Safety monitoring system and method for grid-connected photovoltaic power generation energy storage equipment

Abstract

The invention provides a safety monitoring system and method for grid-connected photovoltaic power generation and energy storage equipment, and relates to the technical field of power grid monitoring, and the system architecture comprises a data collection layer, a data transmission layer, a data storage and management layer, a data analysis and decision-making layer and a safety protection and emergency response layer. A multi-physical field sensor network widely monitors multiple parameters, a data analysis and decision-making layer deeply mines data, performance degradation prediction is carried out from microcosmic to macroscopic, a quantum encryption communication link and a distributed data transmission network respectively ensure safety and high efficiency of data transmission, block chain data storage ensures that data is true, credible and traceable and verified, and the system is applicable to multiple scenes. A safety system is comprehensively constructed, data leakage is prevented, and the safety protection and emergency response layer rapidly isolates and controls energy when a fault occurs, automatically triggers multiple measures and provides information, so that the fault influence is reduced, continuous power supply is guaranteed, the emergency response capability and toughness of the system are improved, and stable operation is maintained.

Description

Technical Field

[0001] This invention relates to the field of power grid monitoring technology, specifically to a safety monitoring system and method for grid-connected photovoltaic power generation and energy storage equipment. Background Technology

[0002] According to the Chinese patent application CN113346625B, a comprehensive power quality monitoring and dispatching system for distributed photovoltaic grid-connected distribution networks includes multiple grid-connected branch lines. Each grid-connected branch line has multiple distributed photovoltaic grid-connected terminals and multiple user terminals. Grid-connected equipment is installed at each distributed photovoltaic grid-connected terminal, and power distribution equipment is installed at each grid-connected branch line terminal for connection between grid-connected branch lines and power transmission. The dispatching platform is connected to the grid-connected equipment, user terminals, and power distribution equipment respectively. The dispatching platform includes a data acquisition module, a configuration module, a dispatching module, and a power quality monitoring module. It can also provide power dispatching capabilities by monitoring the power quality of the grid-connected branch lines and the power consumption of the user terminals under the grid-connected branch lines.

[0003] The aforementioned patent documents and prior art have the following technical problems when used:

[0004] Problem 1: Existing systems often use a single type or a limited number of sensors, making it difficult to comprehensively acquire multi-dimensional operational data of energy storage devices at both the micro and macro levels. Data processing is mostly limited to simple threshold judgments or basic statistical analysis, lacking the ability to deeply mine the inherent physical relationships between data and conduct cross-scale correlation analysis. This makes it difficult to formulate accurate predictive maintenance strategies, resulting in frequent sudden equipment failures, high operation and maintenance costs, shortened equipment lifespan, low operating efficiency, and impact on the stability of power supply.

[0005] Problem 2: Traditional data transmission is vulnerable to hacker attacks, encryption technology is fragile in the face of increasingly complex quantum attack methods, data storage lacks effective anti-tampering mechanisms, data authenticity and credibility are difficult to guarantee, and data traceability and verification are not easy to operate, resulting in insufficient system data security.

[0006] Thirdly, existing energy storage devices lack effective active isolation mechanisms when faults occur. Fault energy can easily spread, triggering chain reactions and causing wider equipment damage and system shutdowns. Emergency response relies heavily on human experience and manual operation, resulting in slow response speeds, incomplete and inaccurate fault information acquisition, long fault handling times, and significant economic losses and social impact caused by power outages. Summary of the Invention

[0007] To address the shortcomings of existing technologies, this invention provides a safety monitoring system for grid-connected photovoltaic power generation and energy storage equipment, which solves the problems of limited data acquisition and analysis and weak encryption technology protection.

[0008] Another objective of this invention is to provide a safety monitoring method for grid-connected photovoltaic power generation and energy storage equipment.

[0009] To achieve the above objectives, the present invention provides the following technical solution:

[0010] A safety monitoring system for grid-connected photovoltaic power generation and energy storage equipment includes,

[0011] The data acquisition layer is used to acquire operational status data, microscopic physical quantity data, and multi-field parameter data from multiple sources, and to perform preliminary processing on the acquired data.

[0012] The data transmission layer is used for quantum encryption of the acquired and processed data and to optimize the data transmission path.

[0013] The data storage and management layer is used to store the data collected above and integrate the multi-source data transmitted above.

[0014] The data analysis and decision-making layer is used to analyze the physical correlations of the above multi-source data, predict trends, and make decisions.

[0015] A safety monitoring method for grid-connected photovoltaic power generation and energy storage equipment includes the following steps:

[0016] The system acquires operational status data, microscopic physical quantity data, and multi-field parameter data from multiple sources and performs preliminary processing on the acquired data.

[0017] The collected and processed data is quantum encrypted and the data transmission path is optimized.

[0018] Store the collected data and integrate the multi-source data transmitted above;

[0019] Analyze the physical correlations of the above multi-source data and predict trends to make decisions.

[0020] The present invention has the following beneficial effects:

[0021] 1. This invention integrates quantum sensing, multi-physics sensor networks, and IoT edge computing nodes in the system data acquisition layer. The quantum sensing module accurately detects microscopic physical quantities, the multi-physics sensor network comprehensively covers multi-parameter monitoring, and the edge computing nodes use Kalman filtering to preprocess data, providing rich and reliable data sources for subsequent processing. The data analysis and decision-making layer deeply mines data value through cross-scale multi-physics coupling analysis, spatiotemporal dimension expansion analysis engines, and predictive maintenance decision-making systems. It correlates microscopic structural changes with macroscopic performance degradation predictions and formulates optimal maintenance strategies based on Q-learning algorithms. This allows for precise location of potential equipment problems, advance maintenance planning, significantly reduced probability of sudden equipment failures, extended equipment lifespan, reduced maintenance manpower and material costs, and significantly improved equipment operating efficiency and reliability. This ensures the continuous and stable operation of grid-connected photovoltaic power generation and energy storage equipment, providing a solid guarantee for power supply.

[0022] 2. This invention employs a system data transmission layer. The quantum encrypted communication link module utilizes quantum encryption technology to construct a secure channel. Its quantum key distribution technology and high-frequency key update mechanism effectively resist quantum and traditional attacks, ensuring data confidentiality and integrity. The distributed data transmission network module uses routing algorithms to achieve efficient and stable transmission, quickly converges during network changes, and possesses load balancing capabilities. The blockchain data storage module in the data storage and management layer leverages the immutability and encrypted storage characteristics of blockchain, with each node fully backing up data. This not only ensures data authenticity and credibility but also provides reliable evidence for data traceability, verification, and electricity market transactions. This comprehensive data security system effectively prevents data leakage, tampering, and loss, protects the security of sensitive system information, safeguards the company's operational security and reputation, and enhances competitiveness and trustworthiness in the electricity market.

