System, method and equipment for energy storage battery safety algorithm online research and development platform

Abstract

The invention discloses a system, a method and equipment for an energy storage battery safety algorithm online research and development platform. The system comprises a data source selection subsystem, a data point tuple array and a data point tuple array, wherein the data source selection subsystem is used for constructing a temporary data point tuple array for algorithm processing; the algorithm development subsystem is used for executing a test on an algorithm written by a user on a platform so as to transmit generated configuration information to the scheduler subsystem for task arrangement; the scheduler subsystem is used for constructing an algorithm mirror image according to a scheduling rule in the configuration information, storing the algorithm mirror image in a mirror image warehouse, and deploying an algorithm to a container arrangement platform for instantiation; and the trusteeship application subsystem is used for storing a result as a file, a database record or a machine learning model after the algorithm operation is completed, and providing corresponding data acquisition, RESTful API or model deployment services. Through the scheme of the invention, analysis can be carried out according to the mechanism of the battery and the use condition of the user, and algorithm research and development are carried out through the data generated by the energy storage energy management system to finally carry out trusteeship online.

Description

Technical Field

[0001] The present application relates to the field of computer systems, and in particular to a system, method and device for an online research and development platform for energy storage battery safety algorithms. Background Art

[0002] Conduct research on key core technologies, equipment and integrated optimization design of sodium ion batteries, new lithium ion batteries, lead-carbon batteries, flow batteries, compressed air, hydrogen (ammonia) energy storage, heat (cold) energy storage, etc., focus on tackling energy storage technologies such as superconductors and supercapacitors, and develop a new generation of high energy density energy storage technologies such as liquid metal batteries, solid-state lithium ion batteries, and metal air batteries. Breakthroughs in key technologies such as full-process safety technology, intrinsic safety control of batteries, safety warning of electrochemical energy storage systems, multi-level protection structures and key materials of the system, efficient fire extinguishing and prevention of re-ignition, and overall safety design of energy storage power stations support the safe operation of large-scale energy storage power stations. Breakthroughs in energy storage battery cycle life rapid detection and aging status evaluation technology, develop retired battery health assessment, sorting, repair and other cascade utilization related technologies, and study the control, protection and security defense technology of new multi-energy storage access to the power grid system.

[0003] At present, the research on the safety of the energy storage system big data platform is still in its infancy. It is particularly important to ensure the safety of the energy storage system through algorithms to reduce the probability of thermal runaway. This requires the analysis of a large amount of data from the energy storage system and a large amount of statistical calculations based on the battery mechanism. However, due to its own development conditions, algorithm development is currently carried out by battery experts who analyze a large number of data samples in order to develop safe and reliable algorithms to ensure system safety. At present, the closest technology to the algorithm platform is the algorithm development and online deployment of machine learning by Internet companies for a large amount of data, forming a complete data chain. However, the energy storage industry needs to analyze the battery mechanism and user usage, and this type of algorithm platform has not yet appeared.

[0004] Therefore, there is an urgent need for a technical solution that can analyze the battery mechanism and user usage, and use the data generated by the energy storage energy management system to develop algorithms and ultimately host them online. Summary of the invention

[0005] In order to solve the deficiencies of the prior art, the embodiments of the present application provide a system, method and device for an online R&D platform for energy storage battery safety algorithms. The present application solves the technical problems that the prior art cannot analyze the mechanism of the battery and the user's usage.

[0006] The embodiment of the present application provides a system for an online R&D platform for energy storage battery safety algorithms, including: a data source selection subsystem, an algorithm development subsystem, a scheduler subsystem and a hosted application subsystem; wherein the data source selection subsystem is used to configure real-time, historical or file data sources, build a temporary data point tuple array for algorithm processing, and automatically release the temporary data after the processing is completed; the algorithm development subsystem is used to perform tests on the algorithms written by users on the platform, store them in the object storage system after review, and obtain the user's scheduling policy to transmit the generated configuration information to the scheduler subsystem for task scheduling; the scheduler subsystem is used to build an algorithm image according to the scheduling rules in the configuration information and store it in the image warehouse, and after obtaining the user's specified output target, deploy the algorithm to the container orchestration platform for instantiation; the hosted application subsystem is used to store the results as files, database records or machine learning models after the algorithm operation is completed, and provide corresponding data acquisition, RESTful API or model deployment services based on the usage scenario requirements.

