Industrial and commercial energy storage power station management system and method based on big data analysis

Abstract

The invention provides an industrial and commercial energy storage power station management system and method based on big data analysis, and relates to the technical field of data synchronization. Determining an operation risk instruction recording the power station in the current operation stage in each state partition database according to all the operation attribute linked lists; synchronizing all the energy scheduling labels and dynamic scheduling conditions marked in the operation record database in the current operation stage to a copy of an operation adjustment type, and updating a data field to which cloud edge collaboration belongs through the adjustment data code to obtain an operation management index in the operation record database; and determining an operation response strategy of the industrial and commercial energy storage power station, and copying the operation response strategy to the management request based on big data analysis. According to the invention, full-process management and control can be carried out on the operation data of the industrial and commercial energy storage power station under the cloud-side collaborative architecture, so that the pertinence of operation adjustment during management of the energy storage power station is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data synchronization, more particularly, the present application relates to a business and industrial energy storage power station management system and method based on big data analysis. BACKGROUND

[0002] Data synchronization is widely used in cloud edge node data interaction, device terminal operation data upload, power grid dispatching information issuance and other whole data flow scenarios of energy storage power stations. Business and industrial energy storage power station operation involves battery pack status, converter parameters, power grid load, dispatching instructions and other types of data. The data sources are scattered in edge terminals, cloud platforms, power grid interfaces and other nodes, and have the characteristics of high real-time requirement, large data volume and strong correlation. At the same time, the accuracy of data directly affects the effectiveness of risk assessment and dispatching decision. Relying on data synchronization technology to realize real-time transmission, consistency verification and standardized integration of data from each node can meet the multi-source data interaction needs of energy storage power stations, and is the key foundation to ensure the accuracy of big data analysis, the timeliness of management decision and the pertinence of operation adjustment.

[0003] However, in the existing management of business and industrial energy storage power stations based on big data analysis, the data screening lacks a hierarchical mechanism adapted to the state partition, the generation of operation risk instructions lacks correlation with real-time operation status, the energy dispatching label and dynamic dispatching condition are out of synchronization in time sequence, and the cloud edge data domain update is disconnected from the construction of operation management index, resulting in chaotic processing of multi-source operation status data, lagging risk identification, ambiguous energy dispatching direction, poor cloud edge data consistency, and inability of operation response strategy to accurately match the actual operation needs of the power station, thereby restricting the timeliness and control precision of management decision. Therefore, how to perform whole-process management and control on the operation data of business and industrial energy storage power stations under the cloud edge collaborative architecture to improve the pertinence of operation adjustment during the management of energy storage power stations is a problem faced by the industry. SUMMARY

[0004] The present application provides a business and industrial energy storage power station management system and method based on big data analysis, which can perform whole-process management and control on the operation data of business and industrial energy storage power stations under the cloud edge collaborative architecture to improve the pertinence of operation adjustment during the management of energy storage power stations.

[0005] In a first aspect, the present application provides a business and industrial energy storage power station management method based on big data analysis, which comprises the following steps: obtaining operation status data of a business and industrial energy storage power station in response to a cloud edge collaborative energy storage power station management platform based on big data analysis; determine the data screening rule when the power station operation information is retrieved, perform hierarchical screening on the operation state data according to the data screening rule, obtain the operation attribute chain table corresponding to the multiple state partition databases in the operation process of the power station, and then determine the operation risk instruction of the power station in the current operation stage in each state partition database from all the operation attribute chain tables; extract the energy scheduling label corresponding to the operation state in each state partition database, synchronize all the energy scheduling labels and the dynamic scheduling condition marked in the operation record database in the current operation stage to the copy of the operation adjustment type, obtain the adjustment data code on the target adjustment point, and then update the data domain belonging to the cloud edge cooperation through the adjustment data code to obtain the operation management index in the operation record database; determine the operation response strategy of the industrial and commercial energy storage power station according to all the operation risk instructions and the operation management index, and copy the operation response strategy to the management request based on big data analysis.

[0006] In this embodiment, the data screening rule when the power station operation information is retrieved specifically includes: extract the real-time operation characteristic value of each data source node from the power station operation information; synchronize all the real-time operation characteristic values based on a distributed consensus protocol to obtain a screening constraint attribute; generate the data screening rule when the power station operation information is retrieved from the screening constraint attribute.

[0007] In this embodiment, the data screening rule refers to an executable rule set for accurately screening effective data from massive power station operation information.

[0008] In this embodiment, the state partition database refers to a dedicated data storage unit divided according to the type of energy storage power station equipment and the dimension of data.

[0009] In this embodiment, determining the operation risk instruction of the power station in the current operation stage in each state partition database from all the operation attribute chain tables specifically includes: extract the operation characteristic parameter representing the power station in the current operation stage in each state partition database; submit the operation characteristic parameter to a risk assessment node cluster to obtain the operation risk index corresponding to each operation characteristic parameter; fuse each operation risk index by a risk assessment master node to obtain the operation risk instruction of the power station in the current operation stage in each state partition database.

[0010] In this embodiment, extracting the energy scheduling label corresponding to the operation state in each state partition database specifically includes: determine the original state sequence corresponding to the running state in each state partition database; Each master node sends all original state sequences to a label generation cluster composed of distributed computing nodes to obtain multiple label core values; The label generation cluster asynchronously replicates all core label values to obtain energy scheduling labels corresponding to the running state in each state partition database.

[0011] In this embodiment, all energy scheduling labels and dynamic scheduling conditions marked in the running record database of the current running phase are synchronized to the replica of the running adjustment type to obtain adjustment data encoding at the target adjustment point, which specifically includes: The dynamic scheduling conditions that have reached consensus are extracted and verified by the coordination node of the running record database; All energy scheduling labels and dynamic scheduling conditions are packaged as transaction update logs and synchronized to the replica node of the running adjustment type, and the write confirmation of the replica node is waited; The received energy scheduling labels and dynamic scheduling conditions are mapped according to the pre-defined rule to generate local pre-encoding results; Each replica node submits the local pre-encoding result to a consensus group for comparison to generate adjustment data encoding at the target adjustment point.

