batteries without early warning functions, the early warning message can be directly issued by a server or other computing terminal, and the early warning message contains identity information, location information or batch information that can point to the battery.

[0098] Furthermore, on the one hand, by giving reminders for abnormal production conditions of batteries being produced, the outflow of unqualified batteries can be reduced and the safety of batteries can be improved; on the other hand, for batteries that have already left the factory and are in use, early warning indicators can be used to analyze the actual production data of the batteries in use, and early warnings of battery failures can be given as early as possible. It should be noted that the faulty battery in the implementation of this application refers to a battery that has passed the quality inspection when the production is completed, but fails in the subsequent use process, and does not refer to an unqualified battery that fails in the quality inspection when the production is completed.

[0099] This scenario example also provides a device early warning method combined with the early warning indicators related to the operation fault type mentioned above. The device can be an electric energy device. The electric energy device can include a vehicle, a ship or an aircraft. This scenario example takes a vehicle as an example.

[0100] In the related technology, the historical operation data generated by the battery during the operation on the vehicle side can be obtained, the data features related to the operation failure can be mined, and the operation failure warning model can be trained by combining unsupervised learning and supervised learning methods to complete the early warning of the battery operation failure. However, since the battery is a complex electrochemical system with a high degree of nonlinearity, and is affected by road conditions and driving environment, the vehicle operation conditions are complex and changeable, which leads to large noise in the historical operation data of the vehicle-side battery, which is not conducive to the mining and extraction of operation data features related to the fault. Therefore, the method of using only the historical operation data of the battery for vehicle-side battery failure warning also needs to be improved.

[0101] For batteries that have already left the factory and are running on the vehicle, if the battery on the vehicle shows operating fault data characteristics related to a certain operating fault type during operation, it can be combined with the actual production data of the power battery on the vehicle to further determine whether to issue a warning for this certain operating fault type, which can improve the accuracy of the warning.

[0102] Based on this, as mentioned above, the early warning indicators related to the operation fault type have been constructed, and the corresponding relationship between the operation fault type and the early warning indicator can be further generated. By extracting the characteristics of the operation data of the battery on the vehicle during operation, the operation data characteristics of the battery on the vehicle during operation can be obtained. If the operation data characteristics indicate that the battery has operation fault data characteristics related to the fault type during operation. The actual production data of the battery in the production process is obtained, and the characteristics of the actual production data are extracted to obtain the historical production data characteristics. Based on the corresponding relationship between the operation fault type and the early warning indicator, the target early warning indicator corresponding to the certain operation fault type is determined to further verify whether to issue an early warning for the certain operation fault type. The historical production data characteristics are compared with the target early warning indicators. If the historical production data characteristics match the target early warning indicators, it indicates that the battery on the vehicle is more likely to have an operation fault, and an early warning prompt message is issued. If the historical production data characteristics do not match the target early warning indicators, it indicates that the production process of the battery to be evaluated is normal, and no early warning is currently required. It can be seen that by combining the historical production data characteristics and battery operation data characteristics of the entire life cycle of the battery for vehicle early warning, the accuracy of vehicle fault early warning can be improved.

[0103] See also Figure 1a The vehicle warning method provided in this scenario example can be applied to Figure 1a In the vehicle 110, or applied to other devices having the function of controlling the vehicle (such as Figure 1aThe cloud server 120, mobile terminal 130, etc. in the system. Among them, the vehicle can be an autonomous driving vehicle, and the autonomous driving vehicle can be a vehicle with partial autonomous driving functions or a vehicle with full autonomous driving functions. In other words, the level of autonomous driving of the vehicle can be divided into no automation (L0), driving support (L1), partial automation (L2), conditional automation (L3), high automation (L4) or full automation (L5) with reference to the classification standard of the Society of Automotive Engineers (SAE). A vehicle or other device can implement the method for creating the early warning indicator through the components (including hardware and software) it contains. It is understandable that the vehicle can be a car, truck, motorcycle, bus, recreational vehicle, amusement park vehicle, construction equipment (such as engineering vehicle), tram, golf cart, train, and cart, etc., which use power batteries. The implementation method of this specification does not make any special limitation on this.

[0104] Exemplarily, the vehicle-side Tbox (Telematics box, Internet of Vehicles system) sends the initial operation data of battery A during operation to the cloud server 120, and the cloud server 120 can analyze and extract features of the received initial operation data to obtain operation fault data features related to fault type X. The cloud server 120 stores historical production data features F that describe the situation where the actual production data of the battery deviates from the expected production data of the battery, and can also store the correspondence between the operation fault type and the production warning indicator. The cloud server 120 matches the corresponding relationship between the operation fault type and the production warning indicator based on the fault type X to obtain the production warning indicator Q related to the fault type X. If the historical production data feature F matches the production warning indicator Q, the cloud server 120 sends a warning prompt message to the vehicle-side.

[0105] The cloud server 120 may be an electronic device with certain computing and processing capabilities. It may have a network communication module, a processor, and a memory. The server may be a distributed server, which may be a system with multiple processors, memories, network communication modules, etc. operating in coordination. Alternatively, the cloud server may also be a server cluster formed by several servers. Alternatively, with the development of science and technology, the server may also be a new technical means capable of realizing the corresponding functions of the implementation method of the specification. For example, it may be a new form of "server" based on quantum computing.

[0106] This specification provides a device early warning method, the device includes subcomponents, see Figure 1b , the device early warning method may include the following steps:

[0107] S110. Obtain historical production data characteristics of the sub-component during the production process.

[0108] S120: If the historical production data characteristics of the sub-component match the warning indicator, a warning prompt message is issued.

[0109] Among them, the early warning indicator is created by performing correlation analysis on the sample production data and sample operation failure data based on the preset sub-component samples to obtain the production failure correlation data; the number of preset sub-component samples is multiple; if the production failure correlation data is greater than the predetermined threshold, the early warning indicator is created in a manner of determining the sample production data corresponding to the sample operation failure data. Subcomponents can be different modules, parts or functional modules in the device, which together constitute the function of the entire device or system. For example, the main chassis of a computer can be regarded as a device, and the main chassis includes multiple subcomponents, such as the motherboard, processor, memory, hard disk, graphics card, etc. These subcomponents are responsible for different functions, and work together to enable the normal operation of the entire computer system. The historical production data feature can characterize the situation where the historical production data of the subcomponent in the production process step deviates from the expected production data.

[0110] Among them, the early warning indicator can be a production indicator used to monitor whether abnormal production conditions related to subcomponent operation failures occur during the production process of the subcomponent. The early warning indicator can directly use the target production data that meets the indicator screening conditions, or the early warning indicator can use the target production data features extracted based on the target production data and meet the indicator screening conditions. The early warning indicator can serve as an important basis for process optimization, quality control, and fault warning of existing equipment.

[0111] The early warning information can be used to remind or notify the subassembly to be evaluated of possible abnormal production conditions during the production process. The early warning information can include the unique identification code of the subassembly, the actual production data characteristics of the subassembly, and the failure type of the subassembly. The early warning information is in the form of voice, and can also be sent to an electronic device with a display function, such as at least one of a mobile terminal, a vehicle-side display unit, and a vehicle-mounted tablet computer.

[0112] Specifically, during the production process of the subcomponent, the historical production data of the subcomponent during the production process is obtained. By extracting the features of the historical production data of the subcomponent during the production process, the features of the historical production data of the subcomponent during the production process can be obtained, and then the features of the historical production data of the subcomponent are matched with the early warning indicators. If the two match, it indicates that the production process of the subcomponent is abnormal, and the device sends an early warning prompt message. If the two do not match, it indicates that the production process of the subcomponent is normal and no early warning is required at present. In some embodiments, the historical production data can be stored locally in the device in advance, and the historical production data can be obtained from the local device; the historical production data can also be stored in advance to other computer devices (such as cloud servers), and the historical production data can be obtained from the computer device. In some other embodiments, the historical production data features of the subcomponent during the production process can be stored locally in advance, and the historical production data features of the subcomponent during the production process can be obtained from the local device; the historical production data features of the subcomponent during the production process can also be stored in advance to other computer devices (such as cloud servers), and the historical production data features of the subcomponent during the production process can be obtained from the computer device.