[0023] 3. This invention employs an adaptive energy shield-type fault isolation system module for system safety protection and emergency response layers. This module can rapidly activate upon the occurrence of a fault, isolating the fault area through electromagnetic shielding and energy management technologies, controlling energy diffusion, and preventing the fault from escalating and causing a chain reaction on surrounding equipment and the entire system, thus reducing downtime risk and maintenance difficulty. The emergency response mechanism module, based on fault tree analysis algorithms, automatically triggers various emergency measures when a fault or risk is predicted, such as cutting off power to the faulty equipment, activating backup energy storage devices, and providing maintenance personnel with accurate fault information and response suggestions, enabling rapid fault handling. This mechanism effectively reduces the impact of faults on grid-connected photovoltaic power generation systems, ensures a continuous and stable power supply, reduces economic losses and social impact caused by power outages, improves the system's ability and resilience to cope with emergencies, and ensures stable operation in complex and ever-changing power environments. Attached Figure Description

[0024] Figure 1 This is a system architecture diagram of the present invention;

[0025] Figure 2 This is a diagram illustrating the system operation steps of the present invention;

[0026] Figure 3 This is a visualization of the Kalman filtering algorithm of the present invention;

[0027] Figure 4 This is a schematic diagram of the network topology and shortest path of the present invention;

[0028] Figure 5 This is a schematic diagram comparing the base selection and measurement results of the present invention;

[0029] Figure 6 This is a line graph showing the data from each sensor and the fused data in this invention.

[0030] Figure 7 Line graphs showing the original time series and predicted sequence of this invention;

[0031] Figure 8 This is a heatmap visualization of the action value function of the present invention;

[0032] Figure 9 This is a line graph showing the impact of changes in the basic event probability on the top-level event probability in this invention.

[0033] Figure 10 This is a line graph showing the fault energy and remaining energy of the present invention. Detailed Implementation

[0034] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Specific Implementation Example 1

[0036] like Figure 1-10 As shown, a safety monitoring system for grid-connected photovoltaic power generation and energy storage equipment has an architecture including a data acquisition layer, a data transmission layer, a data storage and management layer, a data analysis and decision-making layer, and a safety protection and emergency response layer, wherein:

[0037] The data acquisition layer is used to sense the operating status of multi-source sensing devices and to perform preliminary data processing. The data acquisition layer includes a quantum sensing module, a multi-physics sensor network module, and an IoT edge computing node module. The quantum sensing module accurately captures microscopic physical quantities, the multi-physics sensor network module comprehensively monitors multiple field parameters, and the IoT edge computing node module uses the Kalman filter algorithm to preprocess the data, providing a rich and reliable source of raw information for subsequent monitoring.

[0038] The quantum sensing module is used to deploy quantum sensors such as quantum magnetometers and quantum gyroscopes to accurately measure key physical quantities such as internal current distribution, magnetic field changes, and device attitude information of energy storage devices, providing microscopic data support for device status assessment. The multiphysics sensor network module is used to install various traditional high-precision sensors to monitor multiphysics parameters such as electric field, magnetic field, thermal field, and stress field, providing a data foundation for cross-scale multiphysics coupling analysis. The IoT edge computing node module is used to connect the above sensors, and uses the Kalman filter algorithm to perform preliminary screening, cleaning, and simple analysis on the collected raw data, and to make preliminary judgments and marks on abnormal data, reducing invalid data transmission, reducing data transmission latency, and improving the real-time monitoring performance of the system. The Kalman filter algorithm effectively integrates multi-sensor data and predicts the system state at the next moment by making the optimal estimate of the system state, and is suitable for processing dynamic system data with uncertainty.

[0039] In the quantum sensing module, the measurement accuracy of the quantum magnetometer should be able to distinguish magnetic field changes on the order of nanoteslas, and the angle measurement resolution of the quantum gyroscope should be better than 0.01 degrees to ensure the high accuracy of the collected microscopic physical quantity data and provide a reliable basis for subsequent accurate equipment status assessment. In the multiphysics sensor network module, the sampling frequency of each sensor should not be less than 1kHz, and it should be able to collect parameters such as electric field, magnetic field, thermal field, and stress field in real time and continuously. The sensors should also have an automatic calibration function, which can automatically calibrate within a preset period or when environmental conditions change beyond a threshold to ensure data accuracy. In the IoT edge computing node module, the parameters of its Kalman filter algorithm should be dynamically adjusted according to the operating characteristics of the energy storage device and the range of data fluctuations, with an adjustment cycle of no more than 1 hour, to adapt to the data processing needs under different operating conditions of the device and maintain the effectiveness and stability of data preprocessing.

[0040] The data transmission layer is used to ensure secure and fast data transmission. The data transmission layer includes a quantum encrypted communication link module and a distributed data transmission network module. The quantum encrypted communication link module ensures confidentiality with quantum encryption, and the distributed data transmission network module optimizes the path with routing algorithms. The two work together to enable data to flow stably and efficiently between nodes such as the acquisition end and the processing center, resisting external attacks and transmission interference.

[0041] The quantum-encrypted communication link module establishes a quantum-encrypted communication channel from the data acquisition end (edge ​​computing node) to the data processing center. It uses quantum encryption technology to encrypt transmitted data, ensuring security during remote data transmission, resisting hacker eavesdropping and attacks, and protecting the confidentiality and integrity of energy storage device operation data. The distributed data transmission network module combines traditional wired and wireless network technologies to build a reliable distributed data transmission network. It employs the OSPF (Open Shortest Path First) routing algorithm to optimize data transmission paths, improving network transmission efficiency and reliability. It dynamically selects the optimal data transmission path based on network topology and link state information, reducing transmission latency and data packet loss. To ensure stable and efficient data transmission to the data processing center or other relevant monitoring nodes, the quantum-encrypted communication link module should use quantum key distribution technology to generate encryption keys, with a key update frequency of no less than 1 minute. The encryption algorithm should comply with international quantum encryption security standards to ensure data confidentiality and integrity even in the face of complex quantum attacks. The distributed data transmission network module should have a reconvergence time of less than 10 seconds when the network topology changes or the link fails, ensuring continuous and efficient data transmission. Furthermore, the network should have load balancing capabilities, automatically allocating data transmission tasks based on the traffic and processing capacity of each node.