[0007] In a possible implementation, a real-time, historical or file data source is configured, a temporary data point tuple array is constructed for algorithm processing, and the temporary data is automatically released after the processing is completed, including: selecting a data source type, the data source type including online selection of real-time data points, retrieval of historical data points and custom upload of data files; generating a temporary data point tuple array according to the selected data source type, the array containing time, voltage and / or current information to construct multidimensional data; displaying some sample data to the user, and automatically clearing the corresponding related data in the cache and / or file after the data is used.

[0008] In a possible implementation, the algorithm written by the user on the platform is tested, stored in the object storage system after review, and the user's scheduling strategy is obtained to transmit the generated configuration information to the scheduler subsystem for task scheduling, including: providing algorithm templates for users to download, or providing online editors for users to develop algorithms to obtain algorithms submitted by users; testing and debugging the algorithms to receive algorithms that have passed debugging; reviewing the received algorithm codes, and storing the reviewed algorithm codes in the target object storage system; receiving the algorithm scheduling rules selected by the user, and generating tasks containing the image addresses of the algorithms to be scheduled according to the scheduling rules; generating configuration information, wherein the configuration information includes data cache addresses, scheduling rules, the image of the algorithm to be scheduled, and the result storage type.

[0009] In a possible implementation, an algorithm image is constructed according to the scheduling rules in the configuration information and stored in the image repository. After obtaining the user's specified output target, the algorithm is deployed to the container orchestration platform for instantiation, including: generating a corresponding algorithm image according to preset scheduling rules, and storing the algorithm image in the image repository; receiving user instructions, saving the algorithm processing results to a file, a database and providing a REST interface, or hosting the model and providing a REST interface according to the result saving type specified by the user; deploying the algorithm image to the container orchestration system according to preset rules, and performing online operations to provide algorithm services.

[0010] In a possible implementation, after the algorithm operation is completed, the result is stored as a file, database record or machine learning model, and based on the usage scenario requirements, corresponding data acquisition, RESTful API or model deployment services are provided, including: after the algorithm calculation is completed, the algorithm result is stored and processed according to the algorithm result storage location in the user's configuration information; if the algorithm result storage configuration is file export, the algorithm result file is stored in the object storage module, and a data list for download is generated; if the algorithm result storage configuration is database storage, the algorithm result is stored in the database; if the algorithm result is a machine learning model, the trained machine learning model is stored in the object storage module; for the algorithm result data stored in the database, a REST interface for query is generated; for the machine learning model stored in the object storage module, a model hosting service is generated and deployed online, and a REST interface service for model prediction is provided.

[0011] In a possible implementation, before displaying part of the sample data to the user, it also includes: randomly generating encryption parameters including a key, an offset, a modulus and a mask value; encrypting each tuple in the temporary data point tuple array in a chronological order based on the encryption parameters and a preset encryption function; selecting sample data at fixed intervals from the encrypted tuple array to construct a sample array for display.

[0012] In a possible implementation, the preset encryption function includes: Where t represents the time point of data collection, X represents the vector containing all battery parameter values, x1, x2, ..., x m represents the value of each battery parameter, k represents the encryption key, r1, r2, ..., r m represents a random mask value, represents the bitwise XOR operator, b represents the offset, m represents the modulus, and f(t, X) represents the encryption function.

[0013] The embodiment of the present application also provides a method for an online R&D platform for energy storage battery safety algorithms, including: configuring real-time, historical or file data sources, building a temporary data point tuple array for algorithm processing, and automatically releasing temporary data after the processing is completed; executing tests on the algorithms written by users on the platform, storing them in the object storage system after review, and obtaining the user's scheduling policy to transmit the generated configuration information to the scheduler subsystem for task scheduling; building an algorithm image according to the scheduling rules in the configuration information and storing it in the image warehouse, and after obtaining the user's specified output target, deploying the algorithm to the container orchestration platform for instantiation; after the algorithm operation is completed, storing the results as files, database records or machine learning models, and providing corresponding data acquisition, RESTful API or model deployment services based on usage scenario requirements.

[0014] An embodiment of the present application also provides a device for an online R&D platform of energy storage battery safety algorithms, comprising: a processor, a memory, and a system bus; wherein the processor and the memory are connected via the system bus; the memory is used to store one or more programs, wherein the one or more programs include instructions, and when the instructions are executed by the processor, the processor executes the method described in the above embodiment.

