Data management method and system of industrial energy storage system

Abstract

The invention relates to the technical field of data management, in particular to a data management method and system for an industrial energy storage system. The method comprises the following steps: carrying out real-time energy storage data acquisition on an industrial energy storage unit, and generating corrected energy storage monitoring data; carrying out distributed redundancy transmission on the corrected energy storage monitoring data, and carrying out uploading data difference analysis to obtain monitoring uploading difference data; performing difference conflict processing on the redundant monitoring and uploading data through the monitoring and uploading difference data to generate monitoring fragment feature data; performing dynamic cache synchronization processing according to the monitoring fragment feature data to obtain monitoring cache state data; and performing management state comprehensive analysis on the monitored fragment feature data by monitoring the cache state data to generate a comprehensive data management index. According to the method, efficient, reliable and safe uploading management optimization of mass industrial energy storage monitoring data is realized through distributed intelligent caching, and the problems of inconsistent data uploading and even data loss are effectively solved.

Description

Technical Field

[0001] The present invention relates to the technical field of data management, and in particular to a data management method and system for an industrial energy storage system. Background Art

[0002] With the rapid development of renewable energy and the advancement of industrial intelligence, industrial energy storage systems play an increasingly important role in power systems, especially in the integration of renewable energy and grid stabilization, which can not only balance power supply and demand, but also improve the reliability of power systems. With the popularization of distributed energy resources (such as solar panels and wind turbines), energy storage systems are usually composed of multiple energy storage nodes distributed in different geographical locations. These nodes can be batteries, supercapacitors or other energy storage devices. Each energy storage node needs to monitor and manage a large amount of data in real time. Distributed energy storage systems need to have energy storage nodes distributed in multiple locations, and data collection and processing need to have high real-time characteristics. However, traditional data management methods for industrial energy storage systems usually rely on centralized storage management, but the collected data is distributed on multiple nodes, and in the case of network delays, node failures or partitions, this centralized management is prone to data inconsistency, storage delays, and even data loss. Summary of the invention

[0003] Based on this, the present invention provides a data management method and system for an industrial energy storage system to solve at least one of the above technical problems.

[0004] To achieve the above object, a data management method for an industrial energy storage system comprises the following steps:

[0005] Step S1: Perform edge computing topology processing on the industrial energy storage unit to generate intelligent perception network topology data; perform real-time energy storage data collection based on the intelligent perception network topology data to obtain time-series energy storage monitoring data; perform abnormal data correction based on the time-series energy storage monitoring data to generate corrected energy storage monitoring data;

[0006] Step S2: Optimizing distributed storage nodes according to the intelligent sensing network topology data to obtain a distributed data storage network; using the distributed data storage network to perform distributed redundant transmission on the corrected energy storage monitoring data to obtain redundant monitoring upload data;

[0007] Step S3: Perform upload voucher processing on the redundant monitoring upload data to generate upload voucher data; perform upload data difference analysis based on the upload voucher data to obtain monitoring upload difference data; perform difference conflict processing on the redundant monitoring upload data through the monitoring upload difference data to obtain energy storage unit monitoring data;

[0008] Step S4: Perform dynamic data slicing processing according to the energy storage unit monitoring data to generate monitoring slicing feature data; perform dynamic cache synchronization processing according to the monitoring slicing feature data to generate cache management log data;

[0009] Step S5: performing cache status evaluation according to the cache management log data to obtain monitoring cache status data;

[0010] Step S6: Perform a comprehensive management status analysis on the monitored slice feature data by monitoring the cache status data to generate a comprehensive data management index.

[0011] The present invention performs edge computing topology processing on industrial energy storage units, establishes an efficient network structure, enables each energy storage unit to achieve rapid data interaction and processing, and improves the real-time response capability and intelligence level of the system. Real-time energy storage data acquisition is performed according to the intelligent perception network topology data to ensure that the system can continuously monitor the operating status of each energy storage unit and obtain timely and accurate monitoring data. Abnormal data correction is performed according to the time-series energy storage monitoring data, which can eliminate noise and errors in the data acquisition process and ensure that subsequent analysis is based on a reliable data source. Distributed storage node optimization based on intelligent perception network topology data effectively improves the reliability and access efficiency of data storage and forms a distributed data storage network. The distributed redundant transmission strategy using this network further enhances the security of data. Even if some nodes fail, the integrity of data can be ensured by redundant monitoring and uploading data. By uploading credential processing and uploading data difference analysis, not only the security of data upload is enhanced, but also the differences or conflicts generated during the data transmission process are effectively identified and resolved, ensuring the consistency and accuracy of the final energy storage unit monitoring data. Large-scale monitoring data is divided into smaller, easy-to-manage and process segments, which improves data access efficiency and flexibility. The dynamic cache synchronization process ensures the high efficiency of data access, and the cache management log data formed helps to monitor and optimize the use of data cache. Cache status evaluation based on cache management log data can provide the system with cache health status and performance indicators, and help identify potential failure risks in a timely manner. By comprehensively considering the monitoring cache status data and conducting a comprehensive analysis of the management status of the monitoring shard feature data, the generated comprehensive data management index is a highly generalized indicator that reflects the efficiency and health status of the entire data management process. This index can not only serve as a basis for data management optimization and adjustment, but also help decision makers make more accurate management and scheduling decisions. Therefore, a data management method for an industrial energy storage system of the present invention collects industrial energy storage monitoring data from energy storage units in real time through edge computing, optimizes distributed storage nodes and implements redundant transmission, ensures the consistency and integrity of data in the event of network failure or node failure, thereby reducing the risk of data loss. Through dynamic data sharding and intelligent cache management, resource allocation is optimized according to data access mode, upload storage delay is further reduced, and efficient and intelligent data upload management is achieved.

[0012] Preferably, step S1 comprises the following steps:

[0013] Step S11: deploy intelligent monitoring sensors on industrial energy storage units to obtain monitoring node distribution data;

[0014] Step S12: Perform wireless network configuration according to the monitoring node distribution data, and perform edge computing topology processing to generate intelligent perception network topology data, wherein the intelligent perception network topology data includes energy storage monitoring nodes and distributed computing nodes;

[0015] Step S13: Setting data collection parameters for energy storage monitoring nodes based on intelligent sensing network topology data, and collecting real-time industrial energy storage data streams to obtain initial energy storage unit monitoring data;

[0016] Step S14: using the distributed computing nodes to timestamp the initial energy storage unit monitoring data to obtain time series energy storage monitoring data;

[0017] Step S15: performing data redundancy filtering according to the time series energy storage monitoring data to obtain simplified energy storage monitoring data;

[0018] Step S16: performing outlier detection on the simplified energy storage monitoring data to generate abnormal data points; performing data resampling according to the abnormal data points to obtain resampled monitoring data;

[0019] Step S17: Based on the abnormal data points, the simplified energy storage monitoring data is corrected by resampling the monitoring data to generate corrected energy storage monitoring data.

[0020] The present invention ensures that the status of each energy storage unit can be fully covered and monitored in real time by deploying intelligent monitoring sensors for industrial energy storage units, thereby enhancing the monitoring capability of the system and the breadth of data acquisition. By configuring the wireless network according to the distribution data of the monitoring nodes, efficient data transmission can be achieved, the network structure is optimized, and the delay of data transmission is reduced. By setting the data acquisition parameters under the guidance of the intelligent perception network topology data, the data acquisition process is made more accurate and efficient, and the collection of real-time industrial energy storage data streams ensures the timeliness of the data. The initial energy storage unit monitoring data is timestamped using distributed computing nodes to ensure the time series characteristics of the data. Data redundancy filtering is performed based on the time series energy storage monitoring data, and repeated or invalid data points can be removed, which effectively reduces the amount of data and improves the efficiency of data processing. The abnormal points in the monitoring data are identified by outlier detection, and the subsequent data resampling operation supplements or corrects these abnormal data. The generated resampled monitoring data is more accurate and continuous, and the data quality is improved. The simplified energy storage monitoring data is corrected using abnormal data points and resampled monitoring data, ensuring the accuracy and completeness of the corrected energy storage monitoring data. It is particularly critical to eliminate erroneous data caused by equipment failure or environmental interference, thereby improving the robustness of the entire data link.

[0021] Preferably, step S2 comprises the following steps:

[0022] Step S21: Performing geographic spatial analysis based on the intelligent sensing network topology data to obtain spatial location distribution data;

[0023] Step S22: evaluating the wireless signal strength of the industrial energy storage unit through the spatial location distribution data, optimizing the communication link of the intelligent sensing network topology data, and generating a sensing node communication network;

[0024] Step S23: Connecting distributed storage nodes according to the sensing node communication network to obtain a distributed data storage network;

[0025] Step S24: using the distributed data storage network to select an upload node for the corrected energy storage monitoring data, and performing parallel task decomposition to obtain multiple upload instruction data;

[0026] Step S25: redundantly copy the corrected energy storage monitoring data to obtain a redundant monitoring data packet;

[0027] Step S26: performing distributed redundant transmission on redundant monitoring data packets through multiple upload instruction data to obtain redundant monitoring upload data.

[0028] The present invention can clearly display the physical location layout of each monitoring node through the geospatial analysis of the intelligent sensing network topology data. The wireless signal strength of the industrial energy storage unit is evaluated by using the spatial location distribution data, and the communication link optimization of the intelligent sensing network topology data can effectively improve the stability and efficiency of data transmission. Distributed storage nodes are connected based on the optimized sensing node communication network to form a distributed data storage network. This architecture can disperse storage pressure, improve data access speed and security, and enhance the scalability and fault tolerance of the system. The upload node selection and parallel task decomposition performed by the distributed data storage network not only optimize the data upload path, but also realize the efficient parallel processing of data upload, which greatly speeds up the data processing speed. By redundantly replicating the corrected energy storage monitoring data, the reliability and durability of the data are significantly enhanced. Even in the case of partial data transmission failure, the complete recovery of the data can be ensured, which improves the robustness of the system. The redundant monitoring data packets are effectively distributed to multiple storage nodes, which not only ensures the real-time backup of the data, but also increases the success rate of data transmission through multi-path transmission.