[0042] The data storage and management layer is used for reliable storage and effective integration of data. It includes a blockchain data storage module and a data fusion and management center module. The blockchain data storage module uses its characteristics to ensure the authenticity and trustworthiness of the data. The data fusion and management center module uses a weighted average method to fuse multi-source data. After standardized storage and indexing, it provides a basis for data traceability and verification, and also lays a solid data foundation for analysis and decision-making.

[0043] The blockchain data storage module encrypts and stores the collected key operational data of energy storage equipment in the blockchain network. Each data node stores a complete copy of the data record. The immutability of the blockchain ensures the authenticity and credibility of the data, providing a reliable basis for data verification in scenarios such as equipment operation history tracing, maintenance record querying, and electricity market transactions. The data fusion and management center module receives data from the data transmission layer and integrates and manages it. It uses the weighted average method in the data fusion algorithm to initially fuse quantum sensing data, multiphysics field data, and data from other sources. It assigns corresponding weights according to the reliability and importance of different sensor data to obtain more accurate fused data results. The data is stored in a unified data format and standard, and a data index and query mechanism are established to facilitate subsequent data retrieval and analysis.

[0044] The data analysis and decision-making layer is used to deeply mine the value of data to make scientific decisions. The data analysis and decision-making layer includes a cross-scale multi-physics coupling analysis module, a spatiotemporal dimension expansion analysis engine module, and a predictive maintenance decision-making system module. The cross-scale multi-physics coupling analysis module uses finite element analysis algorithms to analyze physical relationships, the spatiotemporal dimension expansion analysis engine module uses time series algorithms to predict trends, and the predictive maintenance decision-making system module uses Q-learning algorithms to plan maintenance strategies, comprehensively improving the level of intelligent equipment operation and maintenance.

[0045] The cross-scale multiphysics coupling analysis module uses finite element analysis algorithms to perform cross-scale analysis on fused multiphysics data. It discretizes complex physical systems into a finite number of elements for solution, establishing a model linking microscopic particle behavior and macroscopic device performance. This allows for in-depth analysis of the underlying physical laws within the data, predicting potential macroscopic performance degradation or failure risks due to microstructural changes, such as thermal runaway caused by microscopic defects in battery materials or short-circuit faults caused by electromagnetic stress concentration within power electronic devices. The spatiotemporal dimension expansion analysis engine module integrates historical data, current real-time data, and future data predicted based on external information such as meteorological data. It constructs a spatiotemporal dimension expansion analysis model, employing the ARIMA autoregressive moving average algorithm from time series analysis to analyze historical data, extracting trends and seasonality. This data is then combined with current data for short-term predictions, while external environmental data is used for long-term trend correction. The predictive maintenance decision system module forecasts the operational trends of equipment under different environmental conditions over a future period. For example, it combines weather forecasts to predict the impact of changes in light intensity on the charging and discharging of energy storage equipment. This allows for the advance adjustment of equipment operating parameters and maintenance plans, ensuring the safe and stable operation of the equipment under various temporal and spatial conditions. Based on the above analysis results, the predictive maintenance decision system uses Q-learning reinforcement learning algorithm technology to formulate predictive maintenance decision strategies. The Q-learning algorithm continuously explores and learns in the environment through an intelligent agent, taking corresponding actions (maintenance operations) according to different states (equipment operating states). It updates the action value function based on the rewards obtained (such as improved equipment reliability, reduced costs, etc.) to find the optimal maintenance plan, including maintenance time, maintenance content, and required resources. For example, it determines when to replace battery modules or adjust power conversion device parameters to achieve a balance between equipment reliability and maintenance costs, thereby improving the overall life-cycle benefits of the equipment.

[0046] The safety protection and emergency response layer is used to ensure equipment safety and emergency handling. The safety protection and emergency response layer includes an adaptive energy shield fault isolation system module and an emergency response mechanism module. The adaptive energy shield fault isolation system module isolates energy and reduces hazards in the event of a fault. The emergency response mechanism module uses fault tree analysis algorithms to formulate strategies and automatically or manually handle faults quickly to ensure stable system operation and safety.

[0047] The adaptive energy shield fault isolation system module is used to deploy an adaptive energy shield fault isolation system around energy storage equipment. When a fault is detected, such as abnormal energy fluctuations caused by thermal runaway of a battery cell or short circuit, the system is quickly activated. It uses electromagnetic shielding, energy absorption and conversion technologies to isolate the fault area, prevent the fault from spreading, and convert and store or safely release excess energy, reducing the damage of the fault and the risk of system downtime. The emergency response mechanism module is used to establish a comprehensive emergency response mechanism. When the data analysis layer predicts a serious fault or safety risk, it promptly triggers corresponding emergency measures. It uses the fault tree analysis (FTA) algorithm to analyze the emergency situation, build a fault tree model, and determine the possible causes and propagation paths of the fault, so as to formulate more targeted emergency response strategies and improve the efficiency and accuracy of emergency handling. For example, it can automatically disconnect the power connection of the faulty equipment, start the backup energy storage equipment, or send an emergency alarm to the operation and maintenance personnel and provide detailed fault information and response suggestions to ensure that the fault is handled in the shortest possible time and to ensure the stable operation of the entire grid-connected photovoltaic power generation system. Specific Implementation Example 2

[0049] like Figure 1-10 As shown, based on the content of the above specific embodiments, the following content is further disclosed.

[0050] A method for safety monitoring of grid-connected photovoltaic power generation and energy storage equipment includes the following steps:

[0051] The system collects operational status data, microscopic physical quantity data, and multi-field parameter data from multiple sources and performs preliminary processing on the collected data. It establishes a quantum-encrypted communication channel from the data acquisition end to the data processing center and uses quantum encryption to encrypt the transmitted data. It constructs a distributed data transmission network by combining wired and wireless networks, and optimizes the data transmission path using the OSPF (Open Shortest Path First) routing algorithm. The system dynamically selects the best data transmission path based on the network topology and link status information.

[0052] The data collected and processed above is quantum encrypted and the data transmission path is optimized; the internal current distribution, magnetic field changes, and device attitude information of the energy storage device are collected; multi-physics field parameters such as electric field, magnetic field, thermal field, and stress field are collected; the Kalman filter algorithm is used to perform preliminary screening, cleaning, and analysis of the raw data collected above, and abnormal data is preliminarily judged and marked;

[0053] The system stores the collected data and integrates the multi-source data transmitted above; it encrypts the collected energy storage device operation data and stores it in the blockchain network, with each data node keeping a complete copy of the data record; it receives data from the data transmission layer and integrates and manages it, using the weighted average method in the data fusion algorithm to fuse quantum sensing data, multiphysics field data, and data from other sources, assigning corresponding weights according to the reliability and importance of different sensor data, storing it according to a unified data format and standard, and establishing a data index and query mechanism.