[0113] In the above-mentioned battery early warning method, the historical production data characteristics of the sub-component in the production process are obtained; if the historical production data characteristics of the sub-component match the early warning indicators, the equipment sub-component that actually has a potential failure risk but currently does not show relevant abnormalities is determined, and then the equipment issues an early warning prompt message. For sub-components that have already left the factory and are in use, early warning indicators can be used to analyze the sub-components in use, and accurate early warnings can be issued for sub-component operating failures as early as possible. Furthermore, by giving reminders for production anomalies of sub-components being produced, the outflow of unqualified sub-components can be reduced and the safety of sub-components can be improved; early warning indicators can also be used to optimize the process steps in the sub-component production process to reduce the probability of sub-component failures.

[0114] It should be noted that the early warning indicator is created by performing correlation analysis based on the sample production data and sample operation failure data of the preset subcomponent sample to obtain the production failure correlation data; the number of preset subcomponent samples is multiple; if the production failure correlation data is greater than a predetermined threshold, the early warning indicator is determined according to the sample production data corresponding to the sample operation failure data. Exemplarily, the device is an electric energy device, the subcomponent is a battery, and the preset subcomponent sample is a preset sample battery.

[0115] Among them, the number of preset sub-component samples can be multiple. Correlation analysis can analyze two or more variable elements with correlation, so as to measure the degree of correlation between two variable factors. Through correlation analysis, certain connections between different variable elements are mined. Sample production data can be the original data generated in the production process of any preset sub-component sample, such as the original data generated in the production process of preset sample batteries and the original data generated in the production process of chips. Sample production data can also be production data features obtained by analyzing or feature extracting the original data generated in the production process of preset sub-component samples. Sample operation fault data may include fault type and data features. Sample operation fault data can be operation data generated in the operation process of faulty sub-component samples, such as operation data of faulty batteries and operation data of faulty chips. Operation fault data can also be sample operation fault data features obtained by analyzing or feature extracting operation fault data. Production fault correlation data is used to characterize the degree of correlation between production data and fault type. Exemplarily, production fault correlation data can be divided into complete correlation, no correlation, incomplete correlation, or a numerical value that can represent the size of correlation, such as a correlation of 20%, 50% or 70%. The fault type can be understood as the category of the operating fault that occurs during the operation of the faulty subcomponent sample, or it can be the overall fault of the device caused by the faulty subcomponent sample. Taking the preset sample battery as an example, the fault type can be at least one of excessive self-discharge and internal short circuit. The fault type can also be at least one of difficulty in starting the vehicle engine and vehicle electrical system failure.

[0116] In some cases, by analyzing the correlation between sample production data and sample operation failure data, the obtained production failure correlation data can better reflect the actual correlation between the preset sub-component sample operation failure and the production anomaly shown in the preset sub-component sample production process, and can realize the correlation between the performance of the preset sub-component samples throughout the life cycle.

[0117] Specifically, before making the correlation, it is necessary to obtain sample production data and sample operation failure data. The sample production data and sample operation failure data may have been stored locally in advance, and the sample production data and sample operation failure data may be obtained locally; the sample production data and sample operation failure data may also be pre-stored in other computer devices (such as cloud servers), and the sample production data and sample operation failure data may be obtained from the computer device; the sample production data may also be pre-stored locally, and the operation failure data may be collected, and the sample production data may be obtained locally and the collected sample operation failure data may be received.

[0118] It should be noted that the sample production data in the implementation manner of this specification may be obtained by collecting data generated during the production process of the preset sub-component sample using data collection settings such as sensors deployed on the preset sub-component sample production line. After obtaining the sample production data and the sample operation failure data, a correlation analysis algorithm is used to mine the correlation between the obtained sample production data and the sample operation failure data to obtain production failure correlation data. In some other implementations, after obtaining the sample production data and the sample operation failure data, feature extraction may be performed on the sample production data to obtain target production data features. Feature extraction is performed on the sample operation failure data to obtain sample operation failure data features. A correlation analysis algorithm is used to mine the correlation between the target production data features and the sample operation failure data features to obtain production failure correlation data.

[0119] After obtaining the production fault correlation data based on the obtained sample production data and sample operation fault data, the production fault correlation data is compared with the predetermined threshold value. If the production fault correlation data is greater than the predetermined threshold value, it indicates that the production data corresponding to the production fault correlation data is highly likely to be related to the preset sub-component sample operation fault. Therefore, the early warning indicator can be created based on the sample production data corresponding to the production fault correlation data. For example, the sample production data corresponding to the production fault correlation data is analyzed to obtain the target production data characteristics related to the fault type, and used as the indicator to be created.

[0120] In some other embodiments, several target production data features corresponding to the production data may be stored in advance. When the production fault correlation data is greater than a predetermined threshold, the target production data features related to the fault type may be directly used as the indicator to be created. In some embodiments, as the number of preset sub-component samples produced increases and the number of faulty sub-component samples increases, the early warning indicators may be updated accordingly, such as modifying existing early warning indicators or adding new early warning indicators. Specifically, as the number of preset sub-component samples produced increases, the target production data features stored in the abnormal feature database will be updated, and as the number of faulty sub-component samples increases, the sample operation fault data features related to the fault type stored in the sample fault tracing database will be updated. The updated target production data features and the updated sample operation fault data features are used to perform correlation analysis to obtain new production fault correlation data. According to the new production fault correlation data, new early warning indicators related to the fault type are determined in the updated target production data features.

[0121] In the above index creation method, the production fault correlation data is obtained by performing correlation analysis based on the obtained sample production data and sample operation failure data; if the production fault correlation data is greater than a predetermined threshold, the early warning index is created according to the sample production data corresponding to the sample operation failure data. In the case of failure of some products, the production data of the product can be analyzed by the created indicators to accurately determine the products with potential failures; further, the created indicators can also be used to guide the production process of the product to improve the quality and quality of subsequent products.

[0122] In some embodiments, the device is an electric energy device, the subassembly is a battery, and the predetermined subassembly sample is a predetermined sample battery. Figure 2 , before performing correlation analysis based on sample production data and sample operation failure data of preset sub-component samples, it also includes:

[0123] S210: Acquire actual production data of a sample of a preset sample battery during the production process.

[0124] S220. According to the comparison result between the sample actual production data and the sample expected production data, abnormal production data is determined in the sample actual production data as the sample production data.

[0125] The actual production data of the sample may be data generated by the preset sample battery in the production process steps, and the actual production data of the sample includes but is not limited to the preset sample battery front weight, preset sample battery rear weight, injection volume, current I during the formation process, and the like. HC , voltage V HC , Capacity VAP HC , Temperature T HC , the current I during charging CD , voltage V CD , Capacity CAP CD , Temperature T CD , the current I during the capacity division process FR , voltage V FR , Capacity CAP FR , Temperature T FR The sample actual production data corresponds to the sample expected production data. The sample expected production data can be the expected production data considered when designing the preset sample battery, or it can be the production data that is more in line with the actual production needs after the preset sample battery is put into production based on the sample production data of a large number of preset sample batteries.

[0126] In some cases, the preset sample battery production process steps cannot maintain the sample actual production data of each preset sample battery at a constant ideal value, and the preset sample battery production process steps may cause abnormalities in the sample actual production data of the preset sample battery due to the preset sample battery itself or the production environment.

[0127] Specifically, according to the data generated by the process steps in the production process of the preset sample battery, the sample actual production data of the preset sample battery in the production process can be obtained. The sample actual production data of the preset sample battery is compared with the sample expected production data of the preset sample battery to obtain the comparison result between the sample actual production data and the sample expected production data, so that according to the comparison result between the sample actual production data and the sample expected production data, it can be determined that there is abnormal production data in the sample actual production data of the preset sample battery, that is, abnormal production data, as the sample production data.

[0128] Exemplarily, the sample expected production data of different models of preset sample batteries are different. Therefore, a relational database of each model of preset sample battery and the sample expected production data of each model of preset sample battery can be established. After obtaining the sample actual production data of the preset sample battery during the production process, the model of the preset sample battery can be searched in the relational database, and the sample expected production data of the preset sample battery found is compared with the sample actual production data of the preset sample battery. According to the comparison result, the abnormal production data in the sample actual production data can be determined as the sample production data.