[0029] Preferably, step S3 comprises the following steps:

[0030] Step S31: performing monitoring characteristic analysis on the redundant monitoring upload data to generate key energy storage monitoring characteristic data; performing upload voucher processing according to the key energy storage monitoring characteristic data to generate upload voucher data;

[0031] Step S32: performing credential identification processing on the redundant monitoring upload data by uploading the credential data to obtain the identified monitoring upload data;

[0032] Step S33: Perform multi-node storage network analysis on the redundant uploaded data through the distributed data storage network to generate a multi-node storage network;

[0033] Step S34: using a multi-node storage network to construct a consensus network for the identification monitoring upload data, generating storage consensus network data; performing upload data difference analysis based on the storage consensus network data, obtaining monitoring upload difference data;

[0034] Step S35: Perform node performance evaluation on the storage consensus network data to generate consensus node performance data; perform confidence node recommendation on the storage consensus network data based on the consensus node performance data to obtain confidence node data;

[0035] Step S36: Based on the consensus node performance data, the confidence node data is used to perform a confidence difference weighted judgment on the monitored uploaded difference data, and difference conflict processing is performed to generate uploaded difference conflict processing data;

[0036] Step S37: Based on the confidence node data, the identification monitoring uploaded data is updated for data consistency by uploading the difference conflict processing data to obtain the energy storage unit monitoring data.

[0037] The present invention performs monitoring feature analysis on redundant monitoring uploaded data, can extract the main features and trends of energy storage equipment operation, and performs upload credential processing according to the generated key energy storage monitoring feature data. By generating upload credential, the integrity and security of the data upload process are ensured, the source and verification basis of the data are provided, and the credibility of data management is enhanced. The redundant monitoring uploaded data is processed with credential identification by uploading credential data, which ensures that all uploaded data are accompanied by corresponding credential identification, which is helpful for subsequent tracking and management. The multi-node storage network analysis performed by the distributed data storage network optimizes the layout of data storage, and the generated multi-node storage network enhances the distributed storage capacity and access efficiency of data. By constructing a consensus network, the consistency and integrity of the data are ensured, and the risk of information loss caused by node failure or data conflict is reduced. The node performance evaluation is performed on the storage consensus network data to generate consensus node performance data. The evaluation process can identify the performance advantages and disadvantages of each storage node, ensuring that the system can still maintain good performance under high load conditions. The confidence node recommendation is performed on the storage consensus network data through the consensus node performance data, ensuring that reliable nodes for data storage and processing are used preferentially. The data after weighted judgment is processed for difference conflicts to ensure that conflicts between data can be effectively resolved in the case of multiple data sources. Based on the confidence node data, the identification monitoring uploaded data is updated for data consistency by uploading the difference conflict processing data. This update process ensures that all monitoring data are strictly verified and processed, and the energy storage unit monitoring data finally generated has high credibility and validity.

[0038] Preferably, step S31 includes the following steps:

[0039] Step S311: Identify the data type of the redundant monitoring uploaded data to obtain monitoring type identification data;

[0040] Step S312: performing feature statistics on the redundant monitoring uploaded data through the monitoring type identification data based on the preset monitoring type statistical rules to generate energy storage feature statistical data;

[0041] Step S313: performing feature correlation analysis based on energy storage feature statistical data, and extracting key monitoring features to generate key energy storage monitoring feature data;

[0042] Step S314: using a preset hash algorithm to gradually read the key energy storage monitoring characteristic data, and performing fixed hash value calculation to generate initial hash value data;

[0043] Step S315: rolling update processing is performed on the initial hash value data to obtain key monitoring hash data; metadata is added according to the key monitoring hash data to obtain monitoring credential metadata;

[0044] Step S316: Perform credential serialization processing on the monitoring credential metadata through a preset upload credential template to generate upload credential data.

[0045] The present invention can clearly distinguish different types of data content by identifying the data type of redundant monitoring uploaded data, so that different types of data can be analyzed and processed in a targeted manner. The redundant monitoring uploaded data is subjected to feature statistics through monitoring type identification data, and the feature performance under different monitoring types is quantified, such as the mean, standard deviation, extreme value, rate of change and other indicators of the feature. Feature correlation analysis is performed based on energy storage feature statistical data, and key monitoring feature extraction is performed. This analysis can reveal the relationship between different features and identify the key features that have the greatest impact on system performance. The initial hash value data generated by the step-by-step reading and fixed hash value calculation of the key energy storage monitoring feature data using a preset hash algorithm provides a unique digital fingerprint for the data, enhancing the uniqueness and non-tamperability of the data. The initial hash value data is subjected to rolling update processing to ensure that the hash value can reflect the latest monitoring data status, capture data changes in a timely manner, and enhance the dynamic adaptability of data management. The voucher serialization processing of the monitoring voucher metadata by the preset upload voucher template ensures the consistency and compatibility of data transmission between different systems, simplifies the complexity of data exchange, and improves the standardization and automation of data management.

[0046] Preferably, step S34 includes the following steps:

[0047] Step S341: using a multi-node storage network to perform node timing credential exchange on the identification monitoring uploaded data to obtain inter-node credential exchange data;

[0048] Step S342: performing hash collision calculation based on the credential exchange data between nodes to obtain preliminary credential similarity data;

[0049] Step S343: Based on a preset similarity threshold, the storage node threshold of the multi-node storage network is screened through preliminary credential similarity data to obtain consensus node list data;

[0050] Step S344: construct and store consensus network data according to the consensus node list data;

[0051] Step S345: Using the storage consensus network data to store node redundant data for the identification monitoring upload data, to obtain consensus node redundant mapping data;

[0052] Step S346: performing data comparison between nodes based on the consensus node redundant mapping data to generate preliminary comparison difference data;

[0053] Step S347: Perform metadata difference analysis based on the preliminary comparison difference data, and perform serialization processing to generate monitoring upload difference data.

[0054] The present invention utilizes a multi-node storage network to perform node timing credential exchange on identification monitoring upload data, thereby ensuring the secure synchronization and verification of data between various nodes. Hash collision calculation is performed based on the credential exchange data between nodes to obtain preliminary credential similarity data. This calculation process can identify the similarity between credentials of different nodes and help discover the existence of data redundancy or inconsistency. By setting a threshold, it is ensured that only nodes that meet the similarity requirements are included in the consensus node list. The storage consensus network data constructed based on the consensus node list data ensures the correct replication and synchronization of data between multiple nodes, thereby enhancing the data reliability and fault tolerance of the system. By storing redundant data in multiple consensus nodes, the security and reliability of data are improved. The consensus node redundant mapping data ensures that when a node fails or an error occurs, the system can still quickly restore the data, thereby reducing the risk of data loss. Data comparison between nodes based on the consensus node redundant mapping data can identify data differences between storage nodes, help discover potential data inconsistency problems, and improve the accuracy of data management. Metadata difference analysis is performed based on the preliminary comparison difference data, and serialization processing is performed to obtain the monitoring upload difference data, thereby ensuring that all differences are effectively identified and recorded.

[0055] Preferably, step S4 comprises the following steps:

[0056] Step S41: Perform dynamic data slicing processing on the monitoring data of the energy storage unit to generate monitoring slicing feature data;

[0057] Step S42: Calculate the monitoring update rate according to the monitoring slice feature data to generate slice update frequency data;

[0058] Step S43: Evaluate the cache demand of the energy storage monitoring slice data by slice update frequency data to generate slice cache demand data;

[0059] Step S44: dividing the data update status of the shard cache demand data based on the preset update rate threshold, when the shard cache demand data is higher than or equal to the preset update rate threshold, marking the monitoring shard feature data as high-frequency monitoring shard data; when the shard cache demand data is lower than the preset update rate threshold, marking the monitoring shard feature data as low-frequency monitoring shard data;

[0060] Step S45: Calculate the average update interval of the high-frequency monitoring slice data to generate average update interval data; perform real-time update window cache processing according to the average update interval data to obtain a high-frequency slice cache strategy;

[0061] Step S46: estimating the cache capacity requirement of the low-frequency monitoring shard data to generate cache capacity requirement data; obtaining available memory resource data; performing secure caching processing on the cache capacity requirement data through the available memory resource data to obtain low-frequency shard cache data;

[0062] Step S47: Perform dynamic cache strategy processing according to the high-frequency shard cache strategy and the low-frequency shard cache data to generate dynamic cache strategy data;

[0063] Step S48: Perform cache synchronization management processing on the monitoring slice feature data through dynamic cache strategy data to generate cache management log data.