[0054] This study analyzes the physical correlations and predicts trends of the multi-source data to formulate decisions. It employs finite element analysis (FEM) algorithms to perform cross-scale analysis on the fused multiphysics data, discretizing complex physical systems into a finite number of elements for solution. By establishing a model linking microscopic particle behavior and macroscopic equipment performance, it uncovers physical laws in the data and predicts macroscopic performance degradation or failure risks caused by changes in microstructure. Furthermore, it integrates historical data, current real-time data, and future data predicted based on external information to construct a spatiotemporally expanded analytical model. The ARIMA (Autoregressive Moving Average) time-series analysis algorithm is used to analyze historical data, extracting data trends and seasonal characteristics. This data is combined with current data for short-term predictions, while external environmental data is used for long-term trend correction, predicting the equipment's operating trends under different environmental conditions in the future and adjusting equipment operating parameters and maintenance plans in advance. Based on these analysis results, a Q-learning reinforcement learning algorithm is used to intelligently formulate predictive maintenance decision strategies. The Q-learning algorithm allows the agent to continuously explore and learn in the environment, taking corresponding actions based on different states and updating the action value function based on the obtained rewards, thereby finding the optimal maintenance plan.

[0055] An adaptive energy shield fault isolation system is deployed around the energy storage device. When a fault is detected, the fault area is isolated by electromagnetic shielding, energy absorption and conversion to prevent the fault from spreading and to convert and store or safely release excess energy. An emergency response mechanism is established. When the data analysis layer predicts a fault or safety risk, corresponding emergency measures are triggered in a timely manner. The fault tree analysis (FTA) algorithm is used to analyze the emergency situation, build a fault tree model, determine the possible causes and propagation paths of the fault, and formulate emergency response strategies.

[0056] The core of the data acquisition layer operation: The quantum magnetometer and quantum gyroscope of the quantum sensing module accurately capture microscopic physical quantities, such as magnetic field changes and device attitude information, with extremely high precision. The multi-physics sensor network module monitors multi-physics parameters at a sampling frequency of no less than 1kHz, and the sensors have automatic calibration functions. The IoT edge computing node module uses the Kalman filter algorithm to dynamically adjust parameters, screen, clean and preliminarily analyze the raw data, reduce invalid data transmission, and provide a reliable source of raw information for subsequent processing.

[0057] Key features of the data transmission layer: The quantum encrypted communication link module adopts quantum key distribution technology, updates keys at a high frequency, and encrypts according to international standards to ensure data confidentiality and integrity. The distributed data transmission network module has a fast routing algorithm convergence when the network changes, and has a load balancing function. It combines traditional network technology to build the network, so that data flows stably and efficiently through each node.

[0058] The key points of data storage and management are: the blockchain data storage module uses the characteristics of blockchain to encrypt and store key data on the network, with each node storing a copy to ensure the authenticity and trustworthiness of the data; the data fusion and management center module receives data, uses a weighted average method to fuse multi-source data, processes it according to weights, stores it according to standards, and builds an index query mechanism to lay the foundation for analysis and decision-making.

[0059] The core tasks of the data analysis and decision-making layer are as follows: The cross-scale multiphysics coupling analysis module uses finite element analysis algorithms to analyze multiphysics data, explore the micro- and macro-level connections, and predict faults; the spatiotemporal dimension expansion analysis engine module integrates multiple types of data and uses time series algorithms to predict the operating trends of equipment in different environments; and the predictive maintenance decision system module formulates maintenance strategies based on Q-learning algorithms to balance equipment reliability and cost and improve the intelligence of operation and maintenance.

[0060] The main functions of the safety protection and emergency response layer are as follows: When a fault occurs, the adaptive energy shield fault isolation system module uses technologies such as electromagnetic shielding to isolate energy, convert or release excess energy, and reduce harm. The emergency response mechanism module uses fault tree analysis algorithms to build a model to determine the cause and path of the fault, and triggers emergency measures such as cutting off the power supply and starting backup equipment to ensure the stability and safety of the system.

[0061] The role of the data acquisition layer: The quantum magnetometer and quantum gyroscope in the quantum sensing module can accurately measure microscopic physical quantities, providing high-precision microscopic data support for equipment status assessment. This helps to detect potential faults caused by changes in microstructure in advance. The multiphysics sensor network module can collect parameters such as electric field, magnetic field, thermal field, and stress field in real time and continuously, and has an automatic calibration function to ensure data accuracy. It provides a comprehensive data foundation for cross-scale multiphysics coupling analysis, enabling a more comprehensive understanding of the equipment's operating status. The IoT edge computing node module uses the Kalman filter algorithm to process data and can perform preliminary judgment and marking of abnormal data. It can also dynamically adjust algorithm parameters according to the equipment's operating characteristics, reducing invalid data transmission, lowering data transmission latency, improving the system's real-time monitoring performance, and reducing the burden on subsequent data processing.

[0062] The role of the data transmission layer: The quantum encrypted communication link module uses quantum key distribution technology to generate encryption keys. The key update frequency is high, the encryption algorithm conforms to international standards, and it can effectively resist quantum attacks, ensuring the confidentiality and integrity of data during remote data transmission and protecting the security of energy storage equipment operation data. The distributed data transmission network module can quickly reconverge when the network topology changes or the link fails, and it has a load balancing function. It can automatically allocate data transmission tasks according to the traffic and processing capacity of each node, improve network transmission efficiency and reliability, and ensure that data is stably and efficiently transmitted to the processing center or other monitoring nodes.

[0063] The role of data storage and management: The blockchain data storage module utilizes the immutability of blockchain to encrypt and store key operational data of energy storage equipment. Each data node maintains a complete copy of the data record, ensuring data authenticity and credibility. This provides a reliable basis for data verification in scenarios such as equipment operation history tracing, maintenance record querying, and electricity market transactions. The data fusion and management center module uses a weighted average method to fuse multi-source data, assigning weights based on the reliability and importance of different sensor data to obtain more accurate fused data results. The data is stored in a unified format and standard, while a data index and query mechanism are established to facilitate subsequent data retrieval and analysis, laying a solid data foundation for data analysis and decision-making.