[0015] In the system, method and device for an online R&D platform for energy storage battery safety algorithms provided above, the embodiment of the present application can analyze the battery mechanism and user usage through an algorithm development platform based on an energy storage big data platform, and conduct algorithm development and final hosting and online launch through the data generated by the energy storage energy management system. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0017] Figure 1 A schematic block diagram of a system for an online R&D platform for energy storage battery safety algorithms provided in an embodiment of the present application;

[0018] Figure 2 A flow chart of a method for an online R&D platform for energy storage battery safety algorithms provided in an embodiment of the present application;

[0019] Figure 3 A flowchart of another method for an online R&D platform for energy storage battery safety algorithms provided in an embodiment of the present application. DETAILED DESCRIPTION

[0020] Various exemplary embodiments of the present disclosure will now be described in detail with reference to the accompanying drawings. It should be noted that the relative arrangement of components and steps, numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present disclosure unless otherwise specifically stated.

[0021] Those skilled in the art can understand that the terms "first", "second" and the like in the embodiments of the present disclosure are only used to distinguish different steps, devices or modules, etc., and neither represent any specific technical meaning nor represent the necessary logical order between them. It should also be understood that in the embodiments of the present disclosure, "multiple" can refer to two or more, and "at least one" can refer to one, two or more. It should also be understood that for any component, data or structure mentioned in the embodiments of the present disclosure, in the absence of explicit limitation or contrary revelation given in the context, it can generally be understood as one or more. In addition, the term "and / or" in the present disclosure is only a description of the association relationship of the associated objects, indicating that there can be three relationships, for example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in the present disclosure generally indicates that the associated objects before and after are an "or" relationship. It should also be understood that the description of each embodiment in the present disclosure emphasizes the differences between the embodiments, and the same or similar parts can refer to each other. For the sake of brevity, they will not be repeated one by one.

[0022] At the same time, it should be understood that, for ease of description, the sizes of the various parts shown in the drawings are not drawn according to the actual proportional relationship. The following description of at least one exemplary embodiment is actually only illustrative and is by no means intended to limit the present disclosure and its application or use. The techniques, methods and devices known to ordinary technicians in the relevant fields may not be discussed in detail, but where appropriate, the techniques, methods and devices should be considered part of the specification. It should be noted that similar numbers and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further discussed in subsequent drawings.

[0023] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0024] Figure 1A schematic block diagram of a system for an online R&D platform for energy storage battery safety algorithms provided in an embodiment of the present application. It should be understood that the system shown in the figure is exemplary and not restrictive. This means that the system architecture involved is not limited to a specific form or design, but is presented as an example. In other words, the architecture shown in the figure can be regarded as a way of expression to clearly describe related concepts and relationships, and does not exclude other forms of architecture. Therefore, when interpreting the architecture in the picture, it should be understood that the model is flexible and diverse, and its purpose is to provide an exemplary description rather than a restrictive provision on a specific form.

[0025] The embodiment of the present application provides a system for an online R&D platform for energy storage battery safety algorithms, including: a data source selection subsystem 101, an algorithm development subsystem 102, a scheduler subsystem 103 and a hosted application subsystem 104; wherein the data source selection subsystem 101 is used to configure real-time, historical or file data sources, build a temporary data point tuple array for algorithm processing, and automatically release temporary data after processing is completed; the algorithm development subsystem 102 is used to perform tests on the algorithms written by users on the platform, store them in the object storage system after review, and obtain the user's scheduling policy to transmit the generated configuration information to the scheduler subsystem for task scheduling; the scheduler subsystem 103 is used to build an algorithm image according to the scheduling rules in the configuration information and store it in the image warehouse, and after obtaining the user's specified output target, deploy the algorithm to the container orchestration platform for instantiation; the hosted application subsystem 104 is used to store the results as files, database records or machine learning models after the algorithm operation is completed, and provide corresponding data acquisition, RESTful API or model deployment services based on the usage scenario requirements. (such as Figure 2 The specific process will be shown in Figure 2 (see the instructions for more information).

[0026] Figure 2 A flow chart of a method for an online R&D platform for energy storage battery safety algorithms provided in an embodiment of the present application. Figure 2 As shown, at step S201, a real-time, historical or file data source is configured, a temporary data point tuple array is constructed for algorithm processing, and the temporary data is automatically released after the processing is completed.

[0027] In one embodiment, a data source type is selected, and the data source type includes online selection of real-time data points, retrieval of historical data points, and custom upload of data files; a temporary data point tuple array is generated according to the selected data source type, and the array contains time, voltage and / or current information to construct multidimensional data; some sample data is displayed to the user, and after the data is used, the corresponding related data in the cache and / or file is automatically cleared.