[0064] The present invention performs dynamic data slicing processing on the monitoring data of the energy storage unit, divides the large-scale monitoring data into smaller, easy-to-manage and process data segments, and improves the data access efficiency. This information is organized into a structured format, for example, the ID, data range, feature statistics, etc. of each slicing, and the monitoring update rate is calculated according to the monitoring slicing feature data, so that the update frequency of each data segment can be identified, thereby reflecting the importance and dynamic changes of the data. The cache demand of the energy storage monitoring slicing data is evaluated through the slicing update frequency data, and the cache resources required for each data segment are determined to ensure that the data updated with high frequency can be cached first. Based on the data update state division of the slicing cache demand data based on the preset update rate threshold, the monitoring slicing feature data is divided into two categories: high frequency and low frequency. This classification strategy is conducive to differentiated management of data with different update characteristics and improves the utilization efficiency of cache resources. The average update interval calculation and real-time update window cache processing of high-frequency monitoring slicing data can ensure rapid response and timely update of high-dynamic data, reduce data delay, and improve system response speed. The cache capacity demand estimation for low-frequency monitoring shard data, combined with the secure cache processing of available memory resource data, not only ensures the effective caching of relatively static data, but also avoids excessive occupation of memory resources, and realizes the reasonable allocation of resources. Dynamic cache strategy processing is performed based on the high-frequency shard cache strategy and low-frequency shard cache data. This strategy comprehensively considers the characteristics of high-frequency and low-frequency data to ensure that the system can flexibly adjust the allocation of cache resources and optimize data access efficiency. The monitoring shard feature data is cached and synchronized through dynamic cache strategy data to ensure that all collected industrial energy storage data can be synchronized in a timely manner, and data cache management is realized to avoid errors caused by data asynchrony.

[0065] Preferably, step S5 comprises the following steps:

[0066] Step S51: Perform data update monitoring according to the cache management log data and generate a data update status report;

[0067] Step S52: performing update notification broadcast processing on the distributed data storage network based on the data update status report, and extracting the cache status summary to generate cache status summary data;

[0068] Step S53: performing cache consistency evaluation according to the cache status summary data to generate cache consistency evaluation data; performing inconsistent node identification on the cache consistency evaluation data to obtain inconsistent node list data;

[0069] Step S54: repair the node cache according to the inconsistent node list data, and re-evaluate the cache status to obtain monitoring cache status data.

[0070] The present invention performs data update monitoring based on cache management log data, ensures that the update status of data can be tracked in real time, and provides transparency of the data update process. The generated data update status report provides important information for the system administrator, helps to timely identify the success or failure of the data update, and ensures that the data is always kept up to date. Based on the data update status report, the distributed data storage network is broadcasted for update notification, and the cache status summary is extracted to generate cache status summary data. The broadcast processing can quickly propagate the data update information to all relevant nodes, ensuring that each part of the system can obtain the latest data update status in time. The cache consistency evaluation is performed based on the cache status summary data, and the consistency status of the cache data in the system can be identified, ensuring that the data between different nodes remain synchronized, and helping to determine whether further data repair or update operations are required. By repairing the cache data of these nodes, the data consistency of the system can be effectively restored, avoiding erroneous decisions or operations caused by data inconsistency. The cache status re-evaluation performed subsequently and the monitoring cache status data obtained not only verify the effect of the repair operation, but also provide the latest information for continuous monitoring and optimization of cache management strategies.

[0071] Preferably, step S6 comprises the following steps:

[0072] Step S61: Obtain historical energy storage data access records; perform access pattern recognition on monitoring slice feature data through historical energy storage data access records to generate data access patterns;

[0073] Step S62: Estimating the access probability of the monitored shard feature data based on the data access pattern to generate shard access probability data; constructing a predictive shard data index based on the shard access probability data;

[0074] Step S63: performing a shard data query simulation according to the predictive shard data index, and locating the target data shard to obtain shard query feature data;

[0075] Step S64: Perform dynamic access path evaluation according to the shard query feature data to generate access path evaluation data; perform data prefetch efficiency evaluation according to the shard query feature data to generate data prefetch evaluation data;

[0076] Step S65: performing cache effect evaluation on the shard query feature data by monitoring the cache status data to generate cache effect evaluation data;

[0077] Step S66: Perform a comprehensive analysis of the management status based on the access path evaluation data, the data pre-fetch evaluation data, and the cache effect evaluation data to generate a comprehensive data management index.

[0078] The present invention identifies the access pattern of the monitoring shard feature data through the historical energy storage data access record, which can reveal the user's access habits to data in different time periods and help identify the data fragments with high frequency access. Based on the data access pattern, the access probability of the monitoring shard feature data is estimated, and the possibility of each data fragment being accessed is quantified. The predictive shard data index constructed on this basis can predict the data access demand in advance, accelerate the data retrieval process, and reduce the query delay. According to the predictive shard data index, the shard data query simulation is performed, which provides a preview for the actual data access, ensures that the system can identify the data to be accessed in advance, and improves the timeliness and accuracy of data access. The dynamic access path evaluation performed by the shard query feature data helps to identify the optimal data access path and reduce the data transmission time. At the same time, the data pre-fetch evaluation data generated by the data pre-fetch efficiency evaluation can guide the system to load the accessed data in advance, further improving the efficiency of data processing and user experience. The cache effect evaluation of the shard query feature data using the monitoring cache status data evaluates the effectiveness of the current cache strategy, provides feedback information for adjusting the cache configuration and optimizing the cache hit rate, and ensures the efficient use of cache resources. Comprehensive analysis of management status based on access path evaluation data, data prefetch evaluation data, and cache effect evaluation data reflects the overall health and efficiency of data management.

[0079] Preferably, the present invention further provides a data management system for an industrial energy storage system, which executes the data management method for an industrial energy storage system as described above, and the data management system for an industrial energy storage system comprises:

[0080] The real-time data acquisition module is used to perform edge computing topology processing on the industrial energy storage unit to generate intelligent perception network topology data; perform real-time energy storage data acquisition based on the intelligent perception network topology data to obtain time-series energy storage monitoring data; perform abnormal data correction based on the time-series energy storage monitoring data to generate corrected energy storage monitoring data;

[0081] A distributed redundant transmission module is used to optimize distributed storage nodes according to the intelligent perception network topology data to obtain a distributed data storage network; and to perform distributed redundant transmission of the corrected energy storage monitoring data using the distributed data storage network to obtain redundant monitoring upload data;

[0082] The data consistency maintenance module is used to process the upload voucher of the redundant monitoring upload data to generate the upload voucher data; perform upload data difference analysis according to the upload voucher data to obtain the monitoring upload difference data; perform difference conflict processing on the redundant monitoring upload data through the monitoring upload difference data to obtain the energy storage unit monitoring data;

[0083] A dynamic data cache module is used to perform dynamic data sharding processing according to the monitoring data of the energy storage unit to generate monitoring shard feature data; perform dynamic cache synchronization processing according to the monitoring shard feature data to generate cache management log data;

[0084] The cache status analysis module is used to evaluate the cache status according to the cache management log data to obtain the monitoring cache status data;

[0085] The comprehensive data management evaluation module is used to conduct a comprehensive analysis of the management status of the monitoring shard feature data by monitoring the cache status data and generate a comprehensive data management index. BRIEF DESCRIPTION OF THE DRAWINGS

[0086] Figure 1 A schematic diagram of the steps of a data management method for an industrial energy storage system according to the present invention;

[0087] Figure 2 for Figure 1 Detailed implementation steps of step S2 in the flowchart;

[0088] Figure 3 for Figure 1 Detailed implementation steps of step S5;

[0089] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0090] The technical method of the present invention is described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0091] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0092] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0093] To achieve this, please refer to Figures 1 to 3 The present invention provides a data management method for an industrial energy storage system, comprising the following steps:

[0094] Step S1: Perform edge computing topology processing on the industrial energy storage unit to generate intelligent perception network topology data; perform real-time energy storage data collection based on the intelligent perception network topology data to obtain time-series energy storage monitoring data; perform abnormal data correction based on the time-series energy storage monitoring data to generate corrected energy storage monitoring data;

[0095] Step S2: Optimizing distributed storage nodes according to the intelligent sensing network topology data to obtain a distributed data storage network; using the distributed data storage network to perform distributed redundant transmission on the corrected energy storage monitoring data to obtain redundant monitoring upload data;

[0096] Step S3: Perform upload voucher processing on the redundant monitoring upload data to generate upload voucher data; perform upload data difference analysis based on the upload voucher data to obtain monitoring upload difference data; perform difference conflict processing on the redundant monitoring upload data through the monitoring upload difference data to obtain energy storage unit monitoring data;

[0097] Step S4: Perform dynamic data slicing processing according to the energy storage unit monitoring data to generate monitoring slicing feature data; perform dynamic cache synchronization processing according to the monitoring slicing feature data to generate cache management log data;

[0098] Step S5: performing cache status evaluation according to the cache management log data to obtain monitoring cache status data;

[0099] Step S6: Perform a comprehensive management status analysis on the monitored slice feature data by monitoring the cache status data to generate a comprehensive data management index.

[0100] In the embodiment of the present invention, reference Figure 1 The above is a schematic diagram of the steps of the data management method of the industrial energy storage system of the present invention. In this embodiment, the data management method of the industrial energy storage system includes the following steps:

[0101] Step S1: Perform edge computing topology processing on the industrial energy storage unit to generate intelligent perception network topology data; perform real-time energy storage data collection based on the intelligent perception network topology data to obtain time-series energy storage monitoring data; perform abnormal data correction based on the time-series energy storage monitoring data to generate corrected energy storage monitoring data;

[0102] In an embodiment of the present invention, an edge computing gateway supporting the Modbus TCP / IP protocol is used to connect to each industrial energy storage unit. For example, by analyzing the network connection information of the industrial energy storage unit, such as the IP address, port number, and network protocol, a network topology diagram containing all energy storage units and edge computing nodes is constructed. According to the generated topological data, a data acquisition program is deployed on the edge computing gateway. The program uses the MQTT protocol to collect real-time monitoring data of key parameters such as voltage, current, temperature, SOC, etc. of each energy storage unit in real time according to a preset sampling frequency (for example, voltage and current are collected once per second, and temperature is collected once per minute). The time series energy storage monitoring data is corrected. For example, an abnormal data elimination method based on the 3σ criterion is adopted, the sliding window size is set to 10 minutes, and the mean and standard deviation of the data in each window are calculated. Data that deviates from the mean value by more than 3 times the standard deviation is judged as abnormal data and eliminated, and the eliminated data points are filled using linear interpolation or data resampling, and finally the corrected energy storage monitoring data is generated.