[0064] The roles of the data analysis and decision-making layers are as follows: The cross-scale multiphysics coupling analysis module uses finite element analysis algorithms to perform cross-scale analysis on the fused multiphysics data, establishing a connection model between the micro and macro levels. This enables early prediction of potential macroscopic performance degradation or failure risks due to changes in microstructure. The spatiotemporal dimension expansion analysis engine module integrates data from different time periods and combines it with external information for prediction. Using time series analysis algorithms, it can predict the operating trends of equipment under different environmental conditions, such as the impact of changes in light intensity on the charging and discharging of energy storage equipment. This allows for early adjustment of equipment operating parameters and maintenance plans, ensuring safe and stable equipment operation. The predictive maintenance decision system module uses Q-learning reinforcement learning algorithms to formulate maintenance decision strategies. Through agent exploration and learning, it takes maintenance actions based on equipment status and updates the action value function to find the optimal maintenance plan, achieving a balance between equipment reliability and maintenance costs, and improving the overall lifecycle benefits of the equipment.

[0065] The role of the safety protection and emergency response layer: When a fault occurs, the adaptive energy shield fault isolation system module uses electromagnetic shielding, energy absorption and conversion technology to isolate the fault area, prevent the fault from spreading, convert and store excess energy or release it safely, reduce the harm of the fault and the risk of system downtime, and reduce the impact of the fault on the entire system. When a serious fault or safety risk is predicted, the emergency response mechanism module will trigger emergency measures in a timely manner. It will use fault tree analysis algorithms to build a fault tree model, determine the cause of the fault and the propagation path, and formulate targeted emergency response strategies, such as automatically cutting off power, starting backup equipment, or sending alarms to maintenance personnel and providing detailed information, thereby improving the efficiency and accuracy of emergency handling and ensuring the stable operation of the system. Specific Implementation Example 3

[0067] like Figure 1-10 As shown, based on the content of the above specific embodiments, the following content is further disclosed.

[0068] The algorithm content at each level of the overall system architecture is as follows:

[0069] The Kalman filter algorithm for the data acquisition layer includes core formulas such as the state prediction equation and the observation update equation, as shown below:

[0070] State prediction equation:

[0071] in:

[0072] This represents the prior state estimate at time k, which is a prediction of the current state based on the state at the previous time.

[0073] F kIt is the state transition matrix, which describes how the system state changes from time k-1 to time k. For example, in some linear dynamic systems, it reflects the evolution of physical quantities over time.

[0074] It is the posterior state estimate at time k-1, that is, the accurate state estimate obtained after updating the observations from the previous time step.

[0075] B k It is a control input matrix. When there are external control inputs to the system (such as adjustable device parameters affecting the system state), it is used to convert the control input quantity u... k When applied to the state transition process, u k It is the control input vector at time k.

[0076] Observational update equation:

[0077] in:

[0078] It is the posterior state estimate at time k, which is the accurate state estimate at the current time obtained after correcting the prior state estimate by incorporating the observations.

[0079] K k It is the Kalman gain matrix. Its calculation involves information such as the system's covariance. Its role is to weigh the weights of the observed values ​​and the predicted values ​​in order to determine how to correct the predicted values ​​based on the observed values ​​in order to obtain a more accurate state estimate. Its calculation also has a corresponding formula.

[0080] z k It is the observation vector at time k, which is the observation data about the system state obtained by actual measurement through sensors. For example, in this grid-connected photovoltaic power generation and energy storage equipment monitoring scenario, it is the actual measured values ​​of physical field parameters such as electric field and magnetic field collected.

[0081] H k It is the observation matrix, which maps the system's state vector to the observation space. In other words, it describes the correspondence between state quantities and actual observable physical quantities, such as the mathematical relationship of how a certain state quantity is transformed into a measurable observation value through a sensor.

[0082] In addition, related formulas such as covariance update are involved to continuously update information such as the error covariance of the state estimation, so as to continuously ensure the accuracy of filtering. For example, the formula for updating the prior estimation error covariance is as follows:

[0083]

[0084] in:

[0085] It is the prior estimation error covariance matrix at time k, used to measure the error of the prior state estimation;

[0086] P k-1 It is the posterior estimation error covariance matrix at time k-1;

[0087] Q k It is the process noise covariance matrix, which describes the covariance of noise introduced by various uncertainties (such as environmental disturbances, model inaccuracies, etc.) during the state transition of the system, and reflects the degree of uncertainty in state prediction.

[0088] And the relevant formulas for calculating Kalman gain:

[0089]

[0090] in:

[0091] R k It is the observation noise covariance matrix, which reflects the degree of uncertainty in the observed values ​​caused by factors such as sensor accuracy and environmental interference during the observation process.

[0092] The implementation steps of the data acquisition layer are as follows:

[0093] Sp1: Initialization: Set the initial state estimate. And the initial estimation error covariance matrix P0 is usually given based on prior knowledge of the initial state of the system or some reasonable assumptions. For example, in the monitoring of energy storage equipment, there is a rough estimate range for the initial state of physical quantities such as current and magnetic field of the equipment to determine the initial state estimate, and the initial error is estimated to determine P0.

[0094] Sp2: State Prediction: At each time k, using the state transition matrix F k State estimate of the previous time step and control inputs (if any) according to the state prediction equation To calculate the prior state estimate at the current time, and simultaneously update the formula based on the prior estimate error covariance. To update the prior estimation error covariance matrix to reflect changes in uncertainty during state prediction;

[0095] Sp3: Observation Acquisition: Obtaining the observation vector z by acquiring actual observation data through sensors. k For example, using multi-physics sensor network modules to collect parameters such as electric field and magnetic field values;

[0096] Sp4: Observation Update: Calculate the Kalman gain matrix K k(Based on the corresponding gain calculation formula), then use the observation update equation By combining the prior state estimate with the observed values, a more accurate posterior state estimate for the current time is obtained. At the same time, the posterior estimation error covariance matrix is ​​updated to prepare for the next round of filtering.