[0028] Specifically, the data source selection system provides users with three flexible and diverse data source selection methods to meet data requirements in different scenarios. The first method is to view and select real-time data points uploaded by the energy management system. In this way, users can obtain real-time operating data of energy storage system equipment, such as key parameters such as battery cell voltage, battery cluster current, and battery cluster discharge power. These real-time data points can reflect the current status and performance of the equipment, providing an important basis for subsequent analysis and decision-making.

[0029] The second way is to view and search the historical data points in the database and configure the required data time period. The historical data points record the operation of the equipment in the past period of time. By analyzing these data, you can find the performance trends and abnormal conditions of the equipment, and provide a reference for optimizing equipment operation and preventing failures. Users can flexibly configure the required data time period according to their needs to obtain historical data for a specific period. It is generated by querying the data of the time series database in a certain time interval. For example, the returned results are time (in seconds), battery cluster terminal voltage, battery cluster terminal current, and battery cluster discharge power. Generate a temporary data point tuple array through the acquired historical data. The sample data is: a = [[2023-05-1000:00:00, 375, 1.5, 3.77], [2023-05-1000:00:01, 376, 1.6, 4.72], ...], which means that the data at 0 seconds on 2023-05-10 is 375mV for the battery cluster terminal voltage, 1.5A for the battery cluster terminal current, and 3.77kW for the battery cluster discharge power, and so on. Time can be replaced by 0, 1, 2... to represent 0 seconds, 1 second, and 2 seconds.

[0030] The third method is for users to provide data files that conform to the platform's data file template, and the platform will parse all the data points in the file. This method provides users with greater freedom, allowing users to use their own data files as long as they conform to the platform's data file template requirements. The platform will automatically parse the data files provided by users, extract the data point information, and generate a standardized data format to facilitate subsequent processing and analysis.

[0031] Regardless of which data source the user chooses, the platform will convert the selected data points into a unified tuple array format for the algorithm platform to call. Each element in the tuple array is a tuple, representing a data record, and each tuple contains multiple attributes, such as timestamp, voltage, current, power, etc. This standardized data format can simplify the design and implementation of the algorithm and improve the efficiency of data processing.

[0032] For different data sources, the system uses different implementation methods to acquire and process data. For the real-time data source of the energy management system, the system will periodically collect real-time data from the equipment, such as battery cluster terminal voltage, current, discharge power, etc., and then convert these data into a temporary data point tuple array, where the timestamp can be replaced by numbers such as 0, 1, 2, etc., indicating the relative time of data collection. For file data sources, users need to provide a data file that conforms to the cSv format template. Each column in the file represents a parameter, such as time, voltage, current, power, etc. The system will parse the file and generate a temporary data point tuple array in the same format as the historical data.

[0033] It should be noted that the temporary data point tuple array is a temporary data structure that only exists in the memory during the algorithm application. Once the algorithm is executed, the data will be cleared from the cache and file to protect the security of business data. To further protect data privacy, users can only view sample data after selecting data, and cannot see all the original data.

[0034] The data source selection system provides high-quality and reliable data support for the algorithm platform by providing a variety of flexible data selection methods, standardized data formats, and secure data protection mechanisms.

[0035] In addition, before showing some sample data to the user, it also includes randomly generating encryption parameters including key, offset, modulus and mask value; encrypting each tuple in the temporary data point tuple array in time sequence based on the encryption parameters and the preset encryption function; selecting sample data at fixed intervals from the encrypted tuple array to construct a sample array for display. The preset encryption function includes, assuming that the original tuple array is A, which contains n elements, each tuple contains a timestamp t and multiple battery parameter values ​​x1, x2, ..., x m Let f(t, X) be the encryption function for a single tuple, defined as follows: Where t represents the time point of data collection, X represents the vector containing all battery parameter values, x1, x2, ..., x m represents the value of each battery parameter, k represents the encryption key, r1, r2, ..., r m represents a random mask value, represents the bitwise XOR operator, b represents the offset, m represents the modulus, and f(t, X) represents the encryption function.

[0036] Perform f(t, X) transformation on each tuple (t, X) of the original array A to obtain the encrypted tuple array A'=[f(t1, X1), f(t2, X2), ..., f(t n , X n)], and then select elements at fixed intervals from A' (such as one every 10) to form a sample data subset S = [A'[i]|i = 1, 11, 21, ...]. Finally, S is provided to the user for preview.