[0129] In this embodiment, the sample production data obtained may be abnormal production data, and then a correlation analysis may be performed based on the abnormal production data and the sample operation fault data to obtain production fault correlation data. The production fault correlation data may be used to characterize the degree of correlation between the abnormal production data and the fault type.

[0130] In the above implementation, the sample actual production data of the preset sample battery in the production process is obtained, and then according to the comparison result between the sample actual production data and the sample expected production data, the abnormal production data is determined in the sample actual production data, and finally the abnormal production data is used as the sample production data. It can be seen that by comparing the sample actual production data with the sample expected production data, the abnormal production data can be determined as the sample production data, and the data range of the relevant analysis can be narrowed, that is, by determining the abnormal production data, the data mining range of the target production data feature can be reduced.

[0131] In some embodiments, the method of determining abnormal production data includes at least one of the following steps:

[0132] Determine the actual production data of the sample that is different from the expected production data of the sample as abnormal production data;

[0133] The actual production data of the sample whose difference with the expected production data of the sample exceeds the preset difference threshold range is determined as abnormal production data.

[0134] Among them, the preset difference threshold range can be set in combination with actual conditions, and different preset difference threshold ranges can be set according to different models of preset sample batteries. Since different process steps in the production process of the preset sample batteries will cause different degrees of errors, the preset difference threshold range can be set according to the error data allowed to exist in different process production steps in the production process of the preset sample batteries. For example: the battery capacity in the expected production data of the preset sample battery is generally set to 100%. In the process steps of the preset sample battery production process, the preset sample batteries with a capacity of 99.8% to 100% can be regarded as normal batches of preset sample batteries. Therefore, the preset difference threshold range of the preset sample battery capacity can be set to 0% to 0.2%.

[0135] Specifically, in some embodiments, the sample actual production data may be compared with the sample expected production data, and when there is a difference between the two, the sample actual production data with the difference may be determined as abnormal production data. It is understandable that when the sample actual production data is the same as the sample expected production data, the corresponding sample actual production data is regarded as normal production data.

[0136] In some other embodiments, the sample actual production data may be compared with the sample expected production data, and when there is a difference between the two, the difference is further compared with a preset difference threshold range. Specifically, by comparing the sample actual production data with the sample expected production data, the difference data between the sample actual production data and the sample expected production data is determined, and the difference data is compared with a preset difference threshold range. If the difference data is within the preset difference threshold range, the corresponding sample actual production data may be considered normal production data, and if the difference data exceeds the preset difference threshold range, the corresponding sample actual production data may be determined as abnormal production data.

[0137] In the above implementation, the sample actual production data that is different from the sample expected production data is determined as abnormal production data, or the sample actual production data whose difference data from the sample expected production data exceeds the preset difference threshold range is determined as abnormal production data. Not only is the abnormal production data determined in a variety of ways, but also by setting the preset difference threshold range, the flexibility of comparing the sample actual production data and the sample expected production data is improved, and the abnormal production data can be determined more accurately.

[0138] In some embodiments, see Figure 3 , based on the correlation analysis of sample production data and sample operation failure data, including:

[0139] S310: Extract features of abnormal production data to obtain target production data features.

[0140] Among them, the situation where the abnormal production data deviates from the expected production data can be obtained by feature extraction, and the deviation is used as the target production data feature. The target production data feature is used to characterize the situation where the actual production data of the battery deviates from the expected production data of the battery. The target production data feature can be understood as the abnormal production data feature that occurs during the battery production process. For example: the voltage of the battery is lower than the expected voltage data, and the target production data feature can be the voltage deviation value between the voltage of the battery and the expected voltage data; the current of the battery is lower than the expected current data, and the target production data feature can be the current deviation value between the current of the battery and the expected current data. For another example, during the production and charging process of the battery, the voltage should show an upward trend under normal circumstances. If the actual production charging voltage of the battery shows a downward trend during the production and charging process, it indicates that there is a production abnormality. Therefore, the target production data feature can be that when there is a charging current in the production and charging process, the battery terminal voltage decreases.

[0141] In some cases, the operating conditions of the electrochemical process treatment involved in the production process of the preset sample battery are smooth and single, the preset sample battery is in a quasi-steady state, and various parameters of the preset sample battery are determined at this stage. Therefore, the production process data of the preset sample battery can better reflect the initial quality of the preset sample battery. The abnormal characteristics of the preset sample battery in the production data may be the root cause of the failure of the preset sample battery during operation. Therefore, it is necessary to determine the abnormal target production data characteristics existing in the production process of the preset sample battery.

[0142] In order to mine the correlation between battery operation failure and battery production data, it is necessary to first obtain the target production data features in the battery production process and the operation failure data features related to the failure type in the operation process of the faulty battery. Specifically, by extracting features from the abnormal production data, it is possible to obtain the situation where the abnormal production data deviates from the expected production data of the battery, and the target production data features can be obtained by the situation where the abnormal production data deviates from the expected production data of the battery.

[0143] Exemplarily, the actual production data of the battery during the production process can be compared with the expected design data to obtain the actual production data with anomalies, and feature extraction can be performed based on the actual production data with anomalies to obtain the target production data features of the battery during the production process.

[0144] S320: Perform correlation analysis on target production data features and sample operation failure data to obtain production failure correlation data.

[0145] Among them, correlation analysis can analyze two or more correlated variable elements to measure the degree of correlation between the two variable factors. Through correlation analysis, certain connections between different variable elements are mined. Production failure correlation data is used to characterize the degree of correlation between the target production data characteristics and the fault type. Exemplarily, production failure correlation data can be divided into complete correlation, no correlation, and incomplete correlation.

[0146] In some cases, by analyzing the correlation between the target production data characteristics and the operation failure data characteristics, the obtained production failure correlation data can better reflect the true correlation between the battery operation failure and the production anomaly manifested in the battery production process, and can realize the correlation of the battery's performance throughout its life cycle.

[0147] Specifically, by using a correlation analysis algorithm to mine the correlation between the target production data features and the sample operation fault data, production fault correlation data between the target production data features and the fault type can be obtained to quantify the degree of correlation between the target production data features and the fault type.

[0148] In the above implementation, by determining the target production data features and performing correlation analysis using the target production data features and sample operation fault data, the data mining scope of the target production data features related to the fault type can be reduced and the data processing speed can be improved.

[0149] In some embodiments, see Figure 4a , after feature extraction of abnormal production data, it also includes:

[0150] S410: Extract features of abnormal production data to obtain features of the abnormal production data.

[0151] Specifically, based on the data generated by the process steps in the production process of the preset sample battery, the abnormal production data of the preset sample battery in the production process can be obtained. By extracting features from the abnormal production data, the production situation corresponding to the abnormal production data can be obtained, and the abnormal production data features can be obtained through the production situation corresponding to the abnormal production data. The feature extraction methods include but are not limited to derivation, threshold range, and other methods.

[0152] For example, during the charging process of the preset sample battery production, the voltage of the preset sample battery should always show a voltage rising trend under normal charging conditions. Figure 4b ,right Figure 4bBy taking the derivative of the voltage curve of the charging part, we can get the position of the zero point in the curve. If there is a zero point, it can be considered that during the charging process of the preset sample battery, the voltage first rises and then falls, or the voltage first falls and then rises, and the preset sample battery is considered to be abnormal. Figure 4b There are zero points in both 410 and 420. The abnormal production data feature of the preset sample battery can be expressed as: when there is a charging current in the production charging process, the terminal voltage of the preset sample battery drops. Then, data feature 410 and data feature 420 belong to the classification abnormal data feature M. It should be noted that when the current is negative, it means that the preset sample battery is charging. Figure 4b 430 in the figure has stopped charging the preset sample battery, and the voltage will drop because the voltage of the preset sample battery will drop after the charging power supply is disconnected. The preset sample battery voltage that drops after stopping charging is reasonable and normal within a certain range.

[0153] For example, the voltage difference is reasonable within a certain range, so a preset threshold range of the voltage difference is set. Figure 4b , Figure 4b The voltage at 04:30 is Figure 4b The voltage difference at A is calculated. If the value exceeds the preset threshold range, it is considered that the preset sample battery is abnormal. The abnormal production data feature of the preset sample battery can be expressed as: the voltage difference exceeds the preset threshold range.