[0097] By initially screening, cleaning, and performing simple analysis on the collected raw data, abnormal data caused by factors such as sensor failure and sudden environmental interference can be identified and eliminated. This makes the data transmitted to subsequent layers cleaner and more reliable, avoiding invalid data from consuming transmission bandwidth and storage resources. The filtered data is more regular and predictable, reducing data complexity and uncertainty. It can flow more efficiently in subsequent transmission and processing. For example, when performing operations such as encrypted transmission at the data transmission layer, it can complete the corresponding process faster when dealing with relatively regular data, thereby reducing the overall data transmission latency and improving the real-time monitoring performance of the system. In the monitoring scenario of grid-connected photovoltaic power generation and energy storage equipment, data collected by various types of sensors (quantum sensing modules, multi-physics sensor network modules, etc.) are involved. The Kalman filter algorithm can reasonably fuse these multi-source heterogeneous data according to the characteristics of different sensor data and the needs of system state estimation, giving full play to the advantages of each sensor and providing a high-quality data foundation for more accurate equipment state assessment in the future.

[0098] The OSPF (Open Shortest Path First) algorithm for the data transfer layer is as follows:

[0099] The OSPF algorithm is built upon graph theory. Essentially, it's a dynamic routing algorithm that leverages the link-state information constantly exchanged between network nodes. It independently calculates the shortest path at each router node, iteratively comparing the path costs to the destination node via different intermediate nodes to find the optimal (shortest) transmission path. As the network topology or link state changes (such as link failures or bandwidth adjustments), this calculation process is retried to update the routing table in real time, ensuring data is always transmitted along the optimal path in the current network environment. Its core is calculating the shortest path between nodes in the network. The commonly used mathematical description is based on Dijkstra's algorithm, calculating the shortest path from node i to node j as follows:

[0100]

[0101] in:

[0102] d(i,j) represents the shortest path distance from node i to node j (which can be a metric defined based on link cost, such as the inverse of link bandwidth, transmission delay, etc., which are comprehensive measures of distance).

[0103] N is the set of all other nodes in the network except for the starting node i;

[0104] d(i, k) is the distance of the currently known shortest path from node i to intermediate node k;

[0105] w(k,j) is the link cost from intermediate node k to target node j (the link weight is a metric determined based on different network performance factors, such as link bandwidth, transmission delay, packet loss rate, etc., and the weight is set by taking these factors into account).

[0106] In actual OSPF algorithm implementations, data structures such as adjacency matrices of the network topology are used to represent the connectivity relationships and link costs between nodes in the network. For example, if the network has n nodes, the adjacency matrix A contains elements a... ij This represents the link cost between node i and node j. If there is no direct connection, it can be represented by a special value such as infinity.

[0107] The steps for implementing the data transmission layer are as follows:

[0108] Sp1: Network Topology Discovery: In the distributed data transmission network module, each router node exchanges information about its own connections and link states (such as bandwidth and latency) by sending and receiving Link State Advertisements (LSAs). This allows them to construct the overall network topology, understanding how the nodes (routers and other network devices) are connected and the performance status of each link. This information is then organized into corresponding data structures (such as adjacency matrices).

[0109] Sp2: Initialization: For the source node for which the shortest path needs to be calculated (e.g., when finding the shortest path from the edge computing node corresponding to the data acquisition end to the data processing center, this edge computing node is the source node), set its shortest path distance to itself to 0, and its initial shortest path distance to other nodes to infinity (indicating that the shortest path has not yet been determined). At the same time, mark the predecessor node of each node (initially empty, and will be continuously updated during the process of determining the shortest path):

[0110] Sp3: Iterative Calculation: Repeat the following steps to traverse all nodes in the network except the source node. For each node j, check all its neighboring nodes k (determined by adjacency relationships in the network topology). According to the shortest path calculation formula above, try to update the shortest path distance from the source node i to node j, i.e., compare the currently known d(i,j) and d(i,k)+w(k,j). If a shorter path distance can be obtained through the intermediate node k, update d(i,j) and the corresponding predecessor node information. Repeat this process until the shortest path distance of all nodes no longer changes (convergence is achieved). At this point, the shortest path from the source node to each node in the network is determined.

[0111] SP4: Routing Table Update: Each router node updates its routing table based on the calculated shortest path information. The routing table records information such as the next-hop node to different destination nodes and the corresponding shortest path distance. This allows the best path for data transmission to be quickly determined based on the routing table when data needs to be transmitted.

[0112] The quantum key distribution technology for the data transmission layer is as follows:

[0113] The BB84 protocol involves mathematical descriptions of the preparation and measurement of quantum states, including related probability calculations. The preparation of a qubit can typically be represented in the following form:

[0114] |ψ>=α|0>+β|1>;

[0115] in:

[0116] |ψ> represents the quantum state of a qubit, which is a superposition state and a vector in two-dimensional Hilbert space;

[0117] |0> and |1> are the two ground states of a qubit, similar to the 0 and 1 states of a classical bit. In physical implementation, they can correspond to different polarization states of a photon, for example.

[0118] α and β are complex numbers and satisfy the normalization condition |α| 2 +|β| 2 =1, and the squares of their moduli represent the probabilities of measuring the ground states |0> and |1>, respectively, that is, the probability of measuring |0> is |α|. 2 The probability of obtaining |1> is |β|. 2 ;

[0119] During key generation, the sender and receiver perform operations such as preparing, transmitting, and measuring qubits with different bases (e.g., rectangular polarization base and diagonal polarization base). The consistency of the measurement results determines the final qubit used for key generation. Assume the sender prepares qubits with a certain probability, choosing different bases for operation. Let p be the probability of choosing a rectangular polarization base (usually denoted as the Z-base). Z The probability of choosing a diagonally polarized basis (usually denoted as the X basis) is p. X , and p Z +p X =1.

[0120] The implementation steps are as follows:

[0121] Sp1: Quantum bit preparation: The sender (such as the related quantum encryption communication equipment at the data acquisition end in the grid-connected photovoltaic power generation and energy storage equipment monitoring system) uses quantum light source and other equipment to prepare a series of qubits. Different bases (such as the Z base or X base mentioned above) are randomly selected according to a certain probability to prepare the qubits into corresponding superposition states. Then, the qubits are sent to the receiver (the corresponding receiving quantum encryption communication equipment such as the data processing center) through a quantum channel (such as optical fiber and other media suitable for transmitting quantum signals).

[0122] Sp2: Quantum bit measurement: The receiver also randomly selects either the Z basis or the X basis to measure the received qubit. Due to the properties of quantum mechanics, the measurement results are highly consistent (in accordance with the correct measurement principle of quantum states) only when the sender and receiver select the same basis for operation; otherwise, the measurement results are basically random.