[0103] Step S2: Optimizing distributed storage nodes according to the intelligent sensing network topology data to obtain a distributed data storage network; using the distributed data storage network to perform distributed redundant transmission on the corrected energy storage monitoring data to obtain redundant monitoring upload data;

[0104] In an embodiment of the present invention, the distributed storage system etcd is optimized for nodes using the distributed consistency algorithm Raft based on the intelligent perception network topology data, especially the geographic location information of the energy storage unit contained therein. For example, three servers that are closer to the energy storage unit are selected as the leader node and follower node of the etcd cluster, and load balancing is performed according to the network bandwidth and storage space of the node, and finally a highly available, low-latency distributed data storage network is obtained. The distributed message queue Kafka is used to perform distributed redundant transmission of the corrected energy storage monitoring data. For example, the corrected energy storage monitoring data is sent to three different partitions of Kafka respectively, and each partition corresponds to a different server node. Using Kafka's message replication mechanism, even if one of the nodes fails, the data will not be lost, and redundant monitoring upload data is finally obtained to ensure the reliability of data transmission.

[0105] Step S3: Perform upload voucher processing on the redundant monitoring upload data to generate upload voucher data; perform upload data difference analysis based on the upload voucher data to obtain monitoring upload difference data; perform difference conflict processing on the redundant monitoring upload data through the monitoring upload difference data to obtain energy storage unit monitoring data;

[0106] In an embodiment of the present invention, a hash calculation is performed on the redundant monitoring upload data, for example, using the SHA256 algorithm to generate a unique hash value as an upload credential. The hash value, the corresponding data block ID and the timestamp are stored in the distributed cache Redis to form the upload credential data. The upload credential data is subjected to upload data difference analysis using a data comparison algorithm (for example, a string comparison algorithm or a hash algorithm) to obtain monitoring upload difference data. For example, the hash values ​​corresponding to the same data block ID in the three partitions are compared to analyze whether there are differences. If the hash values ​​are completely consistent, it means that there is no difference in the data; if there are inconsistent hash values, it means that there are differences in the data, and the difference data block IDs are recorded to form monitoring upload difference data. The redundant monitoring upload data is subjected to difference conflict processing using a data merging algorithm (for example, a three-party merging algorithm or a voting algorithm) to obtain the final energy storage unit monitoring data. For example, using a three-party merging algorithm, the data on the three storage nodes are compared, and the final correct data is determined based on the voting results.

[0107] Step S4: Perform dynamic data slicing processing according to the energy storage unit monitoring data to generate monitoring slicing feature data; perform dynamic cache synchronization processing according to the monitoring slicing feature data to generate cache management log data;

[0108] In an embodiment of the present invention, dynamic data sharding is performed according to the characteristics of the energy storage unit monitoring data, such as time range, data type, energy storage unit, etc. For example, the energy storage unit monitoring data of one day can be sharded by hour, and each shard contains the voltage, current, temperature, SOC and other data of all energy storage units within the hour, and a unique ID is generated for each shard, and the time range, data type, data volume and other information of the shard are stored in the monitoring shard characteristic data. According to the monitoring shard characteristic data, the data shards are evenly distributed to different cache server nodes using a consistent hashing algorithm. For example, a consistent hashing algorithm is used to calculate the hash value corresponding to each shard ID, and the shard data is stored on the corresponding cache server node according to the hash value. At the same time, the location, time, update times and other information of each shard data storage are recorded to generate cache management log data.

[0109] Step S5: performing cache status evaluation according to the cache management log data to obtain monitoring cache status data;

[0110] In the embodiment of the present invention, the cache state is evaluated using a machine learning algorithm based on information such as the load state of the cache server node, the access frequency of the data shard, and the data update time. For example, a cache state evaluation model can be constructed using a logistic regression algorithm to predict the access probability of each data shard in the future, and the cache data can be preheated or eliminated based on the prediction results. Finally, monitoring cache state data is generated, including information such as the predicted access probability, cache state, and storage location of each data shard.

[0111] Step S6: Perform a comprehensive management status analysis on the monitored slice feature data by monitoring the cache status data to generate a comprehensive data management index.

[0112] In an embodiment of the present invention, the predicted access probability, cache status, storage location and other information of each data shard can be visualized as a chart, and the data shards can be classified and displayed according to different indicators, such as marking the shards with high access probability as red and the shards with low access probability as green, performing statistical analysis on the monitoring data, and generating a comprehensive data management index based on the analysis results. For example, the comprehensive data management index can be calculated based on indicators such as cache hit rate, data update frequency, and data volume, and used as an important indicator to measure the data management efficiency and performance of industrial energy storage systems. For example, for data shards with high access probability, they can be cached on cache server nodes with higher performance, or the number of their copies can be increased to improve data access efficiency; for data shards with low access probability, they can be eliminated from the cache, or migrated to a storage system with lower storage cost to reduce storage cost, thereby realizing intelligent management of industrial energy storage data.

[0113] Preferably, step S1 comprises the following steps:

[0114] Step S11: deploy intelligent monitoring sensors on industrial energy storage units to obtain monitoring node distribution data;

[0115] Step S12: Perform wireless network configuration according to the monitoring node distribution data, and perform edge computing topology processing to generate intelligent perception network topology data, wherein the intelligent perception network topology data includes energy storage monitoring nodes and distributed computing nodes;

[0116] Step S13: Setting data collection parameters for energy storage monitoring nodes based on intelligent sensing network topology data, and collecting real-time industrial energy storage data streams to obtain initial energy storage unit monitoring data;

[0117] Step S14: using the distributed computing nodes to timestamp the initial energy storage unit monitoring data to obtain time series energy storage monitoring data;

[0118] Step S15: performing data redundancy filtering according to the time series energy storage monitoring data to obtain simplified energy storage monitoring data;

[0119] Step S16: performing outlier detection on the simplified energy storage monitoring data to generate abnormal data points; performing data resampling according to the abnormal data points to obtain resampled monitoring data;

[0120] Step S17: Based on the abnormal data points, the simplified energy storage monitoring data is corrected by resampling the monitoring data to generate corrected energy storage monitoring data.

[0121] In an embodiment of the present invention, according to the specific situation of the industrial energy storage unit, a suitable intelligent monitoring sensor is selected, such as a battery voltage sensor, a current sensor, a temperature sensor, a capacity sensor, etc. Intelligent monitoring sensors are deployed at different locations of the industrial energy storage unit, and the installation location, type, model, and other information of each sensor are recorded to form monitoring node distribution data. According to the monitoring node distribution data, a suitable wireless communication protocol, such as ZigBee, WiFi, LoRa, etc., is selected to configure the sensor network and establish a communication connection between the sensor node and the gateway. For example, according to the connection relationship between the sensor and the edge computing node, a network topology diagram containing all energy storage monitoring nodes and distributed computing nodes is constructed. Using the configuration tool provided by the gateway device, data acquisition parameters are set for each sensor node, such as setting the sampling frequency of the temperature sensor to 1 second, the sampling frequency of the voltage sensor to 0.5 seconds, and the sampling frequency of the current sensor to 2 seconds. After completing the parameter setting, the sensor node starts to collect real-time data at the set frequency, and uploads the collected data to the gateway through the wireless network, and finally converges to form the initial energy storage unit monitoring data. Using a high-precision clock service, such as an NTP service, an accurate timestamp is added to each piece of data to record the data collection time. Perform data redundancy filtering on the time-series energy storage monitoring data. For example, use a sliding window algorithm or a time window algorithm to remove duplicate data points or overly dense data points. For example, an outlier detection method based on the 3σ criterion can be used to calculate the mean and standard deviation of each data point and the data in a period of time before and after it. If the data point deviates from the mean by more than 3 times the standard deviation, the data point is considered to be an abnormal data point. For the detected abnormal data points, data resampling can be performed according to actual conditions. For example, the sampling frequency can be reduced, or a more stable and accurate sensor can be used for data collection, and the resampled data can be used as resampled monitoring data. Use the resampled monitoring data to correct the abnormal data points in the streamlined energy storage monitoring data. For example, a linear interpolation method can be used to interpolate the abnormal data points based on the values ​​of the normal data points before and after the abnormal data points, and the interpolation results can be used as the corrected data.

[0122] As an example of the present invention, refer to Figure 2 As shown, Figure 1 Detailed implementation steps of step S2 in the embodiment are shown in the flowchart. In this embodiment, step S2 includes:

[0123] Step S21: Performing geographic spatial analysis based on the intelligent sensing network topology data to obtain spatial location distribution data;

[0124] In an embodiment of the present invention, the geographic location information of all nodes, such as longitude and latitude coordinates, is extracted from the intelligent sensing network topology data. Then, a geographic information system (GIS) tool, such as ArcGIS or QGIS, is used to visualize the node location information and perform spatial analysis. For example, a spatial clustering algorithm can be used to identify areas with dense nodes and areas with sparse nodes, or a spatial interpolation algorithm can be used to estimate the signal strength of areas where no nodes are deployed. Finally, the analyzed node distribution characteristics, node density, signal coverage, and other information are summarized to form spatial location distribution data.