[0154] For example, during the charging process of the preset sample battery, the temperature of the preset sample battery may fluctuate within a certain range, so the preset sample battery temperature is set with a preset threshold range. Figure 4c , the preset sample battery is in the charging process, if Figure 4c If the temperature at position 440 exceeds the preset threshold range, it is considered that the preset sample battery is abnormal. The abnormal production data feature of the preset sample battery can be expressed as: when the production charging process has a charging current, the battery temperature is too high.

[0155] Exemplarily, the preset sample battery capacity is reasonable within a certain range, so a preset threshold range of the preset sample battery capacity is set. Figure 4d During the charging process of the preset sample battery, if the capacity of the preset sample battery reaches Figure 4d If the value does not rise after reaching the 450 position in the figure and the value is not within the preset threshold range, it is considered that the preset sample battery is abnormal. The abnormal production data feature of the preset sample battery can be expressed as: the capacity of the preset sample battery does not reach the calibrated battery capacity.

[0156] S420: Filter out abnormal production data features that meet a first preset filtering condition as target production data features.

[0157] Among them, the first preset screening condition may be the true correlation and credibility between the abnormal production data characteristics and the battery failure.

[0158] In some cases, not all abnormal production data features can be stored in the abnormal feature database. Therefore, the first preset screening condition is used as the storage condition for the abnormal feature database to ensure the true relevance and credibility of the production abnormal features stored in the abnormal feature database as much as possible. If the data are not stored according to the first preset screening condition, there will be risk events such as early warning errors.

[0159] Specifically, by analyzing the abnormal production data characteristics, it can be determined whether the abnormal production data characteristics meet the first preset screening condition, and the production data characteristics that meet the first preset screening condition are used as target production data characteristics, and the target production data characteristics are stored in the abnormal characteristic database.

[0160] In some other embodiments, the abnormal production data feature corresponds to a second cause, which is the fault cause that causes the abnormal production data feature. The second cause can be a cause that can cause the abnormal production data feature to appear and is manually derived based on relevant knowledge and experience.

[0161] In some cases, not all production abnormality data features can be stored in the abnormal feature database. Instead, the second cause of production is used as the entry condition of the abnormal feature database to ensure the true relevance and credibility of the production abnormality features stored in the abnormal feature database as much as possible. If the data is not entered into the database according to the second cause of production, there will be risk events such as early warning errors. Specifically, by analyzing the cause of the abnormal production data feature, the cause of the abnormal production data feature, that is, the second cause of production, can be determined. In practical applications, the correlation between the second cause of production and the abnormal production data feature is verified to determine whether the second cause of production is the true cause of the abnormal production data feature. If the second cause of production is the true cause, it can be considered that the second cause of production meets the first preset screening condition, and the abnormal production data feature can be stored in the abnormal feature database.

[0162] For example, please refer to Figure 4b, the abnormal production data feature of the preset sample battery is represented by a drop in the terminal voltage of the preset sample battery when the production charging process has a charging current. The cause analysis of the abnormal production data feature of the preset sample battery shows that the reason why the preset sample battery produces the abnormal production data feature is that during the production process of the preset sample battery, dust adheres to the pole piece or leaks liquid, causing a slight internal short circuit inside the preset sample battery and a decrease in internal resistance. Verify the true correlation and accuracy between the dust adheres to the pole piece or leaks liquid during the production process of the preset sample battery, causing a slight internal short circuit inside the battery and a decrease in internal resistance and the drop in the terminal voltage of the preset sample battery when the production charging process has a charging current. If there is a true and credible correlation between the two, it can be considered that the second cause meets the first preset screening condition, and the production data feature of the drop in the terminal voltage of the preset sample battery when the production charging process has a charging current can be used as the target production data feature that needs to be stored in the abnormal feature database.

[0163] Exemplarily, the abnormal production data feature of the preset sample battery is represented by the preset sample battery temperature being too high when the production charging process has a charging current. The cause analysis of the production data feature of the preset sample battery can be considered that the reason why the preset sample battery produces the abnormal production data feature is: there is a problem with the transmission characteristics of the preset sample battery itself, and the internal circuit is abnormal. The true correlation and accuracy between the transmission characteristics of the preset sample battery itself, the internal circuit abnormality and the preset sample battery temperature being too high when the production charging process has a charging current is verified. If there is a true and credible correlation between the two, it can be considered that the second cause meets the first preset screening condition, and the production data feature of the battery temperature being too high when the production charging process has a charging current can be used as the target production data feature that needs to be stored in the abnormal feature database.

[0164] In the above embodiment, by constructing an abnormal feature database, abnormal production data features that meet the first preset screening condition are stored in the abnormal feature database, which not only ensures the true correlation and credibility of the abnormal production data features stored in the abnormal feature database with the battery failure as much as possible, but also can precipitate the production abnormal data features of the preset sample battery, which can provide a data source for the optimization of the preset sample battery production process or process.

[0165] In some implementations, before performing a correlation analysis based on the sample production data and the sample operation failure data, the method may further include: acquiring sample operation failure data features from a sample failure tracing database as the sample operation failure data.

[0166] The sample operation fault data feature is used to characterize the abnormality in the actual operation data of the preset sample battery. For example, during the charging process, the preset sample battery has a voltage rise that is too fast compared to other preset sample batteries of the same model, and the terminal voltage of the preset sample battery reaches the cut-off voltage first. Therefore, the sample operation fault data feature can be that the charging voltage of the preset sample battery rises too fast and reaches the cut-off voltage before the normal preset sample battery.

[0167] In some cases, when the preset sample batteries are put into use, different battery failures may occur depending on the usage of the batteries in different vehicles. Therefore, it is necessary to use sample operation failure data characteristics to describe the abnormal data conditions existing in the operation process of the batteries in use.

[0168] Specifically, a sample fault tracing database is provided in advance, and a number of sample operation fault data features are stored in the sample fault tracing database. The sample operation fault data features can be obtained from the sample fault tracing database as sample operation fault data.

[0169] In the above implementation, by obtaining sample operation fault data features from the sample fault tracing database, correlation analysis can be performed between the sample operation fault data features and the target production data, thereby reducing the data mining scope of the target production data features related to the fault type and improving the data processing speed.

[0170] In some embodiments, a method for constructing a sample fault tracing database includes: obtaining operating data characteristics of a faulty battery during operation; screening out operating data characteristics that meet a second preset screening condition as sample operating fault data characteristics; and storing the sample operating fault data characteristics in the sample fault tracing database.

[0171] The faulty battery is a preset sample battery that has a fault during operation. The second preset screening condition may be the true correlation and credibility between the operating data feature and the battery fault.

[0172] In some cases, not all operating data features can be stored in the fault tracing database. Therefore, the second preset screening condition is used as the storage condition for the sample fault tracing database to ensure the true relevance and credibility of the operating data features stored in the sample fault tracing database as much as possible. If the data is not stored according to the second preset screening condition, there will be risk events such as early warning errors.

[0173] Specifically, by analyzing the operation data characteristics, it can be determined whether the operation data characteristics meet the second preset screening condition, and the operation data characteristics that meet the second preset screening condition are used as sample operation fault data characteristics, and the sample operation fault data characteristics are stored in the sample fault tracing database.

[0174] In some other embodiments, the operating data feature corresponds to a first cause, which is the fault cause causing the operating data feature. The first cause can be a cause that can cause the operating data feature to appear and is manually derived based on relevant knowledge and experience.

[0175] In some cases, not all operating data features can be stored in the fault tracing database. Instead, the first cause is used as the entry condition for the sample fault tracing database to ensure the true relevance and credibility of the operating data features stored in the sample fault tracing database with the battery fault. If the data is not entered into the database based on the first cause, there will be risk events such as early warning errors.

[0176] Specifically, by analyzing the cause of the operation data feature, the cause of the operation data feature, i.e., the first cause, can be determined. In practical applications, the correlation between the first cause and the operation data feature is verified to determine whether the first cause is the real cause of the operation data feature. If the first cause is the real cause, it can be considered that the first cause meets the fault feature screening condition, and the operation data feature can be stored in the sample fault tracing database.