[0012] In this embodiment, the running response strategy of the industrial and commercial energy storage power station is determined according to all running risk instructions and the running management index, which specifically includes: Extract the running response condition of each state partition from the running management index; Submit all running risk instructions and all running response conditions to a matching node cluster based on majority consensus to obtain risk-response adaptation weights under each partition; Generate the running response strategy of the industrial and commercial energy storage power station according to all risk-response adaptation weights.

[0013] In this embodiment, the management request based on big data analysis refers to a structured request initiated by the energy storage power station big data management platform, carrying the running response strategy and associated data.

[0014] In a second aspect, the present application provides an industrial and commercial energy storage power station management system based on big data analysis, which is used to execute an industrial and commercial energy storage power station management method based on big data analysis, and the management system includes: A data acquisition module is used to respond to a cloud-edge collaborative energy storage power station management platform based on big data analysis to acquire the running state data of the industrial and commercial energy storage power station; The data screening module is configured to determine a data screening rule for searching power station operation information, perform hierarchical screening on the operation state data according to the data screening rule, obtain operation attribute chain tables corresponding to a plurality of state partition databases in a power station operation process, and then determine, from all the operation attribute chain tables, an operation risk instruction in each state partition database recording a current operation stage of the power station. The data synchronization module is configured to extract an energy scheduling tag corresponding to an operation state in each state partition database, synchronize all the energy scheduling tags and dynamic scheduling conditions marked in the operation record database in the current operation stage to a copy of the operation adjustment type, obtain adjustment data coding on a target adjustment point, and then update a data domain belonging to the cloud edge collaboration through the adjustment data coding to obtain an operation management index in the operation record database. The data replication module is configured to determine an operation response strategy of the industrial and commercial energy storage power station according to all the operation risk instructions and the operation management index, and replicate the operation response strategy to a management request based on big data analysis.

[0015] The technical scheme provided by the embodiments disclosed in the application has the following beneficial effects: The technical scheme provided by the embodiments disclosed in the application has the following beneficial effects:

[0016] Therefore, in the present application, the precision of cloud-edge collaborative management can be improved when the commercial and industrial energy storage power station management data is chaotic. The running state data is obtained through the cloud-edge collaborative platform, which can break through the barriers of device terminals, power grid interfaces and other multi-source data dispersion, and ensure the universality and real-time of data collection. Through hierarchical screening data and generating running risk instructions, the precise correlation between risk and working condition can be realized by relying on the screening rules and double model evaluation of state partition adaptation, and timely and clear risk identification is ensured. By generating adjustment data coding and updating the running management index, the scheduling label and constraint conditions can be associated with standardized coding, and the consistency of cloud-edge data is ensured by relying on the consensus mechanism, and the problem of disconnection between index and data update is solved. By determining the running response strategy and encapsulating it into the management request, a feasible scheme can be formed through simulation verification and standardized transmission, solving the problem of poor landing and chaotic transmission format of traditional strategies, and improving the intelligent and collaborative level of energy storage power station management.

[0017] In summary, the technical solution adopted by the present application can control the running data of commercial and industrial energy storage power stations in the whole process under the cloud-edge collaborative architecture to improve the pertinence of running adjustment in energy storage power station management. BRIEF DESCRIPTION OF DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0019] Figure 1 is an exemplary flowchart of a commercial and industrial energy storage power station management method based on big data analysis provided by the present application; Figure 2 is a flowchart of determining a running attribute chain table provided by the present application; Figure 3 is a flowchart of determining a running management index provided by the present application; Figure 4 is a module structure diagram of a commercial and industrial energy storage power station management system based on big data analysis provided by the present application. DETAILED DESCRIPTION

[0020] The technical solutions in the embodiments of the present application will be described clearly and completely in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0021] This application provides a management system and method for industrial and commercial energy storage power stations based on big data analysis. Its core is to respond to a cloud-edge collaborative energy storage power station management platform based on big data analysis, acquire operational status data of the industrial and commercial energy storage power station; determine data filtering rules for power station operation information retrieval, and perform hierarchical filtering of the operational status data according to the data filtering rules to obtain operational attribute linked lists corresponding to multiple status partition databases during power station operation; then determine the operational risk instructions recorded in each status partition database for the current operational stage from all operational attribute linked lists; extract energy scheduling tags corresponding to the operational status in each status partition database, synchronize all energy scheduling tags and the dynamic scheduling conditions marked in the current operational stage's operational record database to a copy of the operational adjustment type, obtain adjustment data codes at the target adjustment points, and then update the data domain to which the cloud-edge collaboration belongs through the adjustment data codes to obtain the operational management index in the operational record database; determine the operational response strategy of the industrial and commercial energy storage power station based on all operational risk instructions and the operational management index, and copy the operational response strategy to the management request based on big data analysis.

[0022] Example 1: To better understand the above technical solution, the following will provide a detailed description of the technical solution in conjunction with the accompanying drawings and specific implementation methods. (Refer to...) Figure 1 As shown in the figure, this is an exemplary flowchart of a management method for industrial and commercial energy storage power stations based on big data analysis according to this embodiment of the application. The management method includes the following steps: In step S1, the system responds to the cloud-edge collaborative energy storage power station management platform based on big data analysis to obtain the operating status data of industrial and commercial energy storage power stations.

[0023] In practice, firstly, industrial-grade sensors and data acquisition modules are deployed on the battery clusters, energy storage converters, battery management systems, cooling systems, and grid connection interface equipment of commercial and industrial energy storage power stations. These sensors collect various parameters such as battery output voltage, charging and discharging current, remaining capacity, energy storage converter conversion efficiency, grid connection node voltage, and ambient temperature in the equipment room. Data noise is eliminated and invalid data exceeding the normal operating range of the equipment is removed using a moving average method. The standardized data is then uploaded to the cloud platform in real time via a 5G communication network or industrial Ethernet according to the TCP / IP protocol. Simultaneously, a complete copy of the data is stored locally at the edge nodes. The cloud uses a Hadoop distributed file system to segment and store the received data and generate backups to ensure data security. The operating status data of the commercial and industrial energy storage power station can be obtained by reading the cloud file system, which will not be elaborated further here.