[0125] Step S22: evaluating the wireless signal strength of the industrial energy storage unit through the spatial location distribution data, optimizing the communication link of the intelligent sensing network topology data, and generating a sensing node communication network;

[0126] In an embodiment of the present invention, the signal strength from each sensor node to the gateway node can be calculated based on the node distance and the signal propagation model, and the signal quality can be judged whether it meets the data transmission requirements based on a preset threshold. The communication link is optimized based on the signal strength evaluation results and the network topology data. For example, for nodes with weaker signal strength, they can be connected to a gateway node that is closer, or the signal can be enhanced by adding relay nodes. The routing strategy can also be dynamically adjusted according to the network load to select the optimal communication path to reduce data transmission delay and packet loss rate. Finally, a perception node communication network containing optimized node connection relationships, communication paths, signal strength and other information is generated.

[0127] Step S23: Connecting distributed storage nodes according to the sensing node communication network to obtain a distributed data storage network;

[0128] In the embodiment of the present invention, multiple gateway nodes that are geographically close can be connected to the same data center, and the data can be stored in the distributed storage nodes in the data center. At the same time, in order to improve data reliability and access efficiency, data copies can be stored in multiple data centers, and the distribution and number of data copies can be dynamically adjusted according to network conditions and node load conditions. Ultimately, a highly available, high-performance, and scalable distributed data storage network is constructed.

[0129] Step S24: using the distributed data storage network to select an upload node for the corrected energy storage monitoring data, and performing parallel task decomposition to obtain multiple upload instruction data;

[0130] In an embodiment of the present invention, a distributed data storage network is used to select an upload node for the corrected energy storage monitoring data according to a load balancing algorithm (e.g., a polling algorithm or a minimum connection number algorithm). For example, a suitable storage node is selected for data upload based on factors such as the remaining storage space of the storage node and the network bandwidth. For example, the nearest data center can be selected based on the geographical location to which the data belongs, and different storage nodes can be selected based on the type of data and importance. Based on the upload node selection result, a task scheduling tool (e.g., Apache Airflow or Luigi) is used to perform parallel task decomposition on the data upload task to generate multi-channel upload instruction data. For example, the corrected energy storage monitoring data is divided into multiple data packets, and a corresponding upload node and upload path are assigned to each data packet based on the upload node selection result.

[0131] Step S25: redundantly copy the corrected energy storage monitoring data to obtain a redundant monitoring data packet;

[0132] In an embodiment of the present invention, before uploading the data, the corrected energy storage monitoring data is redundantly replicated to generate multiple copies. For example, a three-copy strategy can be adopted to generate three completely identical copies of the data. At the same time, in order to ensure the integrity of data transmission, verification information, such as a CRC checksum or an MD5 checksum, can be added to each data packet for the receiver to perform data verification to ensure that no errors occur during data transmission. Finally, the data copy and the verification information are encapsulated into a redundant monitoring data packet and prepared for uploading. For example, for a 100MB data block, three completely identical copies can be generated, and information such as a CRC checksum and a data block number can be added to each copy, ultimately forming three independent redundant monitoring data packets.

[0133] Step S26: performing distributed redundant transmission on redundant monitoring data packets through multiple upload instruction data to obtain redundant monitoring upload data.

[0134] In an embodiment of the present invention, redundant monitoring data packets are uploaded to different storage nodes through different network paths. For example, three data copies can be uploaded to three different data centers, or uploaded to three different nodes in the same data center. Distributed redundant transmission can effectively improve the reliability and efficiency of data transmission. Even if one of the network paths fails or a storage node is abnormal, the data can still be successfully uploaded through other paths and nodes to avoid data loss. Finally, after all data copies are successfully uploaded, redundant monitoring upload data is obtained.

[0135] Preferably, step S3 comprises the following steps:

[0136] Step S31: performing monitoring characteristic analysis on the redundant monitoring upload data to generate key energy storage monitoring characteristic data; performing upload voucher processing according to the key energy storage monitoring characteristic data to generate upload voucher data;

[0137] Step S32: performing credential identification processing on the redundant monitoring upload data by uploading the credential data to obtain the identified monitoring upload data;

[0138] Step S33: Perform multi-node storage network analysis on the redundant uploaded data through the distributed data storage network to generate a multi-node storage network;

[0139] Step S34: using a multi-node storage network to construct a consensus network for the identification monitoring upload data, generating storage consensus network data; performing upload data difference analysis based on the storage consensus network data, obtaining monitoring upload difference data;

[0140] Step S35: Perform node performance evaluation on the storage consensus network data to generate consensus node performance data; perform confidence node recommendation on the storage consensus network data based on the consensus node performance data to obtain confidence node data;

[0141] Step S36: Based on the consensus node performance data, the confidence node data is used to perform a confidence difference weighted judgment on the monitored uploaded difference data, and difference conflict processing is performed to generate uploaded difference conflict processing data;

[0142] Step S37: Based on the confidence node data, the identification monitoring uploaded data is updated for data consistency by uploading the difference conflict processing data to obtain the energy storage unit monitoring data.

[0143] In an embodiment of the present invention, a data analysis tool (e.g., Pandas or Scikit-learn) is used to perform monitoring feature analysis on redundant monitoring upload data to generate key energy storage monitoring feature data. For example, the change trend, abnormal value, fluctuation range and other characteristics of key parameters such as battery voltage, current, temperature, and capacity are analyzed to extract key indicator data. The extracted key energy storage monitoring feature data is hashed to generate a unique hash value as an upload credential, and the hash value is bound to the corresponding data block ID, timestamp and other information to generate upload credential data. For example, the SHA256 algorithm can be used to hash the key feature data to generate a 256-bit hash value as the unique credential of the data block. The upload credential data is added to the corresponding redundant monitoring upload data to identify the source and integrity of the data. For example, the upload credential data can be used as the header information of the data packet, or stored in the metadata of the data packet. A network monitoring tool (e.g., Nagios or Zabbix) is used to perform multi-node storage network analysis on the distributed data storage network to generate multi-node storage network data. According to the upload credential information in the identification monitoring upload data, a distributed consensus algorithm, such as Raft or Paxos, is used to build a consensus network on a multi-node storage network to ensure that all nodes reach a consensus on the authenticity and consistency of the data and generate storage consensus network data. Compare the identification monitoring upload data stored on different nodes to analyze whether there are differences. For example, compare the upload credential data corresponding to the same data block ID. If the hash values ​​are completely consistent, it means that there is no difference in the data; if there are inconsistent hash values, it means that there are differences in the data during the upload process, forming monitoring upload difference data. According to the number of times the node participates in the consensus, response time, data transmission volume and other information recorded in the storage consensus network data, evaluate the node performance and generate consensus node performance data. For example, the node can be scored according to indicators such as the contribution, stability, and reliability of the node. According to the consensus node performance data, a voting mechanism or a scoring mechanism is used to recommend the storage consensus network data for confidence nodes to obtain confidence node data. For example, the node is scored according to its performance indicators, and a high confidence node is selected according to the scoring results. According to the confidence node data, a confidence difference weighted judgment is made on the monitoring upload difference data. For example, different weights can be assigned to data uploaded by different nodes based on the confidence node's score or reputation value. The higher the confidence node, the higher the weight of the data uploaded by it. Then, based on the weighted data, a preset conflict resolution strategy, such as "majority compliance" or "time priority" strategy, is used to handle conflicts in data with differences. For example, if the data of two of the three nodes are consistent, the data of these two nodes shall prevail; if the data of the three nodes are inconsistent, the data with the latest timestamp shall prevail, and upload difference conflict handling data shall be generated.According to the final data version recorded in the uploaded difference conflict processing data, the identification monitoring uploaded data is updated to ensure that the data stored on all nodes remains consistent. For example, the final data version can be synchronized to all nodes, or the old version data can be deleted and only the final version data can be retained.

[0144] Preferably, step S31 includes the following steps:

[0145] Step S311: Identify the data type of the redundant monitoring uploaded data to obtain monitoring type identification data;

[0146] Step S312: performing feature statistics on the redundant monitoring uploaded data through the monitoring type identification data based on the preset monitoring type statistical rules to generate energy storage feature statistical data;

[0147] Step S313: performing feature correlation analysis based on energy storage feature statistical data, and extracting key monitoring features to generate key energy storage monitoring feature data;

[0148] Step S314: using a preset hash algorithm to gradually read the key energy storage monitoring characteristic data, and performing fixed hash value calculation to generate initial hash value data;

[0149] Step S315: rolling update processing is performed on the initial hash value data to obtain key monitoring hash data; metadata is added according to the key monitoring hash data to obtain monitoring credential metadata;

[0150] Step S316: Perform credential serialization processing on the monitoring credential metadata through a preset upload credential template to generate upload credential data.

[0151] In an embodiment of the present invention, the data structure, data label, data content and other information of the redundant monitoring uploaded data are analyzed to identify the type of each data, such as voltage data, current data, temperature data and the like. Regular expression matching, dictionary matching and other methods can be used to identify data types. Preset statistical rules are applied to perform feature statistics. For example, for voltage data, features such as maximum value, minimum value, average value, standard deviation and the like can be counted; for current data, features such as charging current, discharging current, average current and the like can be counted; for temperature data, features such as maximum temperature, minimum temperature and average temperature can be counted. Correlation analysis methods such as Pearson correlation coefficient and Spearman rank correlation coefficient are used to analyze the correlation between different monitoring features and identify redundant features and key features. For example, if the maximum and minimum values ​​of voltage data are highly correlated, one of the features can be selected as the key feature. A secure hash algorithm, such as SHA256 or MD5, is selected to perform hash calculation on key energy storage monitoring feature data. In order to improve the calculation efficiency, a step-by-step reading method can be adopted, where only a part of the feature data is read each time for hash calculation, and the calculation result is iterated with the previous hash value to finally generate a fixed-length hash value as the initial hash value data. As new monitoring data arrives, the initial hash value data is rolled over to ensure that the hash value can reflect the latest data changes. For example, a sliding window method can be used. Each time an update is made, the hash value of the latest data block is added to the calculation, and the hash value of the oldest data block is removed, thereby achieving a rolling update of the hash value and obtaining key monitoring hash data. Use a data serialization tool (such as JSON or XML) to serialize the monitoring credential metadata through a preset upload credential template to generate upload credential data.

