Multi-source data fusion distributed energy management platform and management method thereof

Abstract

The invention relates to a multi-source data fusion distributed energy management platform and a management method thereof. The platform comprises a data acquisition module, a data fusion module, a data analysis module, a distributed control module and a communication module. The platform collects multi-source heterogeneous data in real time, carries out data fusion and dynamic updating based on local cache nodes, generates an energy optimization strategy by utilizing machine learning and carries out adaptive adjustment, and meanwhile, guarantees the traceability and tamper-proofing performance of data and control instructions by adopting a block chain technology; the method comprises the steps of data acquisition and synchronization, data fusion processing, optimization analysis and prediction, control instruction execution and exception processing, and the scheduling efficiency, robustness and operation reliability of the distributed energy system can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of energy management, in particular to a multi-source data fusion distributed energy management platform and its management method, specifically to multi-source heterogeneous data collection and fusion, energy optimization scheduling strategy generation, distributed control execution and blockchain data protection technologies. BACKGROUND

[0002] With the development of renewable energy technology and the diversification of energy structure, distributed energy systems play an increasingly important role in smart grids. Distributed energy systems usually include photovoltaic, wind energy, energy storage devices and other energy equipment, whose operating state is limited by geography, environment, load and other factors, and has high dynamicity and uncertainty. In order to achieve efficient management of such systems, the energy management platform needs to have the ability of real-time collection, fusion analysis and scheduling decision of multi-source heterogeneous data.

[0003] Existing energy management systems are mainly based on centralized architecture, which has the problems of high response delay, poor system stability, and large single-point failure risk, and is difficult to adapt to the requirements of edge processing, autonomous control and collaborative scheduling of distributed energy systems. At the same time, the authenticity of data and the traceability of instructions are also key issues in distributed energy scenarios, and traditional systems often lack effective protection mechanisms.

[0004] In addition, existing solutions rely heavily on data fusion and analysis capabilities, and usually ignore the collaborative design of data preprocessing, semantic mapping, and exception recovery, resulting in poor system scalability and limited ability to adapt to complex environments. Therefore, there is an urgent need for an energy management platform and method that supports multi-source data fusion, has distributed scheduling capability, and can improve data security and control reliability. SUMMARY

[0005] The purpose of the present application is to provide a multi-source data fusion distributed energy management platform and its management method to solve the technical problems of unstable data collection, insufficient fusion capability, slow scheduling response and lack of traceability of control instructions in the prior art.

[0006] According to the present application, a multi-source data fusion distributed energy management platform is provided, which comprises: a data collection module for collecting multi-source heterogeneous data from distributed energy devices, sensors and external data sources in real time, the multi-source heterogeneous data including energy production data, consumption data, environmental data and market price data, and periodically synchronizing the collection results to local cache nodes; a data fusion module configured to preprocess, standardize and fuse the multi-source heterogeneous data, generate an energy management data set in a unified format, and realize dynamic updating and rollback protection of data based on the local cache mechanism; a data analysis module configured to analyze the energy management dataset based on a machine learning algorithm, generate an energy optimization strategy and a prediction model, and periodically adapt the energy optimization strategy according to real-time feedback data; a distributed control module configured to send control instructions to distributed energy devices according to the energy optimization strategy, and support each device node to make fine-tuning decisions according to local states, forming a master-autonomous collaborative scheduling system; a communication module configured to provide a secure data transmission channel between the data collection module, the data fusion module, the data analysis module, and the distributed control module, and support asynchronous communication and chain instruction confirmation across regional nodes; wherein the platform realizes traceability and tamper resistance of the data and the control instructions through blockchain technology, and each node has the ability to independently record and verify the blockchain interaction log.

[0007] Further, the data collection module comprises: an Internet of Things interface unit supporting multiple communication protocols to adapt to distributed energy devices and sensors, and dynamically switching data collection frequency according to node load; a data verification unit configured to perform integrity and consistency checks on the multi-source heterogeneous data, and trigger a local cache node re-sampling mechanism when abnormal data is detected to improve data collection reliability.

[0008] Further, the data fusion module comprises: a data cleaning unit for handling missing values, duplicates, and noise in the multi-source heterogeneous data, and supporting real-time adjustment of cleaning rules according to environmental changes; a semantic mapping unit configured to map data from different sources to a unified energy management data model based on semantic analysis, and seamlessly integrate new data sources through incremental mapping when they are accessed.

[0009] Further, the data analysis module comprises: an optimization strategy unit that generates the energy optimization strategy using an optimization algorithm, the energy optimization strategy is calculated by the following objective function: min

[0010] wherein T is the optimization time period, is the energy production cost at time t, is the carbon emissions of energy consumption, is the energy distribution scheme at time t, is the environmental weight coefficient, and is dynamically adjusted according to actual execution results after each period ; a prediction submodule, which trains a prediction model based on a deep learning algorithm, and periodically fine-tunes the prediction model online based on historical data and real-time feedback data to improve the accuracy of the prediction model.

[0011] Further, the communication module comprises: a secure transmission unit that adopts an elliptic curve encryption-based communication protocol to ensure the integrity and confidentiality of cross-node data transmission; a chain synchronization unit that is configured to synchronize instruction states among multiple nodes asynchronously, and quickly restore a communication link through a chain broadcast mechanism when a node is abnormal, thereby ensuring the complete execution of the control instructions.

[0012] Further, the distributed control module comprises: a real-time scheduling unit that is configured to dynamically adjust the operating parameters of distributed energy equipment according to the energy optimization strategy, and support each node to autonomously correct the control instruction parameters according to a local state; a fault response unit that is configured to trigger a backup energy scheduling process when a device fault or communication anomaly is detected, and synchronize the fault state to related nodes through a chain broadcast mechanism for collaborative recovery.

[0013] Further, each node in the platform supports local fast data processing and breakpoint resume functions through the local cache node mechanism, thereby reducing the pressure on the central node and improving the overall stability of the system.