[0177] Exemplarily, the operation data feature of the preset sample battery is represented by the battery voltage rising too fast and reaching the cut-off voltage first during the charging process. The cause analysis of the operation data feature of the preset sample battery can be considered that the reason why the preset sample battery produces the operation data feature is that compared with other preset sample battery cells, the internal resistance of this preset sample battery is smaller. The true correlation and accuracy between the smaller internal resistance of the preset sample battery and the fact that the voltage of the preset sample battery rises too fast and reaches the cut-off voltage first during the charging process is verified. If there is a true and credible correlation between the two, it can be considered that the first cause meets the fault feature screening condition, and the operation data feature that the voltage of the preset sample battery rises too fast and reaches the cut-off voltage first during the charging process can be used as a sample operation fault data feature that needs to be stored in the sample fault tracing database.

[0178] In the above implementation, by constructing a sample fault tracing database, the operating data features that meet the second preset screening condition are stored in the sample fault tracing database, which not only ensures the true relevance and credibility of the operating data features stored in the sample fault tracing database to the battery fault as much as possible, but also can precipitate the operating data features of the faulty battery, providing an accurate data source for battery fault analysis.

[0179] In some embodiments, see Figure 5a , obtaining the operating data characteristics of the faulty battery during operation may include the following steps:

[0180] S510: Acquire initial operation data of the faulty battery during operation.

[0181] S520: Clean the initial operation data to obtain target operation data.

[0182] The initial operation data may include but is not limited to the current I of the single battery. Cell , voltage V Cell , Capacity CAP Cell , battery state of charge SOC, temperature T Cell .

[0183] In some cases, the initial operating data of a faulty battery during operation may contain some interference data and incomplete data. Therefore, data cleaning is used to reduce the impact of noise or inaccurate data on the extraction of operating data features and improve the accuracy of operating data features.

[0184] Specifically, the vehicle to which the faulty battery belongs will send the initial operation data of the faulty battery during operation to the cloud server through the vehicle-side Tbox, and the cloud server will receive the initial operation data. The cloud server then cleans the initial operation data to obtain the target operation data.

[0185] S530: Extract features of the target operation data to obtain operation data features.

[0186] The operating data feature is used to characterize the situation where the target operating data of the faulty battery during operation is abnormal.

[0187] Specifically, by extracting features from the target operating data, the abnormal operating conditions of the target operating data during the operating process of the faulty battery can be obtained, and the operating data features related to the operating process can be obtained through the abnormal operating conditions. The feature extraction methods include but are not limited to derivation, threshold range, and other methods.

[0188] For example, see Figure 5b , Figure 5b This is the vehicle battery cell voltage diagram sent by the vehicle-side Tbox and after data cleaning. Figure 5b There is an operation data feature 310 in which the voltage of a certain battery rises faster than other batteries of the same period and the charging terminal voltage reaches the cut-off voltage first. This operation data feature 310 belongs to the classification fault data feature Y. The voltage of the battery can be compared with the data of the voltage of batteries of the same model, and the location of the abnormal voltage of the battery and the manifestation of the abnormal voltage can be extracted according to the threshold range of the battery voltage. Then the operation data feature of the battery can be expressed as the battery voltage rises too fast during the charging process and reaches the cut-off voltage first.

[0189] For example, different battery discharge rates may result in voltage differences between single cells. During battery discharge, the battery releases stored charge to supply current load, and at different discharge rates, the speed of the chemical reaction inside the battery and the rate of energy release will also be different. See Figure 5c , Figure 5c is the voltage diagram of the vehicle battery cell. According to the formula self-discharge rate = ((V1-V2) / (t-t0))*100%, we can get Figure 5c The self-discharge rates of different single cells are different. Figure 5c There is an operating data feature 510 in which the self-discharge rate of a certain single cell is obviously different from that of other single cells. This operating data feature 510 belongs to the classification fault data feature Z. The battery self-discharge rate is different, and the battery cells in the vehicle-end battery pack will have voltage differences. The accumulation of differences will cause faults such as battery pack discharge SOC limitation. The operating data feature of the battery can be expressed as the battery self-discharge rate is different, and the battery cells in the vehicle-end battery pack will have voltage differences. Among them, V1 is the battery voltage at the initial moment, V2 is the battery voltage after a period of time t, t0 is the initial moment, that is, the time before the self-discharge measurement begins, and t is the elapsed time.

[0190] In the above implementation, the initial operation data of the faulty battery during operation is obtained, and the initial operation data is cleaned to obtain the target operation data; the influence of noise or inaccurate data on the extraction of operation data features is reduced by data cleaning, thereby improving the accuracy of the operation data features.

[0191] In some embodiments, see Figure 6 , after obtaining the target production data characteristics, the following steps may be included:

[0192] S610. Classify the target production data features to obtain classified abnormal data features.

[0193] After obtaining the sample running fault data characteristics, including:

[0194] S620. Classify the sample operation fault data features to obtain classified fault data features.

[0195] S630: Perform correlation analysis on the classified abnormal data features and the classified fault data features to obtain production fault correlation data.

[0196] In some cases, by classifying the target production data features and the sample operation failure data features respectively and performing correlation analysis based on the classified features, rather than randomly and aimlessly conducting correlation mining on the target production data features and the sample operation failure data features, correlation analysis can be performed in a targeted manner, thereby improving the accuracy and credibility of the correlation analysis results to a certain extent.

[0197] Specifically, by classifying the target production data features, the classified abnormal data features can be obtained. By classifying the sample operation fault data features, the classified fault data features can be obtained. By performing correlation analysis on the classified abnormal data features and the classified fault data features, and mining the correlation between the classified abnormal data features and the classified fault data features, production fault correlation data describing the degree of correlation between the target production data features and the fault type can be obtained.

[0198] In the above implementation, correlation analysis is performed in combination with the classified abnormal data features and the classified fault data features, and the obtained correlation data can better reflect the true correlation between the battery operation failure and the production anomaly manifested in the battery production process, thereby realizing the correlation between the performance of the battery throughout its life cycle.

[0199] In some implementations, the analysis method for performing correlation analysis to obtain production failure correlation data includes at least one of clustering, variance analysis, and proportion statistics.

[0200] Specifically, by using at least one of clustering, variance analysis, and proportion statistics, the obtained production data and operation fault data can be subjected to correlation analysis, and production fault correlation data describing the degree of correlation between the target production data characteristics and the fault type can be obtained based on the results of the correlation analysis.

[0201] In this embodiment, the obtained production data can be understood as the characteristics of classified abnormal data, and the operation fault data can be understood as the characteristics of classified fault data; thus, at least one of clustering, variance analysis, and proportion statistics can be used to perform correlation analysis on the characteristics of classified abnormal data and the characteristics of classified fault data to obtain production fault correlation data. At least one of clustering, variance analysis, and proportion statistics can be used to perform correlation analysis on the characteristics of classified abnormal data and the characteristics of classified fault data, and according to the results of the correlation analysis, production fault correlation data describing the degree of correlation between the target production data characteristics and the fault type can be obtained.

[0202] For example, a correlation analysis is performed on the classified abnormal data features and the classified fault data features by using a percentage statistical method to obtain production fault correlation data. Figure 7For example, the number of preset battery samples is 10,000, the corresponding vehicle on which the preset battery sample is installed and operated is the sample vehicle, and the battery with the classification abnormal data feature M in the preset battery sample is the target battery. Among the 10,000 sample vehicles, the proportion of faulty vehicles with the classification fault data feature Y is 0.734%. Among the vehicles equipped with the target battery, the proportion of faulty vehicles with the classification fault data feature Y is 20.86%. The proportion of faulty vehicles in vehicles with abnormal battery production data characteristics is 20.86%, which is much higher than the proportion of faulty vehicles in the total number of vehicles, which is 0.734%. It can be considered that the abnormal production data characteristics are significantly correlated with vehicle battery failures. Among them, the proportion data between the two can be confirmed based on the difference in order of magnitude to confirm whether there is a significant difference in the proportion data between the two.

[0203] For example, cluster analysis can usually classify elements with significant correlation into the same cluster. Cluster analysis is performed on the classification abnormal data features and classification fault data features. If the classification abnormal data feature A and the classification fault data feature B are classified into the same cluster C through cluster analysis, it can be considered that there is a potential correlation between the classification abnormal data feature A and the classification fault data feature B, and further analysis is needed to obtain the corresponding production fault correlation data.