[0152] Preferably, step S34 includes the following steps:

[0153] Step S341: using a multi-node storage network to perform node timing credential exchange on the identification monitoring uploaded data to obtain inter-node credential exchange data;

[0154] Step S342: performing hash collision calculation based on the credential exchange data between nodes to obtain preliminary credential similarity data;

[0155] Step S343: Based on a preset similarity threshold, the storage node threshold of the multi-node storage network is screened through preliminary credential similarity data to obtain consensus node list data;

[0156] Step S344: construct and store consensus network data according to the consensus node list data;

[0157] Step S345: Using the storage consensus network data to store node redundant data for the identification monitoring upload data, to obtain consensus node redundant mapping data;

[0158] Step S346: performing data comparison between nodes based on the consensus node redundant mapping data to generate preliminary comparison difference data;

[0159] Step S347: Perform metadata difference analysis based on the preliminary comparison difference data, and perform serialization processing to generate monitoring upload difference data.

[0160] In an embodiment of the present invention, a fixed time interval is set, for example, every 5 minutes, so that each node in the multi-node storage network exchanges the upload credentials of the identification monitoring uploaded data stored by each other. Each node packages the data block ID and the corresponding upload credentials stored by itself and sends them to other nodes, while receiving data packets from other nodes. Finally, each node collects the data block ID and upload credential information of all nodes, and uses a hash algorithm (for example: SHA-256 or MD5) to perform hash collision calculation on the credential exchange data between nodes to obtain preliminary credential similarity data. For example, each node performs hash calculation on the received credential data, and compares it with the credential data stored by itself, and calculates the proportion of the same hash value as the preliminary credential similarity. A similarity threshold is set in advance, for example, 80%. According to the preliminary credential similarity data, data blocks with a consensus degree higher than the threshold are screened out, and the nodes storing these data blocks are added to the consensus node list. For example, if a data block has the same upload credentials in 9 nodes among 10 nodes, the consensus degree of the data block is 90%, which exceeds the preset threshold of 80%, so these 9 nodes are added to the consensus node list. According to the consensus node list data, a new network topology is constructed, called the storage consensus network. The storage consensus network only contains consensus nodes, and the connection relationship between the nodes is adjusted according to the storage situation of the data block. For example, if two consensus nodes store the same data block, a connection is established between the two nodes. The identification monitoring upload data is stored in the storage consensus network, and each data block is ensured to be stored on at least two different consensus nodes to achieve data redundant storage. Record which consensus nodes each data block is stored on, and generate consensus node redundant mapping data. According to the consensus node redundant mapping data, find the consensus nodes that store the same data block, and compare the data stored on these nodes byte by byte. Record the byte position, original data value, node information, etc. of inconsistent data to generate preliminary comparison difference data. Analyze the preliminary comparison difference data to identify the cause of the data difference, such as data transmission error, data writing error, malicious data tampering, etc. Extract the location, type, cause and other information of the difference data, and combine it with metadata such as data block ID and timestamp to perform serialization processing, such as conversion to JSON format, to generate the final monitoring upload difference data.

[0161] Preferably, step S4 comprises the following steps:

[0162] Step S41: Perform dynamic data slicing processing on the monitoring data of the energy storage unit to generate monitoring slicing feature data;

[0163] Step S42: Calculate the monitoring update rate according to the monitoring slice feature data to generate slice update frequency data;

[0164] Step S43: Evaluate the cache demand of the energy storage monitoring slice data by slice update frequency data to generate slice cache demand data;

[0165] Step S44: dividing the data update status of the shard cache demand data based on the preset update rate threshold, when the shard cache demand data is higher than or equal to the preset update rate threshold, marking the monitoring shard feature data as high-frequency monitoring shard data; when the shard cache demand data is lower than the preset update rate threshold, marking the monitoring shard feature data as low-frequency monitoring shard data;

[0166] Step S45: Calculate the average update interval of the high-frequency monitoring slice data to generate average update interval data; perform real-time update window cache processing according to the average update interval data to obtain a high-frequency slice cache strategy;

[0167] Step S46: estimating the cache capacity requirement of the low-frequency monitoring shard data to generate cache capacity requirement data; obtaining available memory resource data; performing secure caching processing on the cache capacity requirement data through the available memory resource data to obtain low-frequency shard cache data;

[0168] Step S47: Perform dynamic cache strategy processing according to the high-frequency shard cache strategy and the low-frequency shard cache data to generate dynamic cache strategy data;

[0169] Step S48: Perform cache synchronization management processing on the monitoring slice feature data through dynamic cache strategy data to generate cache management log data.

[0170] In an embodiment of the present invention, dynamic data sharding is performed according to the characteristics of the energy storage unit monitoring data, such as time, data type, energy storage unit, etc. For example, the energy storage unit monitoring data of one day can be sharded by hour, and each shard contains the voltage, current, temperature, SOC and other data of all energy storage units within the hour, and a unique ID is generated for each shard, and the time range, data type, data volume and other information of the shard are recorded. The time range and data volume of each shard in the monitoring shard feature data are analyzed, and the average data update rate of each shard is calculated, such as the amount of new data added per second. For example, the number of data points in each shard can be counted and divided by the time length of the shard (in seconds) to obtain the average update rate of the shard. According to the shard update frequency data, combined with the preset cache strategy, such as the LRU (least recently used) algorithm or the FIFO (first in first out) algorithm, the cache demand of each shard is evaluated. For example, according to the update frequency of the shard and the preset cache size, it can be calculated how long the shard can store data in the cache, or how much data needs to be cached to meet the access requirements within a certain period of time. Preset an update rate threshold, such as 100 pieces / second, and divide the monitoring shard feature data into high-frequency monitoring shard data and low-frequency monitoring shard data according to the shard update frequency data and the threshold. For example, if the update frequency of a shard is 200 pieces / second, which is higher than the preset threshold, it is marked as high-frequency monitoring shard data; if the update frequency of a shard is 50 pieces / second, which is lower than the preset threshold, it is marked as low-frequency monitoring shard data. For high-frequency monitoring shard data, calculate the average update interval, such as the time interval between two data points, according to its update frequency. According to the average update interval, set a real-time update window, such as 10 seconds, and only cache the data within the time window, and eliminate the old data. For example, if the average update interval of a shard is 0.1 seconds, you can set a 10-second real-time update window to cache only the data within the update window. For low-frequency monitoring shard data, estimate the required cache capacity according to its data volume and preset cache time. For example, if the data volume of a shard is 1GB and the preset cache time is 1 hour, the required cache capacity is 1GB. Get the available memory resource data of the current system, such as the free memory size, the allocatable memory size, etc. Safely cache the low-frequency monitoring shard data based on the available memory resource data and the cache capacity requirement data. For example, you can set a cache upper limit, such as 50% of the total memory, and only cache data that does not exceed the upper limit. If the cache capacity requirement exceeds the upper limit, you can selectively cache part of the data, such as caching only the latest data, or caching more important data. Combine the high-frequency shard cache strategy and the low-frequency shard cache data to generate the final dynamic cache strategy data, which contains information such as the real-time update window size of the high-frequency shard, the cache capacity upper limit of the low-frequency shard, and the cache elimination algorithm.According to the dynamic cache strategy data, cache synchronization management is performed on the monitoring shard feature data. For example, for high-frequency monitoring shard data, cache is performed according to the real-time update window; for low-frequency monitoring shard data, cache is performed according to the cache capacity limit and elimination algorithm. Log information of cache operations, such as cache time, cache data ID, cache result, etc., is recorded to generate cache management log data.

[0171] As an example of the present invention, refer to Figure 3 As shown, Figure 1 Detailed implementation steps of step S5 in the flowchart, in this example, step S5 includes:

[0172] Step S51: Perform data update monitoring according to the cache management log data and generate a data update status report;

[0173] In the embodiment of the present invention, by analyzing indicators such as cache hit rate, cache elimination rate, data update frequency, etc., it is possible to identify which data is being updated frequently and which data has not been updated for a long time. For example, a time window can be set, such as 5 minutes, and the number of updates, cache hits, cache eliminations, and other indicators of each data shard within the time window can be counted. If the number of updates of a data shard exceeds a preset threshold, it is considered that the shard is in an active state and the cache needs to be updated in time; if a data shard has not been updated for a long time, such as more than 1 hour, it can be considered to be eliminated from the cache to release cache space.

[0174] Step S52: performing update notification broadcast processing on the distributed data storage network based on the data update status report, and extracting the cache status summary to generate cache status summary data;

[0175] In the embodiment of the present invention, the data shards that need to be updated in the cache are identified according to the data update status report, and the update notification is broadcast to all nodes in the distributed data storage network. The publish / subscribe mode can be used to send the update notification to all nodes that have subscribed to the data shard. At the same time, key cache status information is extracted from the data update status report, such as the latest update time, cache status, cache node information, etc. of each data shard, to generate cache status summary data.