[0014] In an embodiment, the present application also provides a distributed energy management method, which comprises the following steps: Step 1: Real-time collection of multi-source heterogeneous data from distributed energy equipment, sensors and external data sources, and synchronization to local cache nodes; Step 2: Preprocessing, standardization and fusion of the multi-source heterogeneous data to form an energy management data set in a unified format, and dynamic updating of the data set based on an incremental synchronization mechanism; Step 3: Analysis of the energy management data set based on a machine learning algorithm to generate the energy optimization strategy and periodically adjust the optimization parameters; Step 4: Sending of the control instructions to distributed energy equipment according to the energy optimization strategy, and execution of fine-tuning optimization based on a local state by each device node to form a master-autonomous collaborative mechanism; Step 5: Recording of the data processing and the control instruction execution process through a blockchain technology, independent maintenance of an interaction log by each node, and realization of traceability and tamper resistance of the data and instruction flow.

[0015] Further, the step 3 further comprises: Step 31: training the energy consumption prediction model based on historical data and real-time feedback data, and real-time fine-tuning the prediction results through incremental learning; Step 32: dynamically calculating the energy optimization scheduling weight according to the following formula:

[0016] wherein, is the scheduling weight of the i-th energy device, is the available energy supply of the device, is the energy conversion efficiency of the device, and N is the total number of energy devices; Step 33: generating the optimization scheduling strategy in real time according to the scheduling weight.

[0017] Further, the method further comprises: Step 6: monitoring the running state of the distributed energy device in real time, and synchronizing to the related node through chain broadcast when an exception is detected; Step 7: in the case of device failure or communication interruption, automatically switching to a backup energy scheduling scheme, and completing the rapid issuance of emergency scheduling instructions through the local cache node.

[0018] The application provides a multi-source data fusion distributed energy management platform and a management method thereof, which can effectively solve the problems of poor real-time performance, insufficient data processing capacity, poor system stability and the like in the data acquisition, fusion, control and scheduling process of the existing distributed energy management system. By introducing a local cache node mechanism in the platform, the burden of the center node can be reduced, and the stability and fault tolerance of the system can be improved; through the joint optimization of the data acquisition module and the data fusion module, multi-source heterogeneous data can be efficiently processed, data preprocessing, standardization, fusion and dynamic updating can be realized, so that the data accuracy and real-time performance of the system are ensured; in addition, through the energy optimization strategy and prediction model based on machine learning, combined with the deep learning algorithm for periodic online fine-tuning, the accuracy and self-adaptive ability of energy consumption prediction and optimization scheduling can be improved; the master-autonomous collaborative scheduling mechanism is adopted, which supports the distributed energy device to execute fine-tuning decisions according to the local state, and ensures that the system can still operate efficiently under fault or abnormal conditions. The introduction of the blockchain technology makes the data and control instructions of the system have traceability and tamper resistance, which guarantees the safety and reliability of the platform operation. Through the application, the intelligent level of the distributed energy management system can be improved, the energy scheduling efficiency can be optimized, and the fault tolerance and sustainable development ability of the system can be enhanced. BRIEF DESCRIPTION OF DRAWINGS

[0019] Figure 1 Fig. 1 is a structural schematic diagram of a multi-source data fusion distributed energy management platform provided by the application; Figure 2 A multi-source data fusion distributed energy management method flow chart is provided for the present application.

[0020] Reference signs: Distributed energy equipment 100, data acquisition module 101, data fusion module 102, data analysis module 103, distributed control module 104, communication module 105.

[0021] The implementation, functional features and advantages of the present application will be further described with reference to the accompanying drawings. DETAILED DESCRIPTION

[0022] In order to make the purpose, technical solutions and advantages of the present application clearer, the technical solutions in the present application will be described clearly and completely below with reference to the drawings in the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0023] The terms "first", "second" are only for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features. Therefore, the features defined with "first", "second" can explicitly or implicitly include one or more of the features; in the description of the present application, unless otherwise specified, the meaning of "multiple" is two or more.

[0024] It should be noted that the term "node" in the specification and claims can include device nodes, data cache nodes, communication coordination nodes or other data processing units with specific functions according to the context, all of which are logical entities that can independently execute instructions or interact in the distributed system of the present application.

[0025] In order to make the technical solutions of the present application clearer, the present application will be described in detail below in combination with specific embodiments, but should not be understood as limiting the scope of protection of the present application.

[0026] In the present embodiment, as shown in Figure 1 A multi-source data fusion distributed energy management platform is provided, which includes data acquisition module 101, data fusion module 102, data analysis module 103, distributed control module 104 and communication module 105, and guarantees the traceability and tamper resistance of system data and control instructions through blockchain technology.

[0027] Specifically, the data acquisition module 101 is configured to acquire multi-source heterogeneous data in real time from a plurality of distributed energy devices 100, various sensors and external data sources, which covers energy production data, consumption data, environmental monitoring data and energy market price data. The collected data is synchronized to the local cache node in a periodic manner. The cache mechanism can establish an intermediate buffer layer between the data acquisition node and the central system, thereby improving the data processing efficiency of the system and reducing the risk of transmission interruption caused by network fluctuations.

[0028] Further, the data fusion module 102 is configured to process the multi-source heterogeneous data collected above. First, data preprocessing is performed to eliminate outliers and incomplete items, and then standardization operations are performed on data from different sources to make them have a unified format and facilitate subsequent processing. On this basis, the fusion module fuses index data from different sources using a set of rules to form a consistent energy management data set. To ensure data timeliness and stability, the module combines the local cache node mechanism to realize the rollback protection and dynamic updating function of data in the updating process, ensuring that the system can maintain normal operation state when facing sudden abnormal data.