[0037] The optimized encryption formula is more suitable for the online R&D platform of energy storage battery safety algorithm: In view of the characteristics of time series data, the timestamp information is retained during the encryption process, which is convenient for users to understand the time span and granularity of the data. According to the data type and value range of different battery parameters, the mask XOR method is used for differentiated processing to improve data security. Selecting sample data at fixed intervals not only reduces the amount of data, but also reflects the overall trend and change law of the data. In combination with the actual needs of the platform, an encryption and sample data generation scheme that is more in line with actual application scenarios is designed. In this way, the platform can provide users with the necessary data preview function while ensuring the security of the original business data, helping users to better select and use data resources.

[0038] At step S202, the algorithm written by the user on the platform is tested, stored in the object storage system after review, and the user's scheduling strategy is obtained to transmit the generated configuration information to the scheduler subsystem for task scheduling.

[0039] In one embodiment, an algorithm template is provided for users to download, or an online editor is provided for users to develop algorithms, so as to obtain algorithms submitted by users; the algorithms are tested and debugged to receive algorithms that have passed debugging; the received algorithm codes are reviewed, and the reviewed algorithm codes are stored in a target object storage system; the algorithm scheduling rules selected by the user are received, and tasks containing the addresses of the algorithm images to be scheduled are generated according to the scheduling rules; configuration information is generated, and the configuration information includes data cache addresses, scheduling rules, the algorithm images to be scheduled, and the result saving types.

[0040] Specifically, users can develop, test and schedule algorithms through the system. First, users need to download the algorithm template to the local computer or open it directly on the platform page. Users can develop the algorithm in their own editor and then upload it to the algorithm platform, or edit it directly on the platform. However, other files in the algorithm template cannot be modified. When editing the algorithm, users can test the algorithm and view the results of the algorithm process, but they need permission to view the process results.

[0041] After the user successfully self-tests the algorithm, he can upload and submit the algorithm to the platform. After submission, the code will be manually reviewed. After the review is passed, the algorithm code will be submitted to the object storage for future scheduler calls. Finally, the user selects the scheduling rules, which include: real-time running algorithms, one-time running algorithms, machine learning algorithms, and periodic scheduling algorithms. Then the user's results are sent to the scheduler for orchestration.

[0042] Before developing the algorithm, the system will generate a temporary data point tuple array through the data source selection system. The temporary data can be sorted and combined in the background and stored in the cache and file system. After downloading the algorithm template and developing the algorithm, the data can be directly obtained from the cache for algorithm operation, saving the time required for data sorting.

[0043] Users can test algorithms on the platform, that is, they can edit algorithm codes directly in the browser. Because the selected data source can be used directly for algorithm debugging, users can debug directly on the platform. If there is no problem with the output result and the code does not report an error, it means that the algorithm test is successful.

[0044] Algorithm audit is a review of the code by the backend administrator of the algorithm platform to check whether the code has any illegal behavior, such as whether there is any data crawling behavior, whether there is any code behavior that violates regulations, etc.

[0045] The algorithm to be scheduled is generated by the process of packaging the program into a code software image through Docker. Docker is an open source application container engine that allows developers to package their applications and dependent packages into a portable image and then publish it to any popular Linux or Windows operating system machine. The packaged image can be directly run in the Docker environment.

[0046] The scheduling rules are selected by the user according to their own needs. After selecting the scheduling rules, the platform can directly generate the Job task of kubernetes, which will contain the image address of the algorithm to be scheduled, so that kubemetes can pull the code software image. Kubernetes is an open source for managing containerized applications on multiple hosts in the cloud platform. The goal of Kubernetes is to make the deployment of containerized applications simple and efficient. Kubernetes provides a mechanism for application deployment, planning, updating, and maintenance. Scheduling messages include: cache address of temporary data, scheduling rules, software image of the algorithm to be scheduled, and result storage type.

[0047] In step S203, an algorithm image is constructed according to the scheduling rules in the configuration information and stored in the image repository. After obtaining the user's specified output target, the algorithm is deployed to the container orchestration platform for instantiation.

[0048] In one embodiment, according to preset scheduling rules, a corresponding algorithm image is generated, and the algorithm image is stored in an image library; user instructions are received, and according to the result saving type specified by the user, the algorithm processing results are saved to a file, a database and a REST interface is provided, or the model is hosted and a REST interface is provided; according to preset rules, the algorithm image is deployed to a container orchestration system, and an online operation is performed to provide algorithm services.

[0049] The scheduler can generate algorithm images according to the scheduling type and submit them to the image library to prepare for subsequent deployment. Users can choose to save the results to files, databases and provide REST interfaces, or host the model and provide REST interfaces. Finally, the system will deploy the algorithm to the container orchestration system for online operation according to the rules.