[0176] Step S53: performing cache consistency evaluation according to the cache status summary data to generate cache consistency evaluation data; performing inconsistent node identification on the cache consistency evaluation data to obtain inconsistent node list data;

[0177] In an embodiment of the present invention, after all nodes receive the cache status summary data, they compare the data shard information in their own cache with the summary data to evaluate the consistency of the cache. For example, the latest update time of the data shards can be compared. If the update time of the local cache is earlier than the update time in the summary data, it means that the local cache is expired and needs to be updated. A cache consistency evaluation is performed based on the cache status summary data to generate cache consistency evaluation data. For example, the contents of the same data shards on different storage nodes are compared to determine whether the cached data is consistent. The cache consistency evaluation data is analyzed to identify nodes with inconsistent caches, such as nodes with expired cache data or inconsistent data versions, and these nodes are added to the inconsistent node list data.

[0178] Step S54: repair the node cache according to the inconsistent node list data, and re-evaluate the cache status to obtain monitoring cache status data.

[0179] In an embodiment of the present invention, for nodes in the inconsistent node list data, cache repair is performed according to the specific inconsistency situation. For example, if the data cached by the node has expired, the latest data is obtained from the data source or other cache nodes for updating; if the data versions are inconsistent, according to the preset conflict resolution strategy, for example, the system compares the cache data timestamp of the inconsistent node with the latest timestamp in the cache status summary data. If the timestamp of the node cache data is earlier than the summary data, the node cache data is determined to be expired, and the system initiates a data request to the data source or other cache nodes with the latest data, obtains the latest version of the data shard, and replaces the expired cache data with the new data. After the cache repair is completed, the cache status evaluation is performed again to ensure that the cache data of all nodes has been updated to the latest version and is consistent. The final cache status evaluation results are recorded to generate monitoring cache status data.

[0180] Preferably, step S6 comprises the following steps:

[0181] Step S61: Obtain historical energy storage data access records; perform access pattern recognition on monitoring slice feature data through historical energy storage data access records to generate data access patterns;

[0182] Step S62: Estimating the access probability of the monitored shard feature data based on the data access pattern to generate shard access probability data; constructing a predictive shard data index based on the shard access probability data;

[0183] Step S63: performing a shard data query simulation according to the predictive shard data index, and locating the target data shard to obtain shard query feature data;

[0184] Step S64: Perform dynamic access path evaluation according to the shard query feature data to generate access path evaluation data; perform data prefetch efficiency evaluation according to the shard query feature data to generate data prefetch evaluation data;

[0185] Step S65: performing cache effect evaluation on the shard query feature data by monitoring the cache status data to generate cache effect evaluation data;

[0186] Step S66: Perform a comprehensive analysis of the management status based on the access path evaluation data, the data pre-fetch evaluation data, and the cache effect evaluation data to generate a comprehensive data management index.

[0187] In an embodiment of the present invention, historical energy storage data access records are obtained from a database or log file, such as which users or applications accessed which data shards at what time. Data mining techniques, such as association rule mining or sequence pattern mining, can be used to analyze historical access records and identify data access patterns. According to the identified data access pattern, combined with context information such as current time, user identity, and application scenario, the access probability of the monitoring shard feature data is estimated to predict the possibility of each shard being accessed in the future. For example, a machine learning algorithm, such as logistic regression, support vector machine, etc., can be used to build a prediction model to predict the access probability of each shard based on historical access patterns and current context information. According to the shard access probability data, a predictive shard data index is constructed, such as placing shards with high access probabilities in front of the index and shards with low access probabilities in the back of the index, so as to find the data required by the user faster. Simulate a user submitting a data query request, such as querying the voltage data of a certain energy storage unit in a certain time period. According to the predictive shard data index, the candidate shard containing the target data is quickly located. Record the characteristic information of the query request, such as the query time range, data type, data source, etc., as well as the information of the located candidate shards, and generate shard query characteristic data. According to the shard query characteristic data, evaluate different data access paths, such as reading data from the cache, reading data from the local disk, reading data from the remote node, etc., and calculate the access delay, network bandwidth consumption and other indicators of each path. Store the evaluation results in the access path evaluation data. Evaluate the efficiency of data prefetching according to the shard query characteristic data and data access mode. For example, you can calculate the prefetched data hit rate, that is, what proportion of the actually accessed data has been prefetched into the cache. Store the evaluation results in the data prefetch evaluation data. Evaluate the effect of the current cache strategy based on the monitored cache status data, such as the cache hit rate and cache capacity utilization rate. For example, you can analyze the cache hit rate of different data shards, identify the shards with poor cache effect, and adjust the cache strategy according to the analysis results. Comprehensively analyze the access path evaluation data, data prefetch evaluation data and cache effect evaluation data to evaluate the overall performance of the data management system, such as data access delay, data throughput, resource utilization, etc. Weights can be set according to different indicators to calculate a comprehensive data management index, which reflects the overall health and efficiency of data management and realizes the effective management and security of industrial energy storage unit monitoring data.

[0188] Preferably, the present invention further provides a data management system for an industrial energy storage system, which executes the data management method for an industrial energy storage system as described above, and the data management system for an industrial energy storage system comprises:

[0189] The real-time data acquisition module is used to perform edge computing topology processing on the industrial energy storage unit to generate intelligent perception network topology data; perform real-time energy storage data acquisition based on the intelligent perception network topology data to obtain time-series energy storage monitoring data; perform abnormal data correction based on the time-series energy storage monitoring data to generate corrected energy storage monitoring data;

[0190] A distributed redundant transmission module is used to optimize distributed storage nodes according to the intelligent perception network topology data to obtain a distributed data storage network; and to perform distributed redundant transmission of the corrected energy storage monitoring data using the distributed data storage network to obtain redundant monitoring upload data;

[0191] The data consistency maintenance module is used to process the upload voucher of the redundant monitoring upload data to generate the upload voucher data; perform upload data difference analysis according to the upload voucher data to obtain the monitoring upload difference data; perform difference conflict processing on the redundant monitoring upload data through the monitoring upload difference data to obtain the energy storage unit monitoring data;

[0192] A dynamic data cache module is used to perform dynamic data sharding processing according to the monitoring data of the energy storage unit to generate monitoring shard feature data; perform dynamic cache synchronization processing according to the monitoring shard feature data to generate cache management log data;

[0193] The cache status analysis module is used to evaluate the cache status according to the cache management log data to obtain the monitoring cache status data;

[0194] The comprehensive data management evaluation module is used to conduct a comprehensive analysis of the management status of the monitoring shard feature data by monitoring the cache status data and generate a comprehensive data management index.

[0195] The beneficial effect of this application is that by deploying intelligent monitoring sensors in each energy storage unit, and using wireless networks and edge computing nodes to build a distributed perception network, decentralized data source collection and preprocessing are realized. The edge computing node adds a timestamp in real time, and performs redundant filtering, anomaly detection and correction on the data, ensuring the timing, simplicity and accuracy of the data, and reducing the pressure of subsequent data transmission and processing. Using geospatial analysis technology, the wireless signal strength and communication link are optimized, and multiple gateway nodes with close geographical locations are connected to the same data center to achieve data storage nearby and reduce network latency. At the same time, data sharding and multi-copy storage strategies are adopted to store data in multiple nodes in a decentralized manner, so that even if a single node fails, the integrity and availability of the data will not be affected. Each data block will generate a unique hash value as a credential before uploading, and identify data inconsistencies through credential exchange and collision calculation between nodes. At the same time, a consensus algorithm is used to ensure that all nodes agree on the final version of the data, effectively avoiding data conflicts and errors. According to the data update frequency and access mode, data shards are divided into high-frequency and low-frequency categories, and real-time update window cache and capacity demand estimation cache strategies are used to improve cache hit rate and reduce data access latency. At the same time, by analyzing historical data access records, predicting future data access patterns, and building predictive shard data indexes, data query efficiency is further optimized. By real-time monitoring of data update status, cache inconsistencies are discovered in a timely manner, and node caches are repaired according to inconsistency types and preset conflict resolution strategies to avoid errors caused by data asynchrony. This flexible cache management strategy improves the system's response speed and realizes efficient, reliable, and secure upload management of massive industrial energy storage monitoring data.

[0196] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0197] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. 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 invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A data management method for an industrial energy storage system, characterized in that: The following steps are involved: Step S1: Perform edge computing topology processing on the industrial energy storage unit to generate intelligent perception network topology data; perform real-time energy storage data collection based on the intelligent perception network topology data to obtain time-series energy storage monitoring data; Correct abnormal data according to the time-series energy storage monitoring data to generate corrected energy storage monitoring data; Step S2: Optimizing distributed storage nodes according to the intelligent sensing network topology data to obtain a distributed data storage network; using the distributed data storage network to perform distributed redundant transmission on the corrected energy storage monitoring data to obtain redundant monitoring upload data; Step S3: Perform upload voucher processing on the redundant monitoring upload data to generate upload voucher data; perform upload data difference analysis based on the upload voucher data to obtain monitoring upload difference data; perform difference conflict processing on the redundant monitoring upload data through the monitoring upload difference data to obtain energy storage unit monitoring data; Step S4: Perform dynamic data slicing processing according to the monitoring data of the energy storage unit to generate monitoring slicing feature data; Perform dynamic cache synchronization processing based on the monitoring shard feature data to generate cache management log data; Step S5: performing cache status evaluation according to the cache management log data to obtain monitoring cache status data; Step S6: Perform a comprehensive management status analysis on the monitored slice feature data by monitoring the cache status data to generate a comprehensive data management index.