[0029] Further, the data analysis module 103 is configured to process the energy management data set in a unified format based on machine learning algorithms, mine key features in the data and build corresponding energy optimization strategies and prediction models. The module supports regular adaptive adjustment of the optimization strategy according to real-time feedback data from the device end and the environment end to adapt to the dynamic changes of the system running state and continuously optimize energy utilization efficiency. The generation of energy optimization strategies considers multiple constraint conditions and aims to minimize cost and optimize scheduling, thereby providing high-quality decision-making basis for subsequent control instructions. Understandably, the prediction model is used to predict key parameters of the future energy system in advance, including but not limited to energy demand, energy supply capacity, carbon emission trend, meteorological environmental change and energy market transaction price, etc. The prediction model takes historical data and real-time feedback data as input, generates short-term or medium-term prediction results through trained machine learning models or deep neural networks, to assist the data analysis module 103 in formulating forward-looking energy optimization strategies.

[0030] Further, the distributed control module 104 is configured to send control instructions generated by the optimization strategy to the distributed energy devices 100. The module supports sending instructions based on the master node while allowing each device node to execute fine-tuning decisions based on its local running state, thereby giving each node certain autonomy on the basis of overall coordinated control of the system, realizing a collaborative scheduling system combining master control and autonomy. This control mode not only guarantees the realization of the unified scheduling target of the platform, but also improves the flexibility and response speed of the system in response to local emergencies.

[0031] Further, the communication module 105 is used to realize data interaction between modules of the platform, and is configured to provide an encrypted data transmission channel to support high-reliability data exchange in an asynchronous environment. The chain instruction confirmation mechanism ensures that each control instruction can be completely tracked and accurately executed, preventing command failure or misfire due to transmission delay or interruption. The communication mechanism also supports interconnection and intercommunication between nodes in different geographical regions, enabling the platform to be widely deployed in energy scenarios with distribution characteristics.

[0032] In addition, the platform integrates blockchain technology to ensure the traceability and tamper resistance of the data and control instructions. Each node in the platform has the ability to record the blockchain interaction log independently, can generate a timestamp during execution and write it into an unchangeable chain structure, ensuring that all key operations of the platform can be reproduced and verified at a later stage, effectively improving the safety and transparency of the platform in the energy management process.

[0033] The embodiment provides a distributed energy management platform for multi-source data fusion and a management method thereof, which can effectively solve the problems of poor real-time performance, insufficient data processing capacity, poor system stability and the like in the data acquisition, fusion, control and scheduling process of the existing distributed energy management system. By introducing a local cache node mechanism in the platform, the burden on the central node can be reduced, and the stability and fault tolerance of the system can be improved; through joint optimization of the data acquisition module 101 and the data fusion module 102, multi-source heterogeneous data can be efficiently processed, data preprocessing, standardization, fusion and dynamic updating can be realized, so that the data accuracy and real-time performance of the system are ensured; in addition, through the energy optimization strategy and prediction model based on machine learning, combined with the deep learning algorithm for periodic online fine-tuning, the accuracy and self-adaptive ability of energy consumption prediction and optimal scheduling can be improved; the master-autonomous collaborative scheduling mechanism is adopted, which supports the distributed energy equipment 100 to execute fine-tuning decisions according to the local state, ensuring that the system can still operate efficiently under fault or abnormal conditions. The introduction of the blockchain technology makes the data and control instructions of the system have traceability and tamper resistance, ensuring the safety and reliability of the platform operation. Through the present application, the intelligent level of the distributed energy management system can be improved, the energy scheduling efficiency can be optimized, and the fault tolerance and sustainable development ability of the system can be enhanced.

[0034] In some embodiments, the data acquisition module 101 further comprises an IoT interface unit and a data verification unit. Specifically, the IoT interface unit is used to realize the communication connection between the platform and various distributed energy devices 100 and sensors, and specifically supports multiple communication protocols including Modbus, Zigbee, LoRa, MQTT, etc., thereby adapting to terminal devices of different manufacturers and technical standards. The interface unit can dynamically adjust the data acquisition frequency according to the current load condition of each node, for example, reducing the acquisition frequency when the device running state is stable to save network and processing resources, and increasing the acquisition frequency when the fluctuation state or emergency occurs to obtain more detailed state information, thereby realizing the dynamic balance between data acquisition efficiency and system resource consumption.

[0035] Further, the data acquisition module 101 also comprises a data verification unit for checking the integrity and consistency of the multi-source heterogeneous data obtained immediately after data acquisition is completed. The integrity check is mainly used to identify missing fields, abnormal timestamps or data format errors in the data, and the consistency check is used to compare the logical matching conditions between different collection sources, such as whether the temperature sensor reading is consistent with the device running state, whether the electric energy collection value is consistent with the device power configuration, etc. If data anomalies are detected, the verification unit will actively trigger the re-sampling mechanism in the local cache node, that is, automatically identify the collection channel and device identifier corresponding to the abnormal data, instruct the collection unit to re-take the original data from the device end, and if necessary, re-execute the data fusion process, to ensure that the data entering the platform core analysis process has high reliability and consistency.

[0036] It can be understood that through the collaborative design of the IoT interface and the data verification mechanism, the data acquisition module 101 of the platform not only has high adaptability and flexibility, but also has strong anti-interference and self-recovery capabilities, thereby providing an accurate and reliable data basis for subsequent energy efficiency analysis and scheduling optimization. This implementation significantly improves the robustness of the platform running in complex environments and effectively reduces the risk of control instruction misjudgment caused by data defects.

[0037] In some embodiments, the data fusion module 102 further comprises a data cleaning unit and a semantic mapping unit. The main function of the data cleaning unit is to handle missing values, duplicates and noisy data in the multi-source heterogeneous data collected from different data sources. Specifically, the data cleaning unit automatically identifies and fills in missing values through a series of pre-set preprocessing rules, for example, using the nearest neighbor algorithm or the mean filling method to complete the missing data of energy equipment; for duplicate data, the system automatically removes duplicates and only keeps the most timely data records. In order to improve the flexibility and adaptability of the system, the data cleaning unit supports real-time adjustment of cleaning rules according to environmental changes, for example, when the data source changes or the data collection frequency is adjusted, the cleaning strategy can automatically adapt to the new data characteristics, ensuring the quality and consistency of the data.