[0024] It should be noted that, in this application, the cloud-edge collaborative energy storage power station management platform based on big data analysis refers to a platform that, through collaboration between cloud and edge nodes, relies on big data analysis to realize power station data collection, analysis, scheduling, and strategy distribution, thereby ensuring the safe and economical operation of the power station; the operating status data refers to the raw data set of equipment operating conditions, energy flow, and environmental characteristics of industrial and commercial energy storage power stations.

[0025] In step S2, the data filtering rules for retrieving power plant operation information are determined, and the operation status data is filtered in a hierarchical manner according to the data filtering rules to obtain the operation attribute linked list corresponding to multiple status partition databases during the operation of the power plant. Then, the operation risk instructions of the power plant in the current operation stage are determined by all the operation attribute linked lists.

[0026] In this embodiment, the data filtering rules for retrieving power plant operation information can be determined using the following steps: Extract the real-time operating characteristic values ​​of each data source node from the power plant's operating information; Based on a distributed consensus protocol, all real-time running feature values ​​are synchronized to obtain the filtering constraint attributes; The data filtering rules for retrieving power plant operation information are generated from the filtering constraint attributes.

[0027] In practical implementation, firstly, the data source nodes encompass the acquisition terminals of devices such as battery clusters, energy storage converters, and cooling systems. Raw operational information such as voltage, current, and temperature is collected through sensors at each node. The raw data is processed using the Z-Score standardization method, calculating the difference between each data point and the dataset mean, divided by the standard deviation to eliminate the influence of dimensions. Then, a sliding window method is used, setting a fixed time window, to calculate statistical indicators such as the mean, variance, and peak value of the data within the window. Simultaneously, correlation analysis in feature engineering is used to eliminate redundant indicators, ultimately obtaining the real-time operational characteristic values ​​of each data source node. Finally, the Practical Byzantine Fault Tolerance (PBFT) protocol is used to synchronize the characteristic values. This protocol includes four stages: proposal, pre-preparation, preparation, and confirmation. First, the pre-defined master node collects real-time operational feature values ​​from all data source nodes. After verifying the integrity of these feature values, it generates proposal information. Then, the master node broadcasts the proposal to all slave nodes. Slave nodes verify the proposal's validity and return a pre-prepared response. Once the master node collects sufficient pre-prepared responses, it initiates a preparation request. Slave nodes verify the request, record the feature values, and return a preparation response. Finally, when the master node collects preparation responses from more than two-thirds of the nodes, it sends a confirmation command to all nodes, synchronizing all real-time operational feature values. These synchronized real-time operational feature values ​​are then used as filtering constraint attributes. Finally, the filtering constraint attributes are categorized into three types: validity constraints, classification constraints, and priority constraints. For validity constraints, data value range rules are set; data exceeding this range in the operational information is considered invalid. For classification constraints, mapping rules are established between attributes and device types and data dimensions to clarify the category of different data types. For priority constraints, high, medium, and low priority rules are established based on attribute importance, with high-priority data participating in the retrieval first. The three types of constraint attributes are transformed into executable rules in the form of "if-then" using logical programming. These executable rules are then used as data filtering rules for retrieving power plant operation information.

[0028] It should be noted that, in this application, the data source node refers to the information acquisition terminal deployed on the key equipment of the energy storage power station; the real-time operating characteristic value refers to the indicator obtained after processing the raw data collected by the data source node; the distributed consensus protocol refers to the communication protocol used to coordinate the data transmission of multiple data source nodes; the filtering constraint attribute refers to the boundary conditions for clarifying data filtering; and the data filtering rule refers to the set of executable rules for accurately filtering effective data from massive amounts of power station operation information.

[0029] Preferably, in this embodiment, the operating status data is hierarchically filtered according to the data filtering rules to obtain a linked list of operating attributes corresponding to multiple status partition databases during the power plant operation, with reference to... Figure 2As shown in the figure, this is a flowchart illustrating the process of determining the runtime attribute linked list in some embodiments of this application. In this embodiment, determining the runtime attribute linked list can be achieved using the following steps: In step S21, hierarchical filtering conditions corresponding to different state partitions are parsed from the data filtering rules; In step S22, the running status data is distributed in parallel according to all the hierarchical screening conditions to obtain the initial screening marks on multiple processing nodes; In step S23, a hierarchical redundancy strategy is determined in the data copies carrying hierarchical labels based on all the initial screening markers; In step S24, the hierarchical redundancy strategy is aggregated and verified to obtain the running attribute linked list corresponding to multiple state partition databases during the power plant operation.