2. The data management method for an industrial energy storage system according to claim 1, characterized in that: Step S1 includes the following steps: Step S11: deploy intelligent monitoring sensors on industrial energy storage units to obtain monitoring node distribution data; Step S12: Perform wireless network configuration according to the monitoring node distribution data, and perform edge computing topology processing to generate intelligent perception network topology data, wherein the intelligent perception network topology data includes energy storage monitoring nodes and distributed computing nodes; Step S13: Setting data collection parameters for energy storage monitoring nodes based on intelligent sensing network topology data, and collecting real-time industrial energy storage data streams to obtain initial energy storage unit monitoring data; Step S14: using the distributed computing nodes to timestamp the initial energy storage unit monitoring data to obtain time series energy storage monitoring data; Step S15: performing data redundancy filtering according to the time series energy storage monitoring data to obtain simplified energy storage monitoring data; Step S16: performing outlier detection on the simplified energy storage monitoring data to generate abnormal data points; performing data resampling according to the abnormal data points to obtain resampled monitoring data; Step S17: Based on the abnormal data points, the simplified energy storage monitoring data is corrected by resampling the monitoring data to generate corrected energy storage monitoring data.

3. The data management method for an industrial energy storage system according to claim 1, characterized in that: Step S2 includes the following steps: Step S21: Performing geographic spatial analysis based on the intelligent sensing network topology data to obtain spatial location distribution data; Step S22: evaluating the wireless signal strength of the industrial energy storage unit through the spatial location distribution data, optimizing the communication link of the intelligent sensing network topology data, and generating a sensing node communication network; Step S23: Connecting distributed storage nodes according to the sensing node communication network to obtain a distributed data storage network; Step S24: using the distributed data storage network to select an upload node for the corrected energy storage monitoring data, and performing parallel task decomposition to obtain multiple upload instruction data; Step S25: redundantly copy the corrected energy storage monitoring data to obtain a redundant monitoring data packet; Step S26: performing distributed redundant transmission on redundant monitoring data packets through multiple upload instruction data to obtain redundant monitoring upload data.

4. The data management method for an industrial energy storage system according to claim 1, characterized in that: Step S3 includes the following steps: Step S31: performing monitoring characteristic analysis on the redundant monitoring upload data to generate key energy storage monitoring characteristic data; performing upload voucher processing according to the key energy storage monitoring characteristic data to generate upload voucher data; Step S32: performing credential identification processing on the redundant monitoring upload data by uploading the credential data to obtain the identified monitoring upload data; Step S33: Perform multi-node storage network analysis on the redundant uploaded data through the distributed data storage network to generate a multi-node storage network; Step S34: using a multi-node storage network to construct a consensus network for the identification monitoring upload data, generating storage consensus network data; performing upload data difference analysis based on the storage consensus network data, obtaining monitoring upload difference data; Step S35: Perform node performance evaluation on the storage consensus network data to generate consensus node performance data; perform confidence node recommendation on the storage consensus network data based on the consensus node performance data to obtain confidence node data; Step S36: Based on the consensus node performance data, the confidence node data is used to perform a confidence difference weighted judgment on the monitored uploaded difference data, and difference conflict processing is performed to generate uploaded difference conflict processing data; Step S37: Based on the confidence node data, the identification monitoring uploaded data is updated for data consistency by uploading the difference conflict processing data to obtain the energy storage unit monitoring data.

5. The data management method for an industrial energy storage system according to claim 4, characterized in that: Step S31 includes the following steps: Step S311: Identify the data type of the redundant monitoring uploaded data to obtain monitoring type identification data; Step S312: performing feature statistics on the redundant monitoring uploaded data through the monitoring type identification data based on the preset monitoring type statistical rules to generate energy storage feature statistical data; Step S313: performing feature correlation analysis based on energy storage feature statistical data, and extracting key monitoring features to generate key energy storage monitoring feature data; Step S314: using a preset hash algorithm to gradually read the key energy storage monitoring characteristic data, and performing fixed hash value calculation to generate initial hash value data; Step S315: rolling update processing is performed on the initial hash value data to obtain key monitoring hash data; metadata is added according to the key monitoring hash data to obtain monitoring credential metadata; Step S316: Perform credential serialization processing on the monitoring credential metadata through a preset upload credential template to generate upload credential data.

6. The data management method for an industrial energy storage system according to claim 4, characterized in that: Step S34 includes the following steps: Step S341: using a multi-node storage network to perform node timing credential exchange on the identification monitoring uploaded data to obtain inter-node credential exchange data; Step S342: performing hash collision calculation based on the credential exchange data between nodes to obtain preliminary credential similarity data; Step S343: Based on a preset similarity threshold, the storage node threshold of the multi-node storage network is screened through preliminary credential similarity data to obtain consensus node list data; Step S344: construct and store consensus network data according to the consensus node list data; Step S345: Using the storage consensus network data to store node redundant data for the identification monitoring upload data, to obtain consensus node redundant mapping data; Step S346: performing data comparison between nodes based on the consensus node redundant mapping data to generate preliminary comparison difference data; Step S347: Perform metadata difference analysis based on the preliminary comparison difference data, and perform serialization processing to generate monitoring upload difference data.

7. The data management method for an industrial energy storage system according to claim 1, characterized in that: Step S4 includes the following steps: Step S41: Perform dynamic data slicing processing on the monitoring data of the energy storage unit to generate monitoring slicing feature data; Step S42: Calculate the monitoring update rate according to the monitoring slice feature data to generate slice update frequency data; Step S43: Evaluate the cache demand of the energy storage monitoring slice data by slice update frequency data to generate slice cache demand data; Step S44: dividing the data update status of the shard cache demand data based on the preset update rate threshold, when the shard cache demand data is higher than or equal to the preset update rate threshold, marking the monitoring shard feature data as high-frequency monitoring shard data; when the shard cache demand data is lower than the preset update rate threshold, marking the monitoring shard feature data as low-frequency monitoring shard data; Step S45: Calculate the average update interval of the high-frequency monitoring slice data to generate average update interval data; perform real-time update window cache processing according to the average update interval data to obtain a high-frequency slice cache strategy; Step S46: estimating the cache capacity requirement of the low-frequency monitoring shard data to generate cache capacity requirement data; obtaining available memory resource data; performing secure caching processing on the cache capacity requirement data through the available memory resource data to obtain low-frequency shard cache data; Step S47: Perform dynamic cache strategy processing according to the high-frequency shard cache strategy and the low-frequency shard cache data to generate dynamic cache strategy data; Step S48: Perform cache synchronization management processing on the monitoring slice feature data through dynamic cache strategy data to generate cache management log data.

8. The data management method for an industrial energy storage system according to claim 1, characterized in that: Step S5 includes the following steps: Step S51: Perform data update monitoring according to the cache management log data and generate a data update status report; Step S52: performing update notification broadcast processing on the distributed data storage network based on the data update status report, and extracting the cache status summary to generate cache status summary data; Step S53: performing cache consistency evaluation according to the cache status summary data to generate cache consistency evaluation data; performing inconsistent node identification on the cache consistency evaluation data to obtain inconsistent node list data; Step S54: repair the node cache according to the inconsistent node list data, and re-evaluate the cache status to obtain monitoring cache status data.

9. The data management method for an industrial energy storage system according to claim 1, characterized in that: Step S6 includes the following steps: Step S61: Obtain historical energy storage data access records; perform access pattern recognition on monitoring slice feature data through historical energy storage data access records to generate data access patterns; Step S62: Estimating the access probability of the monitored shard feature data based on the data access pattern to generate shard access probability data; constructing a predictive shard data index based on the shard access probability data; Step S63: performing a shard data query simulation according to the predictive shard data index, and locating the target data shard to obtain shard query feature data; Step S64: Perform dynamic access path evaluation according to the shard query feature data to generate access path evaluation data; perform data prefetch efficiency evaluation according to the shard query feature data to generate data prefetch evaluation data; Step S65: performing cache effect evaluation on the shard query feature data by monitoring the cache status data to generate cache effect evaluation data; Step S66: Perform a comprehensive analysis of the management status based on the access path evaluation data, the data pre-fetch evaluation data, and the cache effect evaluation data to generate a comprehensive data management index.

10. A data management system for an industrial energy storage system, characterized in that: For executing the data management method of the industrial energy storage system according to claim 1, the data management system of the industrial energy storage system comprises: The real-time data acquisition module is used to perform edge computing topology processing on the industrial energy storage unit to generate intelligent perception network topology data; perform real-time energy storage data acquisition based on the intelligent perception network topology data to obtain time-series energy storage monitoring data; perform abnormal data correction based on the time-series energy storage monitoring data to generate corrected energy storage monitoring data; A distributed redundant transmission module is used to optimize distributed storage nodes according to the intelligent perception network topology data to obtain a distributed data storage network; and to perform distributed redundant transmission of the corrected energy storage monitoring data using the distributed data storage network to obtain redundant monitoring upload data; The data consistency maintenance module is used to process the upload voucher of the redundant monitoring upload data to generate the upload voucher data; perform upload data difference analysis according to the upload voucher data to obtain the monitoring upload difference data; perform difference conflict processing on the redundant monitoring upload data through the monitoring upload difference data to obtain the energy storage unit monitoring data; A dynamic data cache module is used to perform dynamic data sharding processing according to the monitoring data of the energy storage unit to generate monitoring shard feature data; perform dynamic cache synchronization processing according to the monitoring shard feature data to generate cache management log data; The cache status analysis module is used to evaluate the cache status according to the cache management log data to obtain the monitoring cache status data; The comprehensive data management evaluation module is used to conduct a comprehensive analysis of the management status of the monitoring shard feature data by monitoring the cache status data and generate a comprehensive data management index.

Images