[0050] The scheduler of the present invention first generates an algorithm image according to the scheduling type. The scheduling type is selected by the user and may include but is not limited to batch processing scheduling, real-time stream processing scheduling, machine learning training scheduling, etc. Algorithm mirroring refers to a standardized and portable software unit that packages an algorithm and its operating environment. This mirroring method can ensure the consistent operation of the algorithm in different environments.

[0051] The scheduler supports multiple result saving types, including saving to files, saving to databases and providing REST interfaces, model hosting and providing REST interfaces, etc. These saving types do not correspond one-to-one to specific scheduling types, but are freely selected by users according to their needs. For example, for the scheduling type of machine learning training, users can choose to host the trained model and provide a REST interface for subsequent online prediction. The present invention currently supports the use of the TensorFlow framework for model hosting. TensorFlow is an end-to-end open source machine learning platform with a comprehensive and flexible ecosystem.

[0052] After generating the algorithm image, the scheduler will submit the image to the image library for storage. The image library is used to centrally manage and store the software images of various algorithms to be scheduled, facilitating subsequent deployment and use.

[0053] Finally, the scheduler will deploy the algorithm to the container orchestration system for online operation according to predefined rules. The present invention uses Kubernetes as a container orchestration system, which is mainly used for the scheduling and operation of algorithms. Algorithm orchestration refers to the process of deploying algorithms to the production environment using scheduling rules, and can be put into use directly after deployment.

[0054] Through the above steps, the present invention realizes the automation of the whole process from algorithm development to deployment, greatly improving the delivery efficiency and reliability of the algorithm. At the same time, based on the design of container and microservice architecture, the system has good scalability and flexibility, and can adapt to various complex algorithm scheduling scenarios.

[0055] At step S204, after the algorithm operation is completed, the result is stored as a file, database record or machine learning model, and based on the usage scenario requirements, corresponding data acquisition, RESTful API or model deployment services are provided.

[0056] In one implementation scenario, after completing the algorithm calculation, the algorithm result is stored and processed according to the algorithm result storage location in the user's configuration information; if the algorithm result storage is configured as file export, the algorithm result file is stored in the object storage module, and a data list for download is generated; if the algorithm result storage is configured as database storage, the algorithm result is stored in the database; if the algorithm result is a machine learning model, the trained machine learning model is stored in the object storage module; for the algorithm result data stored in the database, a REST interface for query is generated; for the machine learning model stored in the object storage module, a model hosting service is generated and deployed online, and a REST interface service for model prediction is provided.

[0057] Specifically, the calculation results from the algorithm calculation module are received. These algorithms may be executed by the calculation unit inside the system or provided by an external system. The input data required by the algorithm can come from the system database, user uploads or external data sources. The system decides how to process and store these calculation results based on the pre-set result configuration. The result configuration can be pre-selected by the user through the system interface, or it can be automatically determined by the system based on the algorithm type.

[0058] The system supports three main result processing methods:

[0059] 1. File storage mode: When the algorithm result is configured as a file output, the system will store the result file in the object storage system. At the same time, the system will generate a data list to record all stored file information, including but not limited to file name, storage path, creation time, etc. Users can view this list through the system interface and select the required file to download. This mode is suitable for scenarios with large data sets or offline analysis.

[0060] 2. Database storage mode: If the results are configured to be stored in the database, the system will write the calculation results to the specified database table. At the same time, the system will automatically generate the corresponding REST API interface. These interfaces allow users to query the results stored in the database through HTTP requests. Users can design query parameters such as time range, data type, etc. as needed to achieve flexible data retrieval. This mode is suitable for scenarios that require frequent queries or real-time analysis.

[0061] 3. Machine learning model hosting mode: For machine learning algorithms, when training is completed, the system will store the trained model files in the object storage system. Subsequently, the system will automatically start the model hosting service. This service will generate the corresponding prediction service based on the stored model files and deploy it as an online REST API. Users can send prediction requests through this API to obtain the prediction results of the model. This mode greatly simplifies the deployment and application process of machine learning models, allowing users to easily use trained models without having to worry about the underlying technical details.

[0062] Through these three modes, the system can meet the storage and application requirements of different types of algorithm results, providing users with a comprehensive and flexible data processing solution. Whether it is large-scale data analysis, real-time data query, or machine learning model application, users can easily achieve it through a unified interface. This not only improves data utilization efficiency, but also greatly reduces the technical threshold for algorithm application.