[0030] In practical implementation, firstly, a rule parsing engine can be used to structurally parse the data filtering rules. First, keyword matching identifies the identifiers representing state partitions within the rules. Then, a syntax analysis algorithm extracts the grading criteria corresponding to each state partition, including the value range for validity filtering, the device data mapping relationship for classification filtering, and the importance level threshold for priority filtering. During parsing, a cross-validation algorithm verifies the logical consistency of the conditions, eliminating contradictory conditions, ultimately forming each state partition and its corresponding grading filtering conditions. Next, a MapReduce parallel computing framework can be used. The running state data is first hash-sharded according to the state partition identifier, ensuring that data from the same partition is distributed to adjacent processing nodes to improve processing efficiency. Then, the grading filtering conditions for each state partition are synchronized to the corresponding processing nodes. Each processing node verifies each received data item, determining whether it meets the validity, classification, and priority conditions of the current partition. Data that meets the conditions is marked with its corresponding state partition and priority level; data that does not meet the conditions is marked as invalid. During the filtering process, a load balancing algorithm dynamically adjusts the data volume of each node to avoid overloading a single node, thus obtaining the initial filtering labels across multiple processing nodes. Then, priority levels and corresponding state partition information are extracted from all initial screening tags. A corresponding hierarchical label is added to each valid data entry, and a data copy is generated. A redundancy calculation model is used to determine the number of redundant storage copies based on data priority level, historical access frequency, and fault impact range. High-priority data corresponds to a higher number of redundant copies, while medium- and low-priority data have fewer redundant copies as needed. Simultaneously, an association mapping is established between redundant data and state partitions to ensure that the hierarchical redundancy strategy in the data copy is accurately assigned to the corresponding partition. After the strategy is formulated, the recoverability of the redundant data is verified by simulating fault scenarios, thus obtaining the hierarchical redundancy strategy in the data copy. Finally, during aggregation verification, the aggregation node collects the hierarchical redundant data from all processing nodes using a consistent hashing algorithm. First, the integrity of each redundant data is verified, and data with missing key fields is removed. Then, the consistency of different copies of the same data is verified through data fingerprint comparison, retaining valid data that meets the standards. Valid data is classified by state partition, and a timestamp sorting algorithm is used to arrange the data in each partition in ascending order of collection time, generating a linked list of operating attributes containing data value, collection time, priority level, and the identifier of the device to which it belongs. This yields the linked list of operating attributes corresponding to multiple state partition databases during the power plant's operation.

[0031] It should be noted that, in this application, the state partition database refers to a dedicated data storage unit divided according to the type of energy storage power station equipment and data dimensions; the hierarchical screening conditions refer to a set of multi-level screening criteria formulated for different state partitions; parallel allocation refers to the method of splitting massive operational status data according to preset rules and distributing it to multiple processing nodes for simultaneous processing; the initial screening mark refers to the identifier generated by the processing node after verifying the data according to the hierarchical screening conditions, indicating whether the data meets the conditions and to which the data belongs; the hierarchical label refers to the exclusive identifier that identifies the priority level of the data that passes the initial screening and the state partition to which it belongs; the hierarchical redundancy strategy refers to a differentiated data backup storage scheme formulated according to the data priority level; and the operational attribute linked list refers to a structured data set arranged in the order of data acquisition time in each state partition database.

[0032] In this embodiment, determining the operational risk instructions of the power plant in the current operational phase recorded in the database for each state partition based on all operational attribute linked lists can be achieved through the following steps: Extract the operational characteristic parameters representing the power plant's current operating phase from the database of each state partition; The operational characteristic parameters are submitted to the risk assessment node cluster to obtain the operational risk index corresponding to each operational characteristic parameter; The risk assessment master node integrates each operational risk indicator to obtain the operational risk instructions of the power plant in the current operational stage, which are recorded in the database of each status partition.

[0033] In practical implementation, firstly, the current operating stage of the power station is determined through an operating status determination model. This model is based on key data such as the power of the energy storage converter and the remaining battery capacity in the status partition database, and outputs the charging, discharging, standby, or emergency fault stage results through logical determination rules. Next, for each status partition database, a filtering method from feature engineering is used. Combined with the core requirements of the current operating stage, core indicators related to safe operation and energy conversion are extracted from the operating attribute linked list. Low-discrimination parameters are eliminated using an analysis of variance algorithm, retaining indicators with significant distinguishability as the operating characteristic parameters of the current operating stage. After extraction, data standardization is performed to unify the units of measurement. Then, the risk assessment node cluster is divided into corresponding assessment nodes according to the status partition, and each assessment node pre-deploys a threshold comparison model and a trend analysis model. The extracted operational characteristic parameters are assigned to corresponding evaluation nodes according to state partitions. Each node first compares the parameter values ​​with preset safety thresholds (calibrated based on equipment factory standards, industry specifications, and historical fault data) using a threshold comparison model. Then, a trend analysis model uses a linear regression algorithm to calculate the slope of parameter changes. The results from both methods are then normalized and mapped to the [0,1] interval to obtain the operational risk index corresponding to each operational characteristic parameter. The closer the operational risk index value is to 1, the higher the risk. Finally, the risk assessment master node uses a weighted summation fusion algorithm to integrate the various operational risk indicators. The weights are determined using the analytic hierarchy process (AHP) and, combined with the impact of each characteristic parameter on the safe operation of the power plant, are calibrated after consistency verification. During fusion, the comprehensive risk value of all operational risk indicators is first calculated according to their weights. Then, the comprehensive risk value is compared with a preset risk level threshold (which can be trained using historical fault data and its effectiveness verified through multiple simulations) to determine the risk level. Finally, by combining the status partition identifier and the current operation stage information, a structured operation risk instruction containing risk type, risk level, associated parameters, and scope of impact is generated. After generation, the accuracy of the instruction is verified by comparing it with historical cases, thus determining the operation risk instruction of the power plant in the current operation stage recorded in the database for each status partition.

[0034] It should be noted that, in this application, the operating characteristic parameter refers to the core indicator that quantitatively reflects the operating status of the power plant; the current operating stage refers to the specific operating state of the power plant based on real-time operating conditions; the risk assessment node cluster refers to a parallel processing cluster composed of multiple independent risk assessment nodes; the operating risk indicator refers to a numerical value that quantitatively represents the degree of risk; the risk assessment master node refers to the core coordination node of the risk assessment node cluster; and the operating risk instruction refers to a structured instruction generated based on the fused risk assessment results.

[0035] In step S3, the energy scheduling tag corresponding to the running status in each state partition database is extracted, and all energy scheduling tags and the dynamic scheduling conditions marked in the running record database of the current running stage are synchronized to the copy of the running adjustment type to obtain the adjustment data code on the target adjustment point. Then, the data domain to which the cloud-edge collaboration belongs is updated through the adjustment data code to obtain the running management index in the running record database.

[0036] In this embodiment, extracting the energy scheduling tag corresponding to the running state in each state partition database can be achieved using the following steps: Determine the original state sequence corresponding to the running state in each state partition database; Each master node sends all the original state sequences to the tag generation cluster composed of distributed computing nodes to obtain multiple tag core values; The tag generation cluster asynchronously replicates all core tag values ​​to obtain the energy scheduling tag corresponding to the running status in each state partition database.