[0038] Further, the data fusion module 102 also includes a semantic mapping unit. The semantic mapping unit maps heterogeneous data from different sources into a unified energy management data model through semantic analysis-based techniques. Specifically, the semantic mapping unit first analyzes the data transmitted by different devices and sensors, and generates corresponding semantic labels according to the data fields and contents, thereby achieving effective classification of different data sources. For example, when mapping the output power data of a battery energy storage device and the output power data of a photovoltaic power generation device, the system recognizes that they belong to the category of "energy production data" and integrates them through a unified model. When new data sources are added, the semantic mapping unit can dynamically handle new data types and data structures through incremental mapping, ensuring that the platform can seamlessly access when new or replaced data sources without interrupting or conflicting with the operation and data flow of the existing system.

[0039] It can be understood that through such flexible cleaning and mapping mechanism, the platform can ensure that the complex information obtained from multiple data sources can be effectively converted into unified and usable energy management data sets, thereby improving the decision analysis accuracy and real-time performance of the system. This implementation can provide high compatibility when facing various types of devices and data formats, significantly reducing the maintenance cost of the system, and effectively avoiding the information processing difficulties caused by data heterogeneity.

[0040] In some embodiments, specifically, the data analysis module 103 includes an optimization strategy unit and a prediction sub-module. In the optimization strategy unit, the system generates energy optimization strategies by using optimization algorithms. Specifically, the optimization algorithm uses an objective function to calculate the energy production cost and carbon emissions of energy consumption in each time period. The optimization objective function is: min

[0041] where T is the time period of optimization, for example, T can be 24 hours, representing an optimization schedule for one day; The energy production cost at time t, specifically, represents the energy allocation amount allocated to each distributed energy device 100 at time t. λ is the environmental weight coefficient, which represents the weight of carbon emissions on the optimization result in the optimization objective, and the value of λ ranges from [0, 1]. The value is dynamically optimized according to the actual execution result to ensure a balance between energy production cost and environmental impact. The energy production cost generated at time t can be in monetary units (e.g., RMB). The carbon emissions of energy consumption at time t, representing the carbon emissions generated by the system at time t according to the given energy allocation scheme The carbon emissions generated at time t are in tons (t / CO2). The energy allocation scheme at time t, specifically, represents the energy allocation amount allocated to each distributed energy device 100 at time t. λ is the environmental weight coefficient, which represents the weight of carbon emissions on the optimization result in the optimization objective, and the value of λ ranges from [0, 1]. The value is dynamically optimized according to the actual execution result to ensure a balance between energy production cost and environmental impact.

[0042] During the optimization process, λ will be dynamically adjusted according to the actual execution result after each cycle. For example, if the carbon emissions are relatively high in a cycle, the value of λ may increase, prompting the system to consider more factors in subsequent optimization to reduce carbon emissions; conversely, if the carbon emissions are low, the value of λ may decrease, so that the optimization focuses more on reducing production costs.

[0043] Further, the prediction submodule trains a prediction model based on a deep learning algorithm, and the training of the model relies on historical data and real-time feedback data. Specifically, the system collects historical energy production, consumption, environmental data and other information, uses deep neural networks (such as LSTM, GRU, etc. Sequence model) for modeling to improve the prediction accuracy of future energy demand and supply. According to historical data, the system can understand the running characteristics of different energy devices, the fluctuation rules of energy supply and the changing trend of market prices.

[0044] In addition, real-time feedback data refers to the latest data obtained from distributed energy devices 100, sensors and external data sources in real time, including but not limited to energy production status, device running status, environmental changes, market prices, etc. The prediction submodule combines these real-time feedback data with historical data and performs online fine-tuning periodically to continuously improve the accuracy of the prediction model. This fine-tuning process is usually performed at fixed time intervals (e.g., every hour or every 12 hours) to ensure that the model can adapt to rapidly changing environments and energy demands.

[0045] It can be understood that in this way, the system can dynamically adjust the energy optimization strategy according to real-time feedback data, so that the energy management system can flexibly respond to changing environments and demands, improve energy utilization efficiency and reduce unnecessary carbon emissions.

[0046] In some embodiments, the communication module 105 includes a secure transmission unit and a chain synchronization unit. Specifically, the secure transmission unit can employ an elliptic curve encryption-based communication protocol to ensure the integrity and confidentiality of data transmission across nodes. In particular, the elliptic curve encryption algorithm (ECC) is an encryption technology based on the mathematical principles of elliptic curves, which has lower computational and storage overhead while ensuring high security, making it suitable for application in resource-constrained distributed energy management systems. The secure transmission unit uses ECC to encrypt the transmitted data during data transmission, ensuring the confidentiality and tamper resistance of the data during transmission. Through ECC encryption, all devices and nodes within the platform can ensure the security of information during transmission without revealing the transmission content. For example, the transmitted data includes energy collection data, control instructions, device operating status information, etc. After encryption, these data can only be decrypted by nodes with decryption authority, thereby avoiding the risk of data leakage and tampering.

[0047] Further, the chain synchronization unit is used to synchronize the instruction state among multiple nodes asynchronously and quickly restore the communication link through the chain broadcast mechanism when a node is abnormal, ensuring the complete execution of control instructions. Specifically, the chain synchronization unit uses the chain broadcast mechanism in blockchain technology to ensure real-time synchronization of instructions among multiple nodes. When multiple distributed energy devices 100 or nodes simultaneously execute control tasks, the chain synchronization unit ensures that each node can update its state within a relatively short time when receiving the latest control instruction, avoiding inconsistent execution problems caused by communication delays or network fluctuations.

[0048] In the specific implementation process, when a node sends a control instruction (such as adjusting the energy distribution strategy or device parameters), the chain synchronization unit will deliver the instruction to other related nodes through the blockchain broadcast mechanism. In this process, all nodes that receive the instruction will automatically update their own state and synchronize the updated state information back to the master node. If any node fails to receive or execute the instruction in time due to communication interruption or failure, the chain synchronization unit will ensure the complete transmission of the instruction through the automatic retransmission mechanism, and through the tamper-proofing property of the blockchain, the history of instruction execution can be traced back.