[0063] Figure 3 A flowchart of another method for an online R&D platform for energy storage battery safety algorithms provided in an embodiment of the present application. The present invention provides an algorithm development and management platform that integrates functions such as data source selection, algorithm development, algorithm review, algorithm deployment, and result output, which can significantly improve the work efficiency of algorithm engineers and form a safe algorithm ecosystem.

[0064] In terms of data source selection, this platform can access various types of data sources. These data sources include but are not limited to real-time data from energy management systems, database system data (which contains output data from other algorithms), and file data sources. This diverse data source access capability provides a rich data foundation for algorithm development, which is conducive to improving the accuracy and applicability of the algorithm.

[0065] In the algorithm development system, this platform provides online editing and testing functions. Algorithm engineers can write and debug algorithms directly on the platform and view intermediate results in real time, which greatly facilitates the algorithm verification process. In addition, by selecting the output data of other algorithms in the database system, this platform supports multi-algorithm aggregation, thereby realizing complex algorithm chains and enhancing the functionality and flexibility of the algorithm.

[0066] In order to ensure the security and reliability of the algorithm, this platform has introduced a manual review mechanism. The algorithm must be manually reviewed after uploading, and only the reviewed algorithm can be released and run. This mechanism effectively prevents the exposure of sensitive data and also provides security for external users to use the platform for development.

[0067] In terms of algorithm deployment, the platform implements flexible deployment management through the scheduler. The platform supports the development of various types of algorithms and can output algorithm results to files, databases, or online hosting models. In addition, the platform also provides a REST interface to facilitate other business systems to call algorithm services.

[0068] Through the above design, this algorithm platform greatly reduces the workload of algorithm engineers. Engineers can focus on the development of core business algorithms, and leave basic work such as data processing and algorithm deployment to the platform. At the same time, the platform's algorithms can be selectively exposed to the outside world, which not only protects core algorithm assets, but also lays the foundation for the formation of a secure algorithm ecosystem.

[0069] In summary, through the solution of this application, the real-time data of the energy management system is accessed, and the data uploaded by the real-time energy management can be selected when the data source is selected. The algorithm can be edited online, and the algorithm development engineer can complete the verification of the algorithm directly on the page. Multiple algorithms can be aggregated and used to form an algorithm chain to form more valuable result data. The algorithm platform can be open to the outside world and the algorithm audit can be performed through the algorithm audit system, so there is no need to worry about the exposure of business data, allowing more algorithm enthusiasts and companies to develop algorithms based on this platform and enrich energy storage safety algorithms. The algorithm platform can be used for basic algorithm development, battery mechanism algorithm development, and machine learning algorithm development. Unified management through the platform. The results of the algorithm include a variety of storage methods, which can be used for file storage, database storage and provide REST interface, machine learning model storage, and can be hosted online to provide REST interface for data prediction.

[0070] Furthermore, an embodiment of the present application also provides a device for an online R&D platform of energy storage battery safety algorithms, comprising: a processor, a memory, and a system bus; the processor and the memory are connected via the system bus; the memory is used to store one or more programs, and the one or more programs include instructions, which, when executed by the processor, enable the processor to execute any of the above methods.

[0071] Furthermore, an embodiment of the present application also provides a computer program product, which, when executed on a terminal device, enables the terminal device to execute any one of the above-mentioned processing methods.

[0072] It can be known from the description of the above implementation mode that those skilled in the art can clearly understand that all or part of the steps in the above-mentioned embodiment method can be implemented by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product can be stored in a storage medium such as ROM / RAM, a disk, an optical disk, etc., including several instructions for enabling a computer device (which can be a personal computer, a server, or a network communication device such as a media gateway, etc.) to execute the methods described in the various embodiments of the present application or certain parts of the embodiments.

[0073] It should be noted that the various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments, and the same or similar parts between the various embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description.

[0074] It should also be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0075] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A system for an online R&D platform for energy storage battery safety algorithms, characterized in that: include: Data source selection subsystem, algorithm development subsystem, scheduler subsystem and managed application subsystem; among them, The data source selection subsystem is used to configure real-time, historical or file data sources, build temporary data point tuple arrays for algorithm processing, and automatically release temporary data after processing is completed; The algorithm development subsystem is used to test the algorithms written by users on the platform, store them in the object storage system after review, and obtain the user's scheduling strategy to transmit the generated configuration information to the scheduler subsystem for task scheduling; The scheduler subsystem is used to build an algorithm image according to the scheduling rules in the configuration information and store it in the image warehouse. After obtaining the user's specified output target, the algorithm is deployed to the container orchestration platform for instantiation; The hosted application subsystem is used to store the results as files, database records or machine learning models after the algorithm operation is completed, and provide corresponding data acquisition, RESTful API or model deployment services based on the usage scenario requirements.