[0037] In practice, firstly, for each state partition database, core data reflecting the operating status, including equipment operating parameters and energy flow data, are extracted from the operating attribute linked list. These are then sorted in ascending order by acquisition time using a timestamp sorting algorithm to form an initial sequence. Linear interpolation is used to fill in missing values ​​in the sequence, and outliers are identified and removed using the 3σ principle. A sliding window method is then used to smooth the sequence, eliminating transient fluctuations and ultimately obtaining a continuous and stable original state sequence. Next, each master node distributes the original state sequence to the corresponding computing nodes in the label generation cluster according to the state partition using a load balancing algorithm, ensuring a balanced workload across nodes. Each computing node pre-deploys a K-means clustering model and scheduling requirement matching rules. The K-means clustering model is used to perform cluster analysis on the original state sequence, grouping similar operating states into one category. Then, based on preset scheduling requirement matching rules (which can be formulated by combining industry scheduling standards and historical operating data), the quantified scheduling requirement value corresponding to each state category is calculated, and this quantified value is used as the core label value. Finally, the tag generation cluster uses the Raft consensus algorithm to asynchronously replicate tag core values. The master node broadcasts the generated tag core value to all slave nodes. After receiving the value, the slave nodes verify the data. If the verification is successful, they send back confirmation information. Once the master node collects confirmation from a majority of nodes, it marks the replication as complete. After replication, the cluster uses a preset tag core value-tag mapping table (different scheduling directions correspond to different tag core value ranges) to convert each tag core value into a standardized energy scheduling tag. This results in the energy scheduling tag corresponding to the running state in each state partition database. This energy scheduling tag contains key information such as scheduling type and priority.

[0038] It should be noted that, in this application, the original state sequence refers to the core data set of the running state in the state partition database arranged in chronological order; the distributed computing node refers to the independent computing unit in the tag generation cluster that undertakes data processing tasks; the tag generation cluster refers to a parallel processing cluster composed of multiple distributed computing nodes; asynchronous replication refers to the non-real-time but eventually consistent data transmission method between the master node and the slave node in the tag generation cluster; the tag core value refers to the quantitative indicator of the scheduling demand intensity of the corresponding running state; and the energy scheduling tag refers to the identifier that clarifies the energy scheduling direction and priority corresponding to the running state of each state partition.

[0039] In this embodiment, synchronizing all energy scheduling tags and the dynamic scheduling conditions marked in the current operation phase's operation record database to the copy of the operation adjustment type to obtain the adjustment data code at the target adjustment point can be achieved through the following steps: The coordinating node in the runtime record database extracts and verifies the consensus-reached dynamic scheduling conditions; Package all energy scheduling tags and the dynamic scheduling conditions into a transaction update log, synchronize it to the replica nodes of the running adjustment type, and wait for the replica nodes to confirm the write. The received energy scheduling tags are mapped to dynamic scheduling conditions according to predefined rules to generate local precoding results; Each replica node submits its local precoding result to a consensus group for comparison, generating the adjustment data code at the target adjustment point.

[0040] In practice, the coordinating node first traverses the runtime record database using database indexes to extract dynamic scheduling conditions marked as "consensus," including grid peak and off-peak periods, equipment power limits, and battery remaining power thresholds. A practical Byzantine fault-tolerant protocol is used to verify the extracted conditions. The coordinating node broadcasts the conditions to other slave nodes in the runtime record database. Slave nodes verify the logical consistency and timeliness of the conditions. When more than two-thirds of the slave nodes confirm the validity of the conditions, they are considered consensus-reached dynamic scheduling conditions. Conditions that fail verification are marked and rejected for reconfirmation. Next, all energy scheduling tags and verified dynamic scheduling conditions are packaged into a transaction update log in the format of "tag identifier - condition identifier - associated priority." The log is stored in a JSON structure, containing fields such as data generation timestamp and data checksum. The transaction update log is synchronized to all replica nodes of the runtime adjustment type via TCP / IP protocol. Each replica node receives the log, verifies its integrity using the checksum, writes the log to its local storage module, and returns a "write successful" confirmation signal to the coordinating node after writing. The coordinating node then waits for feedback from all replica nodes. Then, the predefined rules include mapping relationships between four core fields: adjustment point identifier, scheduling type, constraints, and priority. These rules are based on industry scheduling standards and power plant operation requirements, and their rationality has been verified through multiple simulations. Each replica node calls a hash mapping algorithm to match the received energy scheduling tag's corresponding scheduling type and the dynamic scheduling condition's corresponding constraint range with the fields in the predefined rules, generating a string in the order of "adjustment point – scheduling type – constraints – priority." This string is then converted into a standardized local pre-coded result using a Base64 encoding algorithm. Finally, the consensus group consists of three primary verification nodes and multiple secondary verification nodes. Each replica node submits its local pre-coded result and corresponding mapping log to the consensus group. The consensus group uses a practical Byzantine fault-tolerant protocol for comparison. The primary verification nodes first verify the format standardization of the pre-coded results, then compare the character consistency of each result, while the secondary verification nodes synchronously verify the results. When all primary verification nodes confirm that more than 90% of the replica nodes have consistent precoding results, the consistent result is used as the final adjusted data encoding; if there is a discrepancy, the mapping log is retrieved to trace the cause of the difference, the precoding result is regenerated and compared again until a consensus is reached.