[0204] Exemplarily, the classification abnormal data features and the classification fault data features are used as independent variables, and the fault type is used as the dependent variable. The intra-group variation can be obtained through the classification abnormal data features and the classification fault data features, and the inter-group variation can be obtained through the independent variable classification abnormal data features and the classification fault data features and the dependent variable fault type. The analysis of the significant influence of the intra-group variation and the inter-group variation can determine the data that plays a decisive role.

[0205] In the above-mentioned method for creating early warning indicators, correlation analysis is performed using at least one of clustering, variance analysis, and proportion statistics to obtain production failure correlation data. Production failure correlation data can better reflect the true correlation between battery operation failures and production anomalies exhibited in the battery production process, and can realize the correlation between battery performance throughout its life cycle.

[0206] For example, see Figure 8 , the determination of production early warning indicators can include the following steps:

[0207] S810. Determine target related data that meets preset related screening conditions in the production failure related data.

[0208] S820. Use the target production data characteristics corresponding to the target related data as production early warning indicators.

[0209] Among them, there are many degrees of correlation between the target production data characteristics included in the production fault correlation data and the fault type. Therefore, preset relevant screening conditions are set to evaluate the correlation between the target production data characteristics and the fault type, and screen out the target related data whose correlation between the target production data characteristics and the fault type reaches the expected level.

[0210] Specifically, the production fault correlation data is filtered according to preset relevant filtering conditions to obtain target relevant data that meets the preset relevant filtering conditions, so that the target production data characteristics corresponding to the target relevant data are used as production early warning indicators.

[0211] In the above implementation, by determining the target related data that meets the preset related screening conditions, and using the corresponding target production data characteristics as production warning indicators, the preset related screening conditions are used to ensure that the abnormal production data characteristics that are significantly correlated with battery failure are used as production warning indicators as much as possible, thereby improving the accuracy of the production warning indicators.

[0212] The embodiment of this specification provides an electric energy device early warning method applied to a network device, wherein the electric energy device includes a battery. Fig. 9 The electric energy device early warning method may include the following steps:

[0213] S910. Obtain historical production data characteristics of the battery during the production process.

[0214] The early warning indicator is obtained according to the method for creating the early warning indicator in any of the above embodiments. The historical production data feature refers to the situation where the historical production data of the battery in the production process step deviates from the expected production data.

[0215] Specifically, by extracting features from the operating data of the battery on the electric energy device during operation, the operating data features of the battery on the electric energy device during operation can be obtained. If the operating data features indicate that the battery has operating fault data features related to the fault type during operation. In some embodiments, the historical production data features of the battery during the production process can be stored in the storage unit of the electric energy device in advance, so as to directly obtain the historical production data features of the battery during the production process from the storage unit. In some other embodiments, the actual production data of the battery during the production process is obtained, and the features of the historical production data are extracted from the actual production data to obtain the features of the historical production data.

[0216] S920. Match the historical production data characteristics with the warning indicators. If the match is successful, send a corresponding warning signal to the electric energy equipment, so that the electric energy equipment sends a warning prompt information according to the warning signal.

[0217] The method for creating the early warning indicator includes: performing correlation analysis based on sample production data and sample operation failure data of a preset sample battery to obtain production failure correlation data; the number of preset sample batteries is multiple. If the production failure correlation data is greater than a predetermined threshold, the early warning indicator is determined according to the sample production data corresponding to the sample operation failure data.

[0218] Specifically, after obtaining the historical production data features, the early warning indicators related to the fault type can be obtained from the storage unit of the electric energy device, or the early warning indicators related to the fault type can be obtained from the cloud server. Further, the acquired historical production data features are matched with the early warning indicators related to the fault type. If the two match, it indicates that the production process of the battery on the electric energy device is abnormal, and a corresponding early warning signal is sent to the electric energy device so that the electric energy device sends an early warning prompt information according to the early warning signal. If the two do not match, it indicates that the production process of the battery on the electric energy device is normal, and no early warning is currently required.

[0219] It should be noted that the number of electric energy devices in this embodiment can be one or more. When there are multiple electric energy devices, batch warnings can be issued for multiple electric energy devices to achieve batch warnings with higher accuracy. In the above-mentioned electric energy device warning method, the historical production data characteristics of the battery in the production process are obtained; if the historical production data characteristics match the warning indicators related to the fault type, a warning prompt message is issued. The warning is issued in combination with the data of the entire life cycle to improve the accuracy of the battery warning of the electric energy device.

[0220] The embodiment of the present specification provides an early warning method applied to an electric device, the electric device includes a battery, the method may include: receiving an early warning signal sent by a network device, and sending corresponding early warning prompt information according to the early warning signal; the early warning signal is determined by the network device according to the matching of the historical production data characteristics of the battery and the early warning index,

[0221] Among them, the method for creating early warning indicators includes: performing correlation analysis based on sample production data and sample operation failure data of preset sample batteries to obtain production failure correlation data; the number of preset sample batteries is multiple; if the production failure correlation data is greater than a predetermined threshold, determining the early warning indicator based on the sample production data corresponding to the sample operation failure data.

[0222] The electric device end may be any of various devices equipped with batteries, such as cars, drones, ships, aircraft, thermostats, computers, mobile phones, etc. Specifically, the network device determines the historical production data characteristics of the battery, matches the historical production data characteristics of the battery with the production warning indicators, and the network device determines the warning signal according to the matching situation. The network device sends the warning signal to the electric device end. The electric device end receives the warning signal sent by the network device, and then sends corresponding warning prompt information according to the warning signal.

[0223] In the above-mentioned early warning method, an early warning signal is sent to the electric equipment end according to the matching of the historical production data characteristics of the battery and the production early warning index, and then the corresponding early warning prompt information is sent according to the early warning signal. Early warning is carried out in combination with data of the entire life cycle to improve the accuracy of early warning at the electric equipment end.

[0224] The implementation mode of this specification provides an early warning method applied to an electric device, wherein the electric device includes a battery, and the electric device stores historical production data characteristics of the battery. The method includes: receiving early warning indicators issued by a network device; matching the historical production data characteristics with the early warning indicators, and if the match is successful, issuing corresponding early warning prompt information.

[0225] Among them, the method for creating early warning indicators includes: performing correlation analysis based on sample production data and sample operation failure data of preset sample batteries to obtain production failure correlation data; the number of preset sample batteries is multiple; if the production failure correlation data is greater than a predetermined threshold, determining the early warning indicator based on the sample production data corresponding to the sample operation failure data.

[0226] The electric device end may be any of various devices equipped with batteries, such as cars, drones, ships, aircraft, thermostats, computers, mobile phones, etc. Specifically, the network device sends a production warning indicator to the electric device end. The electric device end receives the production warning indicator sent by the network device. The historical production data characteristics of the battery stored locally are matched with the production warning indicator. If the match is successful, the corresponding warning prompt information is issued.

[0227] In the above-mentioned early warning method, by receiving the production early warning index and issuing corresponding early warning information according to the matching between the historical production data characteristics and the production early warning index, early warning is carried out in combination with the data of the whole life cycle to improve the accuracy of the early warning on the electric equipment side.

[0228] An embodiment of the present specification provides a network device, including a memory and a processor, wherein the memory stores a computer program, and the processor implements a method in any of the above embodiments when executing the computer program.

[0229] An embodiment of the present specification provides an electric energy device, including a memory and a processor, wherein the memory stores a computer program, and the processor implements a method in any of the above embodiments when executing the computer program.

[0230] In some embodiments, the electrical energy device is a vehicle, a watercraft, or an aircraft.

[0231] An embodiment of this specification provides a computer-readable storage medium. When instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device can execute a method in any of the above embodiments.

[0232] The embodiments of this specification provide a computer program product, including a computer program. When the computer program is executed by a processor, the method in any one of the above embodiments is implemented.

[0233] This specification provides a vehicle warning method. Fig.10 , the method may include the following steps:

[0234] S1010. If the power battery on the vehicle has operating fault data characteristics related to the target fault type during operation, obtain historical production data characteristics during the production process of the power battery and target production warning indicators related to the target fault type.

[0235] In some cases, if only the operating data generated by the power battery during operation is used for early warning, the abnormal characteristics of the battery in the production data will be ignored, which is the root cause of possible failure of the battery during operation. Therefore, in this embodiment, the operating data generated by the power battery during operation and the production data generated by the battery during production are combined to provide vehicle early warning from the perspective of the entire life cycle, combining the operating data and production data of the entire life cycle.