[0123] Sp3: Key Screening and Negotiation: The sender and receiver inform each other of the basis they chose when preparing and measuring the qubits through a classical communication channel (such as a traditional wired or wireless network). Then, they compare the results and retain only the bits corresponding to the measurement results of those that have chosen the same basis. After further error checking (such as checking for errors caused by eavesdropping through some check codes), these selected consistent bits form the final quantum key, which is used for subsequent encryption and decryption operations of transmitted data.

[0124] The entire process utilizes fundamental properties of quantum mechanics, such as the superposition of quantum states and measurement collapse. By randomly selecting different bases to prepare qubits, the sender increases the difficulty for eavesdroppers to obtain the correct information. This is because if an eavesdropper wants to obtain the qubit state for eavesdropping measurement, since they do not know the base chosen by the sender, the measurement will change the quantum state. This will lead to a large number of inconsistencies when the receiver and sender compare the bases later (this phenomenon is called quantum state perturbation, which can be used to detect eavesdropping behavior). This ensures that only legitimate communicating parties can select consistent bits through correct base selection and measurement to generate a secure key. Furthermore, the randomness and confidentiality of this key are guaranteed by the physical laws of quantum mechanics, which are then used to encrypt the transmitted data to ensure its confidentiality and integrity.

[0125] In the data fusion and management center module of the data storage and management layer, the weighted average method is used. Assuming the fused data comes from n different sensors, the weighted average formula for the fusion of a certain physical quantity is:

[0126]

[0127] in:

[0128] It is the merged data value, representing the final result after weighted averaging.

[0129] x i It is the data collected by the i-th sensor. For example, in the monitoring of grid-connected photovoltaic power generation and energy storage equipment, x1 may be the microscopic physical quantity data collected by the quantum sensor, and x2 may be the electric field data collected by a traditional high-precision sensor in a multi-physics sensor network.

[0130] w i It is the weight of the i-th sensor data, and the magnitude of the weight depends on the reliability and importance of the sensor data;

[0131] This represents the sum of all weights and is used for normalization to ensure that the merged data is within a reasonable range.

[0132] Data from each sensor x i According to its corresponding weight w iIn the final calculation of the fused data, the sensor data with the greater weight has a greater impact on the fusion result. In this way, the characteristics of different sensor data are comprehensively considered, so that the fused data can more accurately reflect the actual operating status of the equipment. Different sensors have different accuracies and error ranges. By weighted averaging, the advantages of each sensor can be used to make up for the shortcomings of a single sensor, thereby obtaining more accurate data on the operating status of energy storage equipment. For example, for a physical quantity, quantum sensors have high accuracy but may be greatly affected by the environment, while traditional sensors have good stability but slightly lower accuracy. By weighted averaging, a better result can be obtained. By fusing data from multiple sources into a unified value, a simpler and more representative data foundation is provided for subsequent data storage, analysis and decision-making, avoiding data redundancy and conflicts.

[0133] The content of the finite element analysis algorithm in the cross-scale multiphysics coupling analysis module for data analysis and decision-making layers is as follows:

[0134] For a two-dimensional steady-state heat conduction problem, its finite element equation can be expressed as:

[0135] KT = Q;

[0136] in:

[0137] K is the thermal conductivity matrix, which is determined by factors such as the thermal conductivity of the material, the geometry of the elements, and the nodal connections. Its elements k ij This represents the thermal conduction coupling relationship between node i and node j. For example, in an energy storage device, if the thermal conductivity between two units is high, then the corresponding k... ij A larger value indicates that heat is more easily transferred between these two nodes;

[0138] T is the nodal temperature vector, which is the unknown quantity to be solved. i This represents the temperature of the i-th node;

[0139] Q is the node heat flux vector, which represents the heat flux input or output applied to each node from the outside. For example, in an energy storage device, if there is a heat source in a certain part, then the corresponding node heat flux vector element will not be zero.

[0140] Solving the above equations usually requires numerical methods, such as Gaussian elimination or iterative methods. Taking the Jacobi iteration method as an example, its iterative formula is as follows:

[0141]

[0142] in:

[0143] T i (k+1) It is the temperature estimate of the i-th node after the (k+1)-th iteration;

[0144] T j (k) It is the temperature estimate of the j-th node after the k-th iteration;

[0145] k represents the number of iterations.

[0146] The implementation steps are as follows:

[0147] Sp1: Model Establishment: Discretize the physical field (such as thermal field, electric field, etc.) region of the energy storage device into a finite number of elements, determine the type of element (such as triangular element, quadrilateral element, etc.) and the node position, and at the same time, determine the thermal conductivity coefficient matrix K and the node heat flux vector Q according to the material properties and boundary conditions;

[0148] Sp2: Initial Guess: Give an initial guess value for the node temperature vector T. For example, you can assume that the temperature of all nodes is an average temperature or give an initial distribution based on experience.

[0149] Sp3: Iterative Solution: Perform multiple iterations according to the iterative formula until a convergence condition is met. The convergence condition can be that the difference between two adjacent iteration results is less than a certain set threshold, such as |T i (k+1) -T i (k) |<∈, where ∈ is a very small positive number;

[0150] Sp4: Result Analysis: After obtaining the final solution of the node temperature vector T, analyze the temperature distribution and related physical quantities such as temperature gradient to evaluate the physical field state inside the energy storage device.

[0151] Finite element analysis (FEM) algorithms solve complex continuous physical field problems by discretizing them into combinations of a finite number of elements. Starting with an initial guess of the temperature distribution, the algorithm uses the physical relationships described by the thermal conductivity matrix and the nodal heat flux vectors to iteratively refine the temperature estimates until the nodal temperature vectors converge to a solution that satisfies the physical laws. This solution reflects the relationship between the microscopic particle behavior and macroscopic device performance within the energy storage device, such as the relationship between temperature distribution and changes in the microstructure of battery materials or device performance degradation. It can establish a model linking microscopic particle behavior and macroscopic device performance. By solving for the physical field distribution, it can delve deeper into the physical laws behind the data. For example, it can predict in advance the macroscopic performance degradation or failure risks that may be caused by changes in microstructure, such as the risk of thermal runaway caused by microscopic defects in battery materials or short-circuit faults caused by electromagnetic stress concentration inside power electronic devices. It supports cross-scale physical field analysis, comprehensively evaluating everything from microscopic material properties to macroscopic overall device performance, providing a powerful tool for a comprehensive understanding of the operating status of energy storage devices.

[0152] The spatiotemporal dimension expansion analysis engine module - time series analysis algorithm adopts the ARIMA autoregressive moving average algorithm. The general form of the ARIMA model is:

[0153]

[0154] in:

[0155] y t It is time series data. In the monitoring of grid-connected photovoltaic power generation and energy storage equipment, it can be a sequence of changes in a certain operating parameter of the equipment (such as voltage, current, etc.) over time.