[0041] It should be noted that, in this application, the operation record database refers to a structured database that stores historical operation data, dispatch execution records, dynamic dispatch conditions, and equipment operating logs of industrial and commercial energy storage power stations; dynamic dispatch conditions refer to the dispatch constraints marked in the operation record database that are updated in real time based on the grid status and power station operation requirements; operation adjustment type refers to the specific category that defines the operation adjustment of industrial and commercial energy storage power stations; target adjustment point refers to the specific object in the energy storage power station that needs to perform energy dispatch and operation status adjustment; transaction update log refers to a structured data log that records the association between energy dispatch tags and dynamic dispatch conditions; replica node refers to a backup node that stores operation adjustment-related data; predefined rules refer to a set of pre-defined mapping rules used to associate energy dispatch tags with dynamic dispatch conditions; local precoding result refers to the preliminary coding generated by the replica node based on local data; and adjustment data coding refers to the standardized coding that uniquely identifies the dispatch requirements and constraints of the target adjustment point.

[0042] Preferably, in this embodiment, the data domain to which cloud-edge collaboration belongs is updated by adjusting the data encoding to obtain the operation management index in the operation record database, with reference to... Figure 3 As shown in the figure, this is a flowchart illustrating the process of determining the runtime management index in some embodiments of this application. In this embodiment, determining the runtime management index can be achieved using the following steps: In step S31, the leading node in cloud-edge collaboration parses the adjusted data encoding into data update instructions for different data partitions; In step S32, all data update commands are synchronized to the replica of the data domain to which the cloud-edge collaboration belongs, and the system waits for confirmation of receipt from the replica. In step S33, each data domain copy performs update operations in parallel locally and generates local index fragments based on the updated data state; In step S34, each copy submits the generated local index fragment to the index coordination service to obtain the runtime management index in the runtime record database.

[0043] In practical implementation, firstly, the master node pre-deploys an encoding parsing engine. This engine has built-in structured parsing rules for adjusting data encoding, which clearly define the splitting logic of fields such as adjustment point identifier, scheduling type, constraints, and priority in the encoding. During parsing, the encoding parsing engine first uses a syntax analysis algorithm to decompose the string structure of the adjustment data encoding, extracting the core information of each field. Then, it uses a hash mapping algorithm to associate and match the field information with the preset data partitions of the cloud-edge collaborative data domain to determine the target data partition corresponding to each encoding. Finally, it generates structured data update instructions in the format of "data partition ID - update field - target value - execution priority". After parsing, a logical verification algorithm verifies the feasibility of the instruction execution, discarding invalid instructions and using the verified instructions as data update instructions. Next, the master node adopts a publish-subscribe model, using 5G industrial Ethernet to batch distribute all data update instructions to the corresponding data domain replicas according to their data partition affiliation. After receiving the instruction, each replica node first verifies the integrity of the instruction data using an MD5 checksum algorithm. After confirming that no data is lost or tampered with, it temporarily stores the instruction in its local cache module and performs write preprocessing. After the write preprocessing is complete, the replica nodes return an acknowledgment signal containing the instruction checksum and the receiving timestamp to the master node. The master node continuously listens for feedback from all replica nodes until it receives acknowledgment signals from all replicas. Replicas that have not received acknowledgment will re-receive the instruction. Then, each data domain replica executes a local data update using a transactional update mechanism based on the received data update instruction. This transactional update mechanism first locks the fields to be updated in the target data partition, modifies the data according to the target value in the instruction, and records the state changes before and after the update through the log to ensure the atomicity of the update operation (either all are executed or all are rolled back). After the update is complete, the replica nodes call the B+ tree index construction algorithm to construct a structured local index fragment using the core retrieval fields of the data partition (such as adjustment point identifier, update timestamp, scheduling type) as the index key. This local index fragment contains key information such as the index key, data storage address, and data validity identifier. Finally, the index coordination service deploys a Byzantine fault-tolerant consensus protocol for consistency verification and aggregation of the local index fragment. Each data domain replica submits local index fragments and corresponding update logs to the index coordination service via TCP / IP protocol. The index coordination service first verifies the format specifications and data integrity of each fragment, eliminating fragments with format errors or missing key information. Subsequently, using a Byzantine fault-tolerant consensus protocol, it compares the index keys and data mapping relationships of all valid fragments to confirm consistency among fragments. When more than 2 / 3 of the fragments reach consensus, an index merging algorithm is used to aggregate the fragments according to their data partition affiliation and retrieval fields, generating a globally structured runtime management index. This globally structured runtime management index is then used as the runtime management index in the runtime record database.

[0044] It should be noted that, in this application, the data domain to which cloud-edge collaboration belongs refers to the collection of storage-wide operational data, scheduling information, and encoding mapping relationships under the cloud-edge collaboration architecture; the leading node refers to the coordinating node that coordinates data encoding parsing, instruction distribution, and synchronization management; data update instructions refer to executable instructions that specify the update content, operation type, and execution standard of data partitions; the partition-level index generated based on the local data state after a single data domain copy of a local index fragment is updated; the index coordination service refers to the service unit in the cloud-edge collaboration architecture that is specifically responsible for the aggregation, verification, and global index generation of local index fragments; and the operation management index refers to the fast retrieval path provided for the operation record database.

[0045] In step S4, the operation response strategy of the industrial and commercial energy storage power station is determined based on all operation risk instructions and the operation management index, and the operation response strategy is copied into the management request based on big data analysis.

[0046] In this embodiment, determining the operational response strategy of a commercial and industrial energy storage power station based on all operational risk instructions and the operational management index can be achieved through the following steps: Extract the running response conditions for each status partition from the running management index; All operational risk instructions and all operational response conditions are submitted to the matching node cluster based on majority consensus to obtain the risk-response adaptation weights for each partition. The operational response strategy for commercial and industrial energy storage power plants is generated based on all risk-response adaptation weights.