[0049] In addition, in the case of node abnormalities, such as a node losing connection with the master node, the chain synchronization unit will promptly notify other nodes of the abnormal state through the chain broadcast mechanism, ensuring that the system can respond quickly when a fault occurs and quickly restore the communication link through backup nodes or recovery strategies. This mechanism not only ensures the complete execution of control instructions, but also improves the robustness of the entire distributed energy management platform in the face of network fluctuations or device failures.

[0050] Through such a design of the communication module 105, the system can ensure the security of data transmission across nodes, the synchronization of instruction execution, and the fault tolerance of the system, thereby improving the stability and reliability of the platform in complex environments.

[0051] In some embodiments, the distributed control module 104 includes a real-time scheduling unit and a fault response unit. The real-time scheduling unit is configured to dynamically adjust the operating parameters of the distributed energy devices 100 according to the energy optimization strategy, and to support each node to autonomously correct the control instruction parameters according to the local state.

[0052] Specifically, the working principle of the real-time scheduling unit is based on the energy optimization strategy generated by the platform. The optimization strategy is generated by analyzing multi-dimensional data such as energy production, consumption, market price, and environmental impact, and generating an energy scheduling plan. On this basis, the real-time scheduling unit monitors the operating state of each distributed energy device 100 in real time, such as load, temperature, power generation efficiency, and other parameters. When the system finds that the operating state of a device has changed (for example, the load exceeds the standard, or there are signs of failure), the real-time scheduling unit will dynamically adjust the operating parameters of the device. For example, when the power generation of a photovoltaic power station is lower than expected, the real-time scheduling unit can adjust the operating parameters of other energy devices (such as energy storage batteries or wind turbines) to ensure the stability and optimal operation of the system. Through such flexible and dynamic adjustment, the system can efficiently respond to different working environments and operating states, ensuring the effectiveness of the overall energy management scheme.

[0053] Further, when executing the control instructions, the real-time scheduling unit allows the devices to autonomously correct the control instruction parameters according to the local state. Specifically, some distributed energy devices 100 (such as energy storage batteries) may be subject to local battery power, temperature, or fault detection limitations, and the real-time scheduling unit can allow the devices to autonomously decide whether to adjust or optimize the control parameters based on these limitations. For example, if the power of the energy storage battery is lower than the preset threshold, the device node will autonomously reduce the output power or stop charging to avoid over-discharge, thereby protecting the battery life and system stability. This autonomous correction mechanism can improve the robustness of the entire system and avoid execution errors caused by network delays or instruction transmission problems.

[0054] The fault response unit is configured to trigger the backup energy scheduling process when a device fault or communication anomaly is detected, and synchronize the fault state to relevant nodes through a chain broadcast mechanism for collaborative recovery. Specifically, the fault response unit detects device anomalies by monitoring the operating state of each device in real time, and immediately starts the backup energy scheduling process when an anomaly is detected. For example, when a wind turbine generator set fails, the system automatically adjusts the operating mode of other devices, such as increasing the discharge capacity of energy storage devices or adjusting the output power of other power generation devices, to make up for the energy gap caused by the failure. In addition, the fault response unit synchronizes fault information to all relevant nodes in a timely manner through a chain broadcast mechanism, ensuring that each node can understand the fault state in the first time and start the corresponding recovery program.

[0055] When a fault occurs, the chain broadcast mechanism not only ensures the rapid dissemination of information, but also effectively coordinates the actions of each node to ensure that the entire system can quickly recover to normal operation. For example, when a node cannot be connected, other nodes will synchronize fault information through the blockchain and start the backup power system according to the predetermined recovery process, thereby ensuring the continuity of energy supply. This design based on fault response and chain broadcast mechanism makes the system have stronger fault tolerance and recovery ability, thereby improving the stability of the platform in complex and dynamic environments.

[0056] Overall, through the cooperation of the real-time scheduling unit and the fault response unit, the system can achieve highly automated energy scheduling and fault handling, not only optimizing energy distribution and reducing the probability of system failure, but also quickly taking measures when a fault occurs to ensure the continuous operation of the system and the stability of energy supply.

[0057] In some embodiments, each node supports local fast data processing and breakpoint resume functions through a local cache node mechanism to reduce the pressure on the central node and improve the overall stability of the system. Specifically, the local cache node is a component of each distributed energy device 100 or sensor node, which is used to temporarily store the collected data and realize fast processing and transmission of data through the local cache mechanism.

[0058] Firstly, the local cache node can efficiently perform temporary storage of data. In the system, the distributed energy devices 100 and sensors continuously collect data from different sources, including energy production data, consumption data, environmental data, etc. Due to the high collection frequency of each device node, if all data is directly transmitted to the central node in real time, the central node will be overloaded, and transmission delay or loss problems may occur. In order to alleviate this problem, the local cache node will first temporarily store the collected data locally, and periodically upload the data to the central node according to the predetermined synchronization strategy. This not only reduces the pressure on the central node, but also avoids the problem of data loss caused by network delay or fluctuation.

[0059] Further, the breakpoint resume function of the local cache node can ensure the continuity and integrity of data transmission. When the communication link between the device node and the central node is abnormal, the local cache node will save the already collected data that has not been uploaded, and continue data uploading after the network is restored. Specifically, the local cache node will record the state information of each data upload, including the last successful upload position and time. When the communication link is restored, the cache node can continue to upload data from the interrupted position, without causing repeated uploading or loss of data. This function ensures the accuracy and integrity of the data in an unstable network environment, and improves the data transmission reliability of the system.