[0236] Specifically, the operation fault data features related to the target fault type can be preset. By extracting the features of the operation data of the power battery on the vehicle during operation, the operation data features related to the operation of the power battery on the vehicle can be obtained. If the operation data features indicate that the power battery has the operation fault data features related to the target fault type during operation, in order to warn the vehicle from the perspective of the entire life cycle, the historical production data features of the power battery during the production process and the target production warning indicators related to the target fault type can be further obtained.

[0237] In this embodiment, the production warning indicator can be stored in a storage unit on the vehicle side, and the production warning indicator can be directly obtained from the storage unit on the vehicle side. The production warning indicator can be obtained according to the creation method of the warning indicator in any of the above embodiments.

[0238] In this embodiment, the historical production data features can be stored in a storage unit on the vehicle side, and the historical production data features can be directly obtained from the storage unit on the vehicle side. The historical production data of the power battery can also be stored in the storage unit on the vehicle side, and the historical production data of the power battery can be obtained from the storage unit on the vehicle side, and feature extraction is performed based on the historical production data of the power battery to obtain the historical production data features.

[0239] In some cases, considering the computing power and storage space of the vehicle side, the operating fault data characteristics are determined on the vehicle side based on the operating data of the power battery during operation. The historical production data can be stored in server A, and server B obtains the historical production data of the power battery from server A. Server B extracts features from the historical production data of the power battery to obtain the historical production data characteristics. Server B can also store production warning indicators, and server B determines the target production warning indicators related to the target fault type. Server B sends the historical production data characteristics and the target production warning indicators related to the target fault type to the vehicle side, and the vehicle side obtains the historical production data characteristics and the target production warning indicators. It should be noted that servers A and B can be the same server or different servers.

[0240] In some cases, in order to further reduce the computing pressure on the vehicle side and reduce the storage space occupied by the vehicle side, the vehicle side sends the operating data of the power battery during operation to the server. The server determines the operating fault data characteristics based on the operating data of the power battery during operation, and as mentioned above, the extraction process of historical production data characteristics also occurs on the server, which will not be repeated here.

[0241] S1020. If the characteristics of the historical production data match the target production warning indicators, a warning prompt message is issued.

[0242] Specifically, the acquired historical production data features are matched with the target production warning indicators. If the two match, it indicates that the production process of the power battery on the vehicle is abnormal, and a warning prompt message is issued. If the two do not match, it indicates that the production process of the power battery on the vehicle is normal, and no warning is currently required.

[0243] In the above vehicle early warning method, if the power battery on the vehicle has an operating fault data feature related to the target fault type during operation, the historical production data feature of the power battery during production and the target production early warning index related to the target fault type are obtained; if the historical production data feature matches the target production early warning index, an early warning prompt message is issued. The vehicle early warning is performed by combining the production data and operation data of the battery throughout its life cycle to improve the accuracy of the early warning.

[0244] The embodiment of this specification also provides a method for creating an early warning indicator, wherein the operating fault data features are stored in a fault tracing database; the target production data features are stored in an abnormal feature database. For example, refer to Fig.11 , the method for creating the early warning indicator may include the following steps:

[0245] S1102: Acquire initial operation data of the faulty battery during operation.

[0246] S1104: Clean the initial operation data to obtain target operation data.

[0247] S1106: Extract features of the target operation data to obtain operation data features related to the operation process.

[0248] Among them, the operating data feature corresponds to the first cause.

[0249] S1108: Taking the operation data characteristics of the first generation cause that meet the fault characteristic screening condition as the operation fault data characteristics that need to be stored in the fault tracing database.

[0250] S1110. Acquire actual production data of the battery during the production process.

[0251] Among them, the actual production data corresponds to the expected production data.

[0252] S1112. Establish a production database based on actual production data generated during the process steps of the battery production process.

[0253] S1114. Determine the actual production data whose difference with the expected production data exceeds a preset difference threshold range as abnormal production data.

[0254] S1116. Extract features of abnormal production data to obtain production data features.

[0255] Among them, the production data feature corresponds to the second generation cause.

[0256] S1118. The production data feature whose second generation cause satisfies the abnormal feature screening condition is taken as the target production data feature that needs to be stored in the abnormal feature database.

[0257] The target production data feature is used to characterize the situation where the actual production data of the battery deviates from the expected production data of the battery;

[0258] S1120, obtaining classified abnormal data features obtained by classifying target production data features, and classified fault data features obtained by classifying operation fault data features.

[0259] S1122. Use at least one of clustering, variance analysis, and proportion statistics to perform correlation analysis on the classified abnormal data features and the classified fault data features to obtain production fault correlation data.

[0260] Among them, the production failure correlation data is used to characterize the degree of correlation between the target production data characteristics and the failure type.

[0261] S1124. Determine target related data that meets preset related screening conditions in the production failure related data.

[0262] S1126. Use the target production data characteristics corresponding to the target related data as production early warning indicators.

[0263] This specification embodiment provides a device 1200 for creating an early warning indicator. Fig.12 The early warning indicator creation device 1200 includes: a data feature acquisition module 1210, a correlation analysis module 1220, and an early warning indicator determination module 1230.

[0264] The data feature acquisition module 1210 is used to acquire target production data features during the production process of the battery, and operation fault data features related to the fault type during the operation process of the faulty battery;

[0265] The correlation analysis module 1220 is used to perform correlation analysis using the target production data characteristics and the operation fault data characteristics to obtain production fault correlation data; wherein the production fault correlation data is used to characterize the degree of correlation between the target production data characteristics and the fault type;

[0266] The warning indicator determination module 1230 is used to determine the production warning indicator related to the fault type in the target production data characteristics according to the production fault correlation data.

[0267] This specification provides a battery warning device 1300, see Fig.13 The battery warning device 1300 includes: a production feature acquisition module 1310 and a warning prompt issuing module 1320.

[0268] The production feature acquisition module 1310 is used to obtain actual production data features of the battery to be evaluated during the production process;

[0269] The warning prompt issuing module 1320 is used to issue a warning prompt message when the actual production data characteristics of the battery to be evaluated match the warning index; wherein the method for creating the warning index includes:

[0270] Performing correlation analysis on sample production data and sample operation failure data of preset sample batteries to obtain production failure correlation data; the number of the preset sample batteries is multiple;

[0271] If the production fault correlation data is greater than a predetermined threshold, a warning indicator is determined according to the sample production data corresponding to the sample operation fault data.

[0272] This specification provides a vehicle warning device 1400, see Fig.14 The vehicle warning device 1400 includes: a production feature acquisition module 1410 and a warning prompt issuing module 1420.

[0273] The production feature acquisition module 1410 is used to acquire historical production data features during the production process of the power battery if an operation fault data feature related to the fault type occurs during the operation of the power battery on the vehicle;

[0274] The warning prompt issuing module 1420 is used to issue a warning prompt message if the historical production data characteristics match the production warning index related to the fault type; wherein the production warning index is obtained according to the method described in any of the above implementation methods.

[0275] This specification provides a vehicle warning device 1500, see Fig.15 The vehicle warning device 1500 includes: a characteristic indicator acquisition module 1510 and a warning prompt issuing module 1520.

[0276] The characteristic index acquisition module 1510 is used to acquire the historical production data characteristics in the production process of the power battery and the target production warning index related to the target fault type if the power battery on the vehicle has an operation fault data characteristic related to the target fault type during operation;

[0277] The early warning prompt issuing module 1520 is used to issue an early warning prompt message if the historical production data characteristics match the target production early warning index.

[0278] This specification provides a battery warning device 1600, see Fig.16 The battery warning device 1600 includes: a production feature acquisition module 1610 and a warning prompt issuing module 1620.

[0279] The production feature acquisition module 1610 is used to obtain actual production data features of the battery to be evaluated during the production process;

[0280] The warning prompt issuing module 1620 is used to issue a warning prompt message when the actual production data characteristics of the battery to be evaluated match the production warning index; wherein the production warning index is obtained according to the method described in any of the above embodiments.

[0281] The embodiment of this specification provides a vehicle warning device 1700, the vehicle includes a power battery. Fig.17 The vehicle warning device 1700 includes a production feature acquisition module 1710 and a warning prompt issuing module 1720.