2. The system according to claim 1, characterized in that in, Configure real-time, historical, or file data sources, build temporary data point tuple arrays for algorithm processing, and automatically release temporary data after processing is complete, including: Select a data source type, which includes selecting real-time data points online, retrieving historical data points, and custom uploading data files; generating a temporary data point tuple array according to the selected data source type, the array containing time, voltage and / or current information to construct multi-dimensional data; Display some sample data to the user, and automatically clear the corresponding related data in the cache and / or file after the data is used.

3. The system according to claim 1, characterized in that in, The algorithm written by the user on the platform is tested and stored in the object storage system after review. The user's scheduling strategy is obtained to transmit the generated configuration information to the scheduler subsystem for task scheduling, including: Providing algorithm templates for users to download, or providing online editors for users to develop algorithms, in order to obtain user-submitted algorithms; Test and debug the algorithm to receive the algorithm that has passed the debugging; Review the received algorithm codes and store the approved algorithm codes in the target object storage system; Receive the algorithm scheduling rules selected by the user, and generate a task containing the image address of the algorithm to be scheduled according to the scheduling rules; Configuration information is generated, wherein the configuration information includes a data cache address, a scheduling rule, an algorithm image to be scheduled, and a result saving type.

4. The system according to claim 1, characterized in that in, The algorithm image is built according to the scheduling rules in the configuration information and stored in the image repository. After obtaining the user's specified output target, the algorithm is deployed to the container orchestration platform for instantiation, including: Generate a corresponding algorithm image according to the preset scheduling rules, and store the algorithm image in the image library; Receive user instructions, save the algorithm processing results to files or databases and provide REST interfaces according to the result saving type specified by the user, or host the model and provide REST interfaces; According to preset rules, the algorithm image is deployed to the container orchestration system, and the online operation is performed to provide algorithm services.

5. The system according to claim 1, characterized in that in, After the algorithm operation is completed, the results are stored as files, database records or machine learning models, and corresponding data acquisition, RESTful API or model deployment services are provided based on the usage scenario requirements, including: After the algorithm calculation is completed, the algorithm result is stored according to the algorithm result storage location in the user's configuration information; If the algorithm result storage is configured as file export, the algorithm result file is stored in the object storage module, and a data list for download is generated; if the algorithm result storage is configured as database storage, the algorithm result is stored in the database; if the algorithm result is a machine learning model, the trained machine learning model is stored in the object storage module; For the algorithm result data stored in the database, a REST interface for query is generated; for the machine learning model stored in the object storage module, a model hosting service is generated and deployed online, and a REST interface service for model prediction is provided.

6. The system according to claim 2, characterized in that in, Before showing some sample data to the user, it also includes: Randomly generate encryption parameters including key, offset, modulus and mask values; Encrypting each tuple in the temporary data point tuple array in a time sequence order based on the encryption parameters and the preset encryption function; Select sample data at fixed intervals from the encrypted tuple array to construct a sample array for display.

7. The system according to claim 6, characterized in that The preset encryption function includes: Where t represents the time point of data collection, X represents the vector containing all battery parameter values, x1, x2, ..., x m represents the value of each battery parameter, k represents the encryption key, r1, r2, ..., r m represents a random mask value, represents the bitwise XOR operator, b represents the offset, m represents the modulus, and f(t, X) represents the encryption function.

8. A method for an online R&D platform for energy storage battery safety algorithms, characterized in that: include: Configure real-time, historical or file data sources, build temporary data point tuple arrays for algorithm processing, and automatically release temporary data after processing is completed; Test the algorithms written by users on the platform, store them in the object storage system after review, and obtain the user's scheduling strategy to transmit the generated configuration information to the scheduler subsystem for task scheduling; The algorithm image is built according to the scheduling rules in the configuration information and stored in the image warehouse. After obtaining the user's specified output target, the algorithm is deployed to the container orchestration platform for instantiation; After the algorithm operation is completed, the results are stored as files, database records or machine learning models, and corresponding data acquisition, RESTful API or model deployment services are provided based on the usage scenario requirements.

9. A device for an online R&D platform of energy storage battery safety algorithm, characterized in that: include: A processor, a memory, and a system bus; wherein the processor and the memory are connected via the system bus; The memory is used to store one or more programs, wherein the one or more programs include instructions, and when the instructions are executed by the processor, the processor is caused to perform the method of claim 8.

Images