[0047] In practical implementation, firstly, an index traversal algorithm can be used to perform structured parsing of the operation management index. The index fragment corresponding to each state partition is located using the state partition identifier as the search keyword. Then, core constraint information related to the operation response is extracted from the fragment, including equipment operating parameter thresholds, scheduling operation permissions, and power grid interaction specifications. During the extraction process, a logical verification algorithm is used to verify the consistency and timeliness of the constraint information, eliminating expired or contradictory conditions, ultimately forming the operation response conditions for each state partition. Next, the matching node cluster allocates tasks according to the state partitions, and a load balancing algorithm ensures balanced workload across nodes. Each node is pre-deployed with the Analytic Hierarchy Process (AHP) and a fuzzy comprehensive evaluation model. The risk level, impact range, constraint strength, and applicable scenarios of the operation risk instruction are first decomposed into indicators. Then, the weight of each indicator is determined using the AHP, and the fuzzy comprehensive evaluation model is used to calculate the fit between risk and response. A majority consensus mechanism is adopted. When the deviation of more than 2 / 3 of the nodes' calculation results is within a preset range, the average value is taken as the risk-response adaptation weight for that partition. Nodes with deviations exceeding the limit need to recalculate and verify. Finally, based on the risk-response adaptation weight from high to low, the response conditions with the highest adaptation degree are selected as the core handling methods. Combined with the operational characteristics of each state partition and the global scheduling requirements, a preliminary strategy framework is formed. The framework includes core modules such as the execution entity, operation steps, parameter thresholds, execution time limits, and emergency plans. The execution entity clarifies the execution permissions of cloud or edge nodes according to the cloud-edge collaborative architecture, and the operation steps are arranged in logical order to ensure feasibility. Digital twin simulation technology is used to simulate the strategy execution effect based on real-time power plant operation data. The feasibility of the strategy is verified by analyzing indicators such as risk elimination rate and energy utilization efficiency. Based on the simulation results, parameter thresholds and execution processes are adjusted and optimized, ultimately forming a formal operational response strategy, namely the operational response strategy for industrial and commercial energy storage power plants.

[0048] It should be noted that in this application, the operational response conditions refer to the set of constraints that limit the operational adjustment boundary; the matching node cluster refers to a distributed computing cluster composed of multiple parallel processing nodes; the risk-response adaptation weight refers to the numerical value that quantifies the degree of fit between the operational risk instruction and the corresponding operational response conditions; and the operational response strategy refers to the handling of operational risks by industrial and commercial energy storage power stations, which clarifies the implementing entity, operation steps, parameter thresholds, execution time limits, and emergency plans for risk handling.

[0049] In specific implementation, copying the execution response strategy to the management request based on big data analytics can be achieved in the following way: First, the strategy formatting interface is called to convert the execution response strategy into standardized data in JSON format according to the hierarchical structure of "execution subject - operation instruction - parameter threshold - emergency trigger condition". Redundant comments are removed and data field descriptions are added to generate structured strategy text. Then, the management request construction program is started to automatically generate a request header containing a unique request identifier, initiation timestamp, and sending node information. An independent strategy storage segment is opened in the data payload area of ​​the request. The structured strategy text is completely embedded into this storage segment using byte stream copying technology. After embedding, the SHA-256 algorithm is called to calculate the checksum of the strategy data and appended to the checksum field at the end of the request. Finally, the encapsulated management request is pushed to the big data management platform through the platform API interface. After receiving the request, the platform verifies the integrity of the strategy data through the checksum. If it is correct, the copying and receiving process is completed. If the verification fails, the request retransmission mechanism is triggered. Further details are omitted here.

[0050] It should be noted that, in this application, the management request based on big data analysis refers to a structured request initiated to the big data management platform of the energy storage power station, which carries the operation response strategy and related data.

[0051] Therefore, this application demonstrates its ability to improve the accuracy of cloud-edge collaborative management when management data for commercial and industrial energy storage power stations is disorganized. Specifically, acquiring operational status data through a cloud-edge collaborative platform breaks down the barriers of fragmented data from multiple sources, such as equipment terminals and grid interfaces, ensuring comprehensive and real-time data acquisition. By hierarchically filtering data and generating operational risk instructions, and relying on state-partitioned filtering rules and dual-model evaluation, accurate correlation between risks and operating conditions can be achieved, ensuring timely and clear risk identification. By generating and adjusting data codes and updating the operational management index, standardized codes can be used to associate scheduling tags and constraints, and a consensus mechanism ensures cloud-edge data consistency, resolving the problem of index and data update disconnect. By determining operational response strategies and encapsulating them into management requests, feasible solutions can be formed through simulation verification and standardized transmission can be achieved, solving the problems of poor implementation and disorganized transmission formats of traditional strategies, and improving the intelligence and collaboration level of energy storage power station management.

[0052] In summary, the technical solution adopted in this application can perform full-process control of the operation data of industrial and commercial energy storage power stations under the cloud-edge collaborative architecture, so as to improve the pertinence of operation adjustments during the management of energy storage power stations.

[0053] Example 2: This application provides a management system for industrial and commercial energy storage power stations based on big data analysis, referencing... Figure 4 As shown in the figure, this is a module structure diagram of a commercial and industrial energy storage power station management system based on big data analysis according to this embodiment of the present application. The management system includes: The data acquisition module 100 is used to respond to the cloud-edge collaborative energy storage power station management platform based on big data analysis and acquire the operating status data of industrial and commercial energy storage power stations. The data filtering module 200 is used to determine the data filtering rules when retrieving power plant operation information, and to perform hierarchical filtering on the operation status data according to the data filtering rules to obtain the operation attribute linked list corresponding to multiple status partition databases during the operation of the power plant, and then to determine the operation risk instructions of the power plant in the current operation stage recorded in each status partition database by all the operation attribute linked lists. The data synchronization module 300 is used to extract the energy scheduling tag corresponding to the running status in each state partition database, synchronize all energy scheduling tags and the dynamic scheduling conditions marked in the running record database of the current running stage to the copy of the running adjustment type, obtain the adjustment data code on the target adjustment point, and then update the data domain to which the cloud-edge collaboration belongs through the adjustment data code to obtain the running management index in the running record database. The data replication module 400 is used to determine the operation response strategy of the industrial and commercial energy storage power station based on all operation risk instructions and the operation management index, and to replicate the operation response strategy to the management request based on big data analysis.