[0282] A production feature acquisition module 1710 is used to acquire historical production data features during the production process of the power battery;

[0283] The warning prompt issuing module 1720 is used to match the historical production data characteristics with the production warning indicators. If the match is successful, a corresponding warning signal is issued to the vehicle, so that the vehicle issues a warning prompt information according to the warning signal; the method for creating the warning indicator includes:

[0284] Performing correlation analysis on sample production data and sample operation failure data of preset sample batteries to obtain production failure correlation data; the number of the preset sample batteries is multiple;

[0285] If the production fault correlation data is greater than a predetermined threshold, the sample production data corresponding to the sample operation fault data is run.

[0286] For the specific description of the warning indicator creation device, the battery warning device and the vehicle warning device, please refer to the description of the indicator creation method, the warning indicator creation method, the battery warning method and the vehicle warning method above, which will not be repeated here. Each module in the above device can be implemented in whole or in part by software, hardware and a combination thereof. The above modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0287] An embodiment of the present specification provides a server, including a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the method in any one of the above embodiments when executing the computer program.

[0288] An embodiment of the present specification provides a vehicle, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of the method in any of the above embodiments are implemented.

[0289] It should be noted that the logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection portion with one or more wirings (electronic device), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or otherwise processing in a suitable manner if necessary, and then stored in a computer memory.

Claims

1. A device early warning method, the device comprising subcomponents, characterized in that: The early warning method comprises: Obtaining historical production data characteristics of the subassembly during the production process; If the historical production data characteristics of the subcomponent match the warning indicator, a warning prompt message is issued; Among them, the early warning indicator is created by obtaining production failure correlation data based on sample production data and sample operation failure data of preset sub-component samples; if the production failure correlation data is greater than a predetermined threshold, the early warning indicator is determined according to the sample production data corresponding to the sample operation failure data; wherein the number of the preset sub-component samples is multiple.

2. The method according to claim 1, characterized in that If the historical production data characteristics of the subcomponent match the warning index, a warning prompt message is issued to the subcomponent.

3. The method according to claim 1, characterized in that The device is an electric energy device, the subassembly is a battery, and the preset subassembly sample is a preset sample battery; Before the correlation analysis is performed between the sample production data based on the preset sub-assembly sample and the sample operation failure data, the method further includes: Acquire sample actual production data of the preset sample battery during the production process; wherein the sample actual production data corresponds to sample expected production data; According to the comparison result between the sample actual production data and the sample expected production data, abnormal production data is determined in the sample actual production data as the sample production data.

4. The method according to claim 3, characterized in that The method of determining the abnormal production data includes at least one of the following steps: determining the sample actual production data different from the sample expected production data as the abnormal production data; The sample actual production data whose difference data with the sample expected production data exceeds a preset difference threshold range is determined as the abnormal production data.

5. The method according to claim 3, characterized in that: The correlation analysis based on the sample production data and the sample operation failure data includes: Extracting features of the abnormal production data to obtain target production data features; A correlation analysis is performed on the target production data features and the sample operation fault data to obtain the production fault correlation data.

6. The method according to claim 5, characterized in that After extracting features from the abnormal production data, the method further includes: Extracting features from the abnormal production data to obtain features of the abnormal production data; The abnormal production data features that meet the first preset screening condition are screened out as the target production data features.

7. The method according to claim 5 or 6, characterized in that: Before performing the correlation analysis based on the sample production data and the sample operation fault data, the method further includes: The sample operation fault data features are obtained from the sample fault tracing database as the sample operation fault data.

8. The method according to claim 7, characterized in that The method for constructing the sample fault tracing database includes: Acquiring operation data characteristics of a faulty battery during operation; the faulty battery is the preset sample battery that fails during operation; Filtering out the operation data features that meet the second preset screening condition as the sample operation fault data features; The sample operation fault data characteristics are stored in the sample fault tracing database.

9. The method according to claim 8, characterized in that The step of obtaining the operating data characteristics of the faulty battery during operation includes: Acquiring initial operation data of the faulty battery during operation; Performing data cleaning on the initial operation data to obtain target operation data; Feature extraction is performed on the target operation data to obtain the operation data feature.

10. The method according to claim 7, characterized in that After obtaining the target production data characteristics, the following steps are included: Classified abnormal data features obtained by classifying the target production data features; After obtaining the sample operation fault data characteristics, the method includes: Classified fault data features obtained by classifying the sample operation fault data features; A correlation analysis is performed on the classified abnormal data features and the classified fault data features to obtain the production fault correlation data.

11. The method according to any one of claims 1 to 6 and 8 to 10, characterized in that The analysis method of performing correlation analysis to obtain the production failure correlation data includes at least one of clustering, variance analysis, and proportion statistics.

12. An electric energy device early warning method applied to a network device, wherein the electric energy device comprises a battery, characterized in that: The method comprises: Obtaining historical production data characteristics of the battery during the production process; Matching the historical production data characteristics with the early warning index, if the match is successful, sending a corresponding early warning signal to the electric energy equipment, so that the electric energy equipment sends out early warning prompt information according to the early warning signal; the method for creating the early warning index includes: Performing correlation analysis on sample production data and sample operation failure data of preset sample batteries to obtain production failure correlation data; the number of the preset sample batteries is multiple; If the production fault correlation data is greater than a predetermined threshold, a warning indicator is determined according to the sample production data corresponding to the sample operation fault data.

13. An early warning method applied to an electric energy device, wherein the electric energy device comprises a battery, characterized in that: The method comprises: Receive a warning signal sent by a network device, and send corresponding warning prompt information according to the warning signal; the warning signal is determined by the network device according to the matching situation between the historical production data characteristics of the battery and the warning indicator, and the creation method of the warning indicator includes: Performing correlation analysis on sample production data and sample operation failure data of preset sample batteries to obtain production failure correlation data; the number of the preset sample batteries is multiple; If the production fault correlation data is greater than a predetermined threshold, a warning indicator is determined according to the sample production data corresponding to the sample operation fault data.

14. An early warning method applied to an electric device, the electric device comprising a battery, characterized in that: The electric device stores historical production data characteristics of the battery, and the method includes: Receive early warning indicators from network devices; Matching the historical production data characteristics with the early warning indicators, and if the match is successful, issuing corresponding early warning prompt information; The method for creating the early warning indicator includes: Performing correlation analysis on sample production data and sample operation failure data of preset sample batteries to obtain production failure correlation data; the number of the preset sample batteries is multiple; If the production fault correlation data is greater than a predetermined threshold, a warning indicator is determined according to the sample production data corresponding to the sample operation fault data.

15. A network device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the method according to any one of claims 1 to 12 when executing the computer program.

16. An electric energy device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the method according to claim 13 or 14 when executing the computer program.

17. The electric energy device according to claim 16, characterized in that The electric energy device is a vehicle, a ship or an aircraft.

18. A computer-readable storage medium, characterized in that: When the instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is enabled to perform the method as claimed in any one of claims 1 to 14.

19. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the method according to any one of claims 1 to 14 is implemented.

20. A battery warning device, characterized in that: The device comprises: A production feature acquisition module is used to obtain actual production data features of the battery to be evaluated during the production process; The early warning prompt issuing module is used to issue an early warning prompt message when the actual production data characteristics of the battery to be evaluated match the early warning index; wherein the method for creating the early warning index includes: Performing correlation analysis on sample production data and sample operation failure data of preset sample batteries to obtain production failure correlation data; the number of the preset sample batteries is multiple; If the production fault correlation data is greater than a predetermined threshold, a warning indicator is determined according to the sample production data corresponding to the sample operation fault data.

21. A vehicle warning device, characterized in that: The vehicle includes a power battery, and the device includes: A production feature acquisition module, used to acquire historical production data features during the production process of the power battery; The warning prompt issuing module is used to match the historical production data characteristics with the production warning indicators. If the match is successful, a corresponding warning signal is issued to the vehicle, so that the vehicle issues a warning prompt information according to the warning signal; the creation method of the warning indicator includes: Performing correlation analysis on sample production data and sample operation failure data of preset sample batteries to obtain production failure correlation data; the number of the preset sample batteries is multiple; If the production fault correlation data is greater than a predetermined threshold, a warning indicator is determined according to the sample production data corresponding to the sample operation fault data.

Images