[0060] In addition, the local cache node mechanism also supports incremental synchronization of data. In actual operation, each device node of the system may dynamically adjust the data collection frequency or change the data upload time according to the local state (such as load, health state, etc.). Through the incremental synchronization mechanism, the local cache node will only upload the newly added data since the last synchronization, rather than repeatedly uploading all data, thereby reducing the data transmission amount and improving the communication efficiency. Each node also synchronously sends the verification information of the data when uploading the data, to ensure the consistency and integrity of the data.

[0061] The design of the local cache node can also improve the overall stability of the system. In traditional centralized systems, the central node may crash or fail due to receiving too much data or network problems. Through the local cache node mechanism, data is preprocessed and cached at each distributed node, and the central node only needs to receive necessary update data periodically, avoiding performance bottlenecks caused by excessive data volume or system load. Through this distributed cache mechanism, the system can maintain efficient and stable operation when facing large-scale distributed energy devices 100 and a large amount of data collection.

[0062] Understandably, the local cache node mechanism effectively reduces the pressure of the central node, improves the reliability and efficiency of data transmission, and enhances the fault tolerance and stability of the system in complex environments by supporting local data processing, breakpoint resume and incremental synchronization functions.

[0063] The application also provides a distributed energy management method based on the above platform, as shown in the figure, which includes the following steps when implemented: Figure 2 Step 1, real-time collection of multi-source heterogeneous data from distributed energy equipment 100, sensors and external data sources, and synchronization to local cache nodes.

[0064] Specifically, the data collection module 101 will collect data in real time according to the different characteristics of the equipment or sensors using the appropriate communication protocol. Through the Internet of Things interface unit, the platform can flexibly support multiple communication protocols, thereby ensuring that it can connect with various types of distributed energy equipment 100 and sensors, and ensuring the comprehensiveness and accuracy of data collection. The collected data includes but is not limited to energy production data, consumption data, environmental data, market price data, etc. To ensure data integrity, the data collection module 101 will also regularly check the quality of the collected data to ensure that every piece of data transmitted is valid.

[0065] Step 2, pre-processing, standardization and fusion of the multi-source heterogeneous data to form a unified format energy management data set, and dynamic updating of the data set based on the incremental synchronization mechanism.

[0066] Specifically, the data fusion module 102 cleans and standardizes the collected data. The cleaning process includes removing redundant data, filling missing values and filtering outliers to ensure data quality. After standardization, heterogeneous data from different sources can be converted to the same standard format, facilitating subsequent analysis and processing. In addition, the platform ensures real-time updating of the data set through the incremental synchronization mechanism. The local cache node of each node will periodically upload the locally stored data to the data fusion module 102 according to the set synchronization strategy, thereby realizing dynamic updating of the data set and maintaining the timeliness and accuracy of the data.

[0067] Step 3, analysis of the energy management data set based on machine learning algorithms to generate energy optimization strategies and periodically adjust optimization parameters.

[0068] ​Specifically, the data analysis module 103 employs a machine learning model to analyze the acquired historical data and real-time feedback data, predicting future energy demand, supply, and market price fluctuations. By training the machine learning model, the system can predict energy demand and supply in the future, providing data support for the generation of energy optimization strategies. The goal of energy optimization strategies is to maximize the overall energy utilization efficiency of the system and minimize energy consumption and carbon emissions. During the generation of energy optimization strategies, the system will combine historical data and real-time feedback data to periodically adjust the strategies, ensuring that the optimization strategies can adapt to changing environments and demands.

[0069] Step 4, send control instructions to distributed energy devices 100 according to the energy optimization strategy.

[0070] Specifically, the energy optimization strategy generated by the data analysis module 103 will be converted into specific control instructions, instructing the distributed energy devices 100 how to adjust energy production and consumption. Control instructions not only consider the balance of energy supply and demand, but also need to be fine-tuned according to the real-time state and operating conditions of the device. Each distributed energy device 100 has a certain degree of autonomy and can make fine-tuning decisions based on local state and local demand, thereby realizing a master-autonomous collaborative scheduling mechanism.

[0071] Step 5, record the data processing and control instruction execution process through blockchain technology, realize the traceability and tamper resistance of data and instruction flow.

[0072] Specifically, the platform records all collected raw data and executed control instructions through blockchain technology to ensure the authenticity and integrity of the data. Each node maintains a copy of the blockchain independently and stores the data encrypted according to the blockchain protocol to ensure that the data cannot be tampered with. In addition, the blockchain technology also enables the execution process of each piece of data and control instruction to be traced, providing transparent operation records for the system and increasing the credibility and security of the system.

[0073] Through the above implementation steps, the distributed energy management method can effectively manage energy production and consumption in large-scale distributed energy systems, maximize the energy utilization efficiency of the system, and ensure the security and transparency of data and control instructions. In practical applications, this method can be widely used in smart grids, renewable energy systems, and energy internet fields, providing support for green, low-carbon, and sustainable energy management.

[0074] In some embodiments, for the calculation of energy optimization scheduling weights, step 3 includes: Step 31: Train the energy consumption prediction model based on historical data and real-time feedback data, and fine-tune the prediction results in real-time through incremental learning. In order to generate more accurate energy optimization scheduling strategies, the system calculates energy optimization scheduling weights through this step. Specifically, the system will optimize scheduling based on the available energy supply of energy equipment, energy conversion efficiency, and real-time state of the equipment. These scheduling weights are used to guide the scheduling decisions of energy equipment, thereby achieving efficient use of energy.

[0075] Specifically, based on historical data and real-time feedback data, the platform will train an energy consumption prediction model through deep learning algorithms. This prediction model can predict the future trend of energy consumption based on the running state of the equipment, the historical record of energy consumption, and external environmental factors such as temperature and humidity. In this way, the system can predict the future energy demand of the equipment in real time and provide accurate data support for scheduling decisions. Incremental learning can enable the model to continuously update based on newly collected real-time feedback data, thereby continuously improving the accuracy and robustness of the model.

[0076] It can be understood that real-time data refers to the current running state and environmental condition data collected in real time from energy equipment, sensors, and external data sources, such as energy production, consumption, equipment state, and market price; historical data is the collected data in the past period of time, which is used to analyze long-term trends and train prediction models, such as past energy production, consumption records, and environmental change data.