[0054] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0055] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, including read-only memory (ROM), random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), one-time programmable read-only memory (OTPROM), electrically-Erasable Programmable Read-Only Memory (EEPROM), compactdisc read-only memory (CD-ROM) or other optical disc storage, disk storage, magnetic tape storage, or any other computer-readable medium capable of carrying or storing data.

[0056] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

Claims

1. A method for managing industrial and commercial energy storage power stations based on big data analysis, characterized in that, The management method comprises the following steps: In response to a cloud-edge collaborative energy storage power station management platform based on big data analysis, obtain operation state data of industrial and commercial energy storage power stations; Determine data filtering rules for power station operation information retrieval, and perform hierarchical filtering on the operation state data according to the data filtering rules to obtain operation attribute chain tables corresponding to a plurality of state partition databases in the operation process of the power station, and then determine, from all the operation attribute chain tables, an operation risk instruction for recording the current operation stage of the power station in each state partition database; Extract energy scheduling tags corresponding to the operation state in each state partition database, synchronize all the energy scheduling tags and dynamic scheduling conditions marked in the operation record database in the current operation stage to a copy of the operation adjustment type to obtain adjustment data codes on target adjustment points, and then update the data domain to which the cloud-edge collaboration belongs through the adjustment data codes to obtain an operation management index in the operation record database; Determine an operation response strategy of the industrial and commercial energy storage power station according to all the operation risk instructions and the operation management index, and copy the operation response strategy into a management request based on big data analysis.

2. A method for managing industrial and commercial energy storage power stations based on big data analysis according to claim 1, characterized in that, The determination of the data filtering rules for power station operation information retrieval specifically comprises: Extract real-time operation characteristic values of each data source node from the power station operation information; Synchronize all the real-time operation characteristic values based on a distributed consensus protocol to obtain filtering constraint attributes; Generate the data filtering rules for power station operation information retrieval from the filtering constraint attributes.

3. A method for managing industrial and commercial energy storage power stations based on big data analysis according to claim 1, characterized in that, The data filtering rules refer to an executable rule set for accurately filtering effective data from a large amount of power station operation information.

4. A method for managing industrial and commercial energy storage plants based on big data analysis according to claim 1, characterized in that, The state partition database refers to a dedicated data storage unit divided according to the type of energy storage power station equipment and the dimension of data.

5. A big data analytics based industrial and commercial energy storage power station management method according to claim 1, characterized in that, The determination of the operation risk instruction for recording the current operation stage of the power station in each state partition database from all the operation attribute chain tables specifically comprises: Extract operation characteristic parameters representing the current operation stage of the power station in each state partition database; Submit the operation characteristic parameters to a risk assessment node cluster to obtain operation risk indicators corresponding to each operation characteristic parameter; Fuse each operation risk indicator by a risk assessment master node to obtain the operation risk instruction for recording the current operation stage of the power station in each state partition database.

6. A big data analytics based industrial and commercial energy storage power station management method according to claim 1, characterized in that, The extraction of the energy scheduling tags corresponding to the operation state in each state partition database specifically comprises: Determine original state sequences corresponding to the operation state in each state partition database; Each master node sends all the original state sequences to a tag generation cluster composed of distributed computing nodes to obtain a plurality of tag core values; The tag generation cluster asynchronously replicates all the core tag values to obtain the energy scheduling tags corresponding to the operation state in each state partition database.

7. A big data analytics based industrial and commercial energy storage power station management method according to claim 1, characterized in that, The synchronization of all the energy scheduling tags and the dynamic scheduling conditions marked in the operation record database in the current operation stage to a copy of the operation adjustment type to obtain adjustment data codes on target adjustment points specifically comprises: Extract and verify the dynamic scheduling conditions that have reached a consensus by a coordination node of the operation record database; Pack all the energy scheduling labels and the dynamic scheduling conditions as transaction update logs, and synchronize to the replica nodes of the running adjustment type, and wait for the write confirmation of the replica nodes; Map the received energy scheduling labels and dynamic scheduling conditions according to a pre-defined rule to generate a local pre-encoding result; Each replica node submits the local pre-encoding result to a consensus group for comparison to generate adjustment data encoding at the target adjustment point.

8. A big data analytics based industrial and commercial energy storage power station management method according to claim 1, characterized in that, The operation response strategy of the industrial and commercial energy storage power station is determined according to all the operation risk instructions and the operation management index, and specifically includes: Extract the operation response condition of each state partition from the operation management index; Submit all the operation risk instructions and all the operation response conditions to a matching node cluster based on majority consensus to obtain the risk-response adaptation weight under each partition; Generate the operation response strategy of the industrial and commercial energy storage power station according to all the risk-response adaptation weights.

9. A big data analytics based industrial and commercial energy storage power station management method according to claim 1, characterized in that, The management request based on big data analysis refers to a structured request initiated by the energy storage power station big data management platform, carrying the operation response strategy and associated data.

10. A big data analysis based industrial and commercial energy storage power station management system for performing a big data analysis based industrial and commercial energy storage power station management method according to any one of claims 1 to 9, characterized in that, The management system includes: A data acquisition module for responding to the cloud-edge collaborative energy storage power station management platform based on big data analysis, acquiring the operation state data of the industrial and commercial energy storage power station; A data filtering module for determining the data filtering rule for power station operation information retrieval, performing hierarchical filtering on the operation state data according to the data filtering rule, obtaining the operation attribute chain table corresponding to the state partition database in the power station operation process, and then determining the operation risk instruction recorded in each state partition database in the current operation stage of the power station from all the operation attribute chain tables; A data synchronization module for extracting the energy scheduling label corresponding to the operation state in each state partition database, synchronizing all the energy scheduling labels and the dynamic scheduling condition marked in the operation record database of the current operation stage to the running adjustment type replica to obtain the adjustment data encoding at the target adjustment point, and then updating the data domain of the cloud-edge collaboration through the adjustment data encoding to obtain the operation management index in the operation record database; A data replication module for determining the operation response strategy of the industrial and commercial energy storage power station according to all the operation risk instructions and the operation management index, and replicating the operation response strategy to the management request based on big data analysis.

Images