[0077] Specifically, historical data is used to preliminarily train the energy consumption prediction model, while real-time data is used for online fine-tuning and incremental learning of the model. When new real-time data is obtained, the model updates the prediction results in real-time through incremental learning, so as to dynamically adjust and improve the prediction accuracy without retraining the model. This way can ensure that the prediction model always adapts to changes in the environment, improving the timeliness and accuracy of prediction.

[0078] Step 32: Dynamically calculate the scheduling weight of each energy equipment according to the following formula:

[0079] Wherein, the parameters in the formula are defined as follows: : The scheduling weight of the i-th energy equipment. This weight determines the proportion of the device in the energy optimization process, the higher the value, the higher the scheduling priority of the device, and the larger the proportion of energy consumption required. The calculation of the scheduling weight takes into account the energy supply capacity, efficiency, and other factors of the equipment.

[0080] : The available energy supply of the i-th device. This value indicates the amount of energy currently available from the i-th energy device. It is typically calculated dynamically based on the device's output capacity, current load, and remaining energy. For example, the available energy supply of a solar panel may be affected by weather and sunshine duration.

[0081] : Energy conversion efficiency of the i-th device. This value reflects the efficiency of the energy device in converting input energy into usable energy and is generally determined by the device's design and technical characteristics. For some renewable energy devices, such as solar panels and wind turbines, efficiency can be affected by external factors such as weather.

[0082] : The available energy supply of the jth device in the system. similar, Indicates the energy supply of other devices. The value reflects the energy supply capacity of each device in the system. The value has an impact on the total weight calculation.

[0083] : Energy conversion efficiency of the jth device. resemblance, Describes the efficiency of other equipment and affects the energy scheduling weight calculation of the entire system.

[0084] N: There are N energy devices in the system. This value is used to calculate the sum of the weights of all devices and to normalize the scheduling strategy of the entire system.

[0085] Furthermore, the calculation of scheduling weights also takes into account changes in real-time feedback data. For example, the operating efficiency and load changes of the equipment will affect its energy conversion efficiency, so the system will dynamically adjust the current status of each device in real-time scheduling. In addition, external factors such as market price fluctuations or weather changes can also affect the energy supply capacity of the equipment, thereby affecting The value of .

[0086] On this basis, as in step 33, the system generates an optimized scheduling strategy for each energy device by updating the scheduling weights in real time. This scheduling strategy takes into account various factors, such as the device's energy supply capacity, conversion efficiency, and operating status, to optimize the overall energy system's operating efficiency. For example, if a device has a large available energy supply and high energy conversion efficiency, the system will prioritize scheduling that device to maximize energy utilization.

[0087] The generated optimized scheduling strategy is periodically fine-tuned based on real-time feedback and device status, ensuring that the system always operates in an optimal state. In this way, the system can effectively respond to different operating conditions and environmental changes, maintaining the stability and efficiency of energy supply.

[0088] In practical applications, this method can be widely applied in various distributed energy systems, such as solar power generation, wind power generation, energy storage devices, and smart grids, providing strong support for sustainable energy management.

[0089] In some embodiments, the distributed energy management method further includes step 6: Real-time monitoring of the operating state of the distributed energy device 100, and synchronizing to the relevant nodes through chain broadcast when an anomaly is detected.

[0090] And step 7: In the case of device failure or communication interruption, the platform automatically switches to a backup energy scheduling scheme, and completes the rapid issuance of emergency scheduling instructions through the local cache node.

[0091] Specifically, the platform obtains the operating data of each distributed energy device 100 in real time through the data acquisition module 101, including the output power of the device, the battery capacity, the device state, the communication signal strength, and other key information. The data acquisition module 101 transmits these data to the data fusion module 102 for preprocessing and standardization, forming a unified format of energy management data set. The platform continuously monitors and analyzes these data, and evaluates the working state of each device in real time.

[0092] During the monitoring process, if an abnormality occurs in a device (such as a decrease in output power, loss of communication, or device failure), the data analysis module 103 will trigger an alarm and send a fault signal to the relevant nodes. To ensure stable operation of the system, the system synchronizes the device abnormality information to all nodes in the network through the chain broadcast mechanism, ensuring timely transmission and sharing of fault information.

[0093] For example, when a wind power generation device fails, the system will immediately calculate the impact of the device and generate a backup energy scheduling scheme. The backup scheme is based on other energy devices in the system (such as solar panels, energy storage devices, etc.) to supplement the missing energy supply. The system will adjust the device operation according to the priority of the backup scheme, and quickly issue scheduling instructions to each device through the local cache node, ensuring that the system can switch to the backup energy scheme in the shortest time.

[0094] Through the chain broadcast mechanism, the system can quickly recover the communication link after a fault occurs, ensuring that each node in the system can synchronize the processing of device state changes and respond in a timely manner. Specifically, when a device fails and is transmitted to other nodes through the broadcast mechanism, the distributed control module 104 of other nodes will adjust its local scheduling strategy according to the received fault information, selectively allocate resources of standby energy devices, ensure the stable energy supply of the entire system, and avoid global impact caused by single-point failure.

[0095] For example, if a node (such as a wind turbine) loses power output due to device failure, other nodes (such as adjacent energy storage devices) in the system will automatically allocate more energy output according to the pre-set standby scheduling rules to ensure continuous power supply. This process is coordinated among multiple nodes to ensure seamless transition to the standby solution.

[0096] At the same time, the role of the local cache node is also very critical, it not only stores the real-time data of each node, but also provides fast data processing and instruction issuing function. When the system detects device failure or other emergencies, the local cache node will immediately play a role, issue emergency scheduling instructions to the affected devices, and ensure the integrity and consistency of the data. The cache node plays a role in relieving the pressure of the central node in this process, and improves the stability and fault tolerance of the system.

[0097] This design ensures that the energy management platform can respond in a timely manner when facing device failure or communication interruption, automatically switch to the standby energy scheduling scheme, and ensure the continuous and stable operation of the energy system.

[0098] The above-described is only an exemplary embodiment of the present application, and does not limit the patent scope of the present application, any equivalent structural transformation made under the technical concept of the present application, or direct / indirect application in other related technical fields is included in the patent protection scope of the present application.

Claims

1. A distributed energy management platform with multi-source data fusion, characterized by: include: A data acquisition module is used to collect multi-source heterogeneous data from distributed energy devices, sensors, and external data sources in real time. The multi-source heterogeneous data includes energy production data, consumption data, environmental data, and market price data, and periodically synchronize the collected results to the local cache node; a data fusion module configured to preprocess, standardize, and fuse the multi-source heterogeneous data to generate an energy management data set in a unified format, and implement dynamic update and rollback protection of the data based on the local cache mechanism; A data analysis module that analyzes the energy management data set based on a machine learning algorithm to generate an energy optimization strategy and a prediction model, and regularly and adaptively adjusts the energy optimization strategy based on real-time feedback data; A distributed control module is used to send control instructions to the distributed energy devices according to the energy optimization strategy, and supports each device node to make fine-tuning decisions based on local status, forming a master-control-autonomous collaborative scheduling system; A communication module, configured to provide a secure data transmission channel between the data acquisition module, the data fusion module, the data analysis module, and the distributed control module, and support asynchronous communication and chain instruction confirmation across regional nodes; The platform uses blockchain technology to achieve traceability and tamper-proofing of the data and control instructions, and each node has the ability to independently record and verify blockchain interaction logs.

2. The distributed energy management platform according to claim 1, characterized in that: The data acquisition module includes: An Internet of Things interface unit, which supports multiple communication protocols to adapt to the distributed energy equipment and the sensors, and dynamically switches the data collection frequency according to the load of each node; The data verification unit is configured to perform integrity and consistency checks on the multi-source heterogeneous data and trigger the local cache node re-collection mechanism when abnormal data is detected to improve data collection reliability.

3. The distributed energy management platform according to claim 1, characterized in that: The data fusion module includes: A data cleaning unit, configured to process missing values, duplicates, and noise in the multi-source heterogeneous data and support real-time adjustment of cleaning rules based on environmental changes; The semantic mapping unit is configured to map data from different sources to a unified energy management data model based on semantic analysis, and to achieve seamless integration through incremental mapping when new data sources are added.

4. The distributed energy management platform according to claim 1, characterized in that: The data analysis module includes: The optimization strategy unit generates the energy optimization strategy using an optimization algorithm. The energy optimization strategy is calculated using the following objective function: min ; Among them, T is the optimization time period, is the energy production cost at time t, Carbon emissions from energy consumption, is the energy allocation plan at time t, It is the environmental weight coefficient, and is dynamically adjusted according to the actual execution results after each cycle. ; The prediction submodule trains a prediction model based on a deep learning algorithm and performs periodic online fine-tuning based on historical data and the real-time feedback data to improve the accuracy of the prediction model.

5. The distributed energy management platform according to claim 1, characterized in that: The communication module includes: Secure transmission unit, which uses a communication protocol based on elliptic curve encryption to ensure the integrity and confidentiality of data transmission across nodes; The chain synchronization unit is used to asynchronously synchronize the instruction status among multiple nodes, and when a node is abnormal, quickly restore the communication link through the chain broadcast mechanism to ensure the complete execution of the control instruction.

6. The distributed energy management platform according to claim 1, characterized in that: The distributed control module includes: A real-time scheduling unit, configured to dynamically adjust the operating parameters of the distributed energy equipment according to the energy optimization strategy, and support each node to autonomously modify the control instruction parameters according to the local state; The fault response unit is configured to trigger the backup energy dispatch process when a device failure or communication anomaly is detected, and synchronize the fault status to related nodes through a chain broadcast mechanism for coordinated recovery.

7. The distributed energy management platform according to claim 1, characterized in that: Each node in the platform supports local fast data processing and breakpoint resume functions through the local cache node mechanism, so as to reduce the pressure on the central node and improve the overall stability of the system.

8. A distributed energy management method based on the platform of claim 1, characterized in that: The following steps are involved: Step 1: Collect the multi-source heterogeneous data from the distributed energy equipment, sensors and external data sources in real time, and synchronize them to the local cache node; Step 2: Preprocess, standardize, and fuse the multi-source heterogeneous data to form an energy management data set in a unified format, and dynamically update the data set based on an incremental synchronization mechanism; Step 3: Analyze the energy management data set based on a machine learning algorithm, generate the energy optimization strategy, and periodically adjust the optimization parameters; Step 4: Send the control instructions to the distributed energy devices according to the energy optimization strategy. Each device node can perform fine-tuning optimization based on the local state to form a master control-autonomy collaborative mechanism; Step 5: The data processing and control instruction execution process are recorded through blockchain technology, and each node independently maintains an interaction log to achieve traceability and tamper-proofing of the data and instruction flow.

9. The distributed energy management method according to claim 8, characterized in that: The step 3 comprises: Step 31: training the energy consumption prediction model based on the historical data and the real-time feedback data, and fine-tuning the prediction results in real time through incremental learning; Step 32: Dynamically calculate the energy optimization scheduling weight according to the following formula: ； in, is the dispatch weight of the i-th energy equipment, is the available energy supply for the device, is the energy conversion efficiency of the equipment, N is the total number of energy devices; Step 33: Generate the optimized scheduling strategy in real time according to the scheduling weight.

10. The distributed energy management method according to claim 8 or 9, characterized in that: The method further comprises: Step 6: Monitor the operating status of the distributed energy equipment in real time, and synchronize to relevant nodes through chain broadcast when an abnormality is detected; Step 7: In the event of equipment failure or communication interruption, automatically switch to the backup energy scheduling plan, and quickly issue the emergency scheduling instructions through the local cache node.

Images