[0156] B is the shift operator, defined as By t =y t-1 B 2 y t =y t-2 wait;

[0157] It is an autoregressive polynomial. These are autoregressive coefficients, which represent the linear regression relationship between the current value of a time series and its p past values. For example, Indicates y t With y t-1 The strength of the correlation between them;

[0158] (1-B) d It is a difference operator used to transform non-stationary time series into stationary time series. d is the difference order. When a time series has non-stationary characteristics such as trend or seasonality, these characteristics can be removed through the difference operation, making the series stationary.

[0159] θ(B) = 1 + θ1B + θ2B 2 +…+θ q B q It is a moving average polynomial, θ j It is the moving average coefficient, which represents the ratio of the current value of the time series to the past q white noise terms. t A linear relationship between -j;

[0160] ∈ t It is a white noise sequence with a mean of 0 and a variance of a constant σ. 2 Furthermore, the white noise terms at different times are independent of each other.

[0161] The ARIMA model first differencing non-stationary time series to make them stationary. Then, it constructs a model based on autoregressive and moving average relationships in historical data. By learning from training data, it estimates the model's parameters, thus establishing a generative model for time series data. In the prediction phase, it uses the established model and known historical data to predict future time series data. Simultaneously, it incorporates external environmental data (such as meteorological data) for long-term trend correction, enabling more accurate predictions of equipment operating trends under different environmental conditions over a future period. It integrates historical data, current real-time data, and future data predicted based on external information such as meteorological data to construct... An analytical model with spatiotemporal dimensions is built to extract data trends and seasonality features. Short-term predictions are made based on current data, while long-term trend corrections are made using external environmental data. This predicts the operating trends of the equipment under different environmental conditions in the future. For example, by combining weather forecasts to predict the impact of changes in light intensity on the charging and discharging of energy storage equipment, operating parameters and maintenance plans can be adjusted in advance to ensure the safe and stable operation of the equipment under different spatiotemporal conditions. This provides data support for the operation management and maintenance decisions of the equipment. Through accurate trend prediction, operating strategies can be planned in advance, such as adjusting charging and discharging power and scheduling maintenance time, thereby improving the operating efficiency and reliability of the equipment.

[0162] The Q-learning algorithm in the predictive maintenance decision system module is as follows:

[0163] The core of the Q-learning algorithm is updating the action value function, and its update formula is:

[0164]

[0165] in:

[0166] s t It is the state of the intelligent agent at time t. In the maintenance decision of energy storage equipment, the state can be the equipment operating state represented by the combination of the equipment's operating parameters (such as temperature, voltage, etc.).

[0167] a t It refers to the action taken by the intelligent agent at time t, such as maintenance operations such as inspecting, repairing, or replacing parts of the equipment;

[0168] r t+1 It is the reward that the agent receives at time t+1. The reward can be a quantitative indicator such as improved equipment reliability or reduced costs. For example, if a maintenance operation reduces the probability of equipment failure, then the corresponding reward value will be higher.

[0169] α is the learning rate, which controls the speed of learning. 0 < α < 1. If α is large, the agent can more easily update the action value function quickly based on new experience. If α is small, the learning process will be slower and will focus more on historical experience.

[0170] γ is a discount factor, 0≤γ≤1. It is used to measure the importance of future rewards. When γ is close to 1, the agent pays more attention to long-term rewards, and when γ is close to 0, the agent pays more attention to short-term rewards.

[0171] max a Q(s t+1 , a) represents the value of the action with the highest value selected by the agent from all possible actions a at time t+1.

[0172] The implementation steps are as follows:

[0173] Sp1: Environmental Modeling: Defines the environment for energy storage device maintenance decisions, including the set of possible device states S, the set of possible maintenance actions A, the reward function r(s, a), and the state transition probability P((s, a)). t+1 )|s t a t The state transition probability describes the probability of transitioning to the next state after taking an action in the current state;

[0174] Sp2: Initialization: Initialize the action value function Q(s, a). You can set its initial value to 0 or give some initial estimates based on experience. At the same time, set parameters such as learning rate α and discount factor γ.

[0175] Sp3: Agent Exploration and Learning: The agent iterates multiple times in the environment. In each iteration, based on the current state s... t The agent selects an action 'a' according to a certain strategy (such as an ∈-greedy strategy, i.e., randomly selecting an action with probability ∈, and selecting the action with the maximum action value with probability 1-∈). t After this action is performed, the environment transitions to the next state s according to the state transition probability. t+1 and return a reward r t+1 Then, update the action value function Q(s) according to the update formula. t a t );

[0176] Sp4: Policy Optimization: As the learning process progresses, the agent continuously optimizes its decision-making strategy based on the updated action value function, so as to select the action that can obtain the maximum long-term reward in each state.

[0177] Q-learning algorithms optimize the action value function by having an agent continuously explore and learn in the energy storage equipment maintenance decision-making environment. The agent tries different actions in each state and updates the action value function based on the reward feedback obtained, thereby gradually learning which actions to take in different states to obtain the optimal long-term reward. This process is a continuous trial and error and optimization process, with the ultimate goal of finding the optimal maintenance plan, including maintenance time, maintenance content, and required resources, to achieve a balance between equipment reliability and maintenance cost.

[0178] The Fault Tree Analysis (FTA) algorithm for the emergency response mechanism module in the security protection and emergency response layer is as follows:

[0179] Assuming the probabilities of occurrence of underlying events (basic events) E1 and E2 are P(E1) and P(E2) respectively, then the probability of occurrence of top-level event (system failure event) T is:

[0180] P(T) = P(E1) × P(E2);

[0181] in:

[0182] P(T): Represents the probability of a system failure event occurring, used to assess the overall failure risk level of the system;

[0183] P(E1) and P(E2): represent the probabilities of two basic events that lead to system failure. The basic events can be a component failure in the energy storage device, a specific erroneous operation, etc.

[0184] For a fault tree with an OR gate structure, assuming the probabilities of the bottom-level events (basic events) E3 and E4 are P(E3) and P(E4) respectively, then the probability of the top-level event (system failure event) T is:

[0185] P(T)=1-(1-P(E3))×(1-P(E4));

[0186] The logic here is that as long as either event E3 or E4 occurs, system failure event T will occur. First, calculate the probability that neither of the two basic events occurs, i.e., (1-P(E3))×(1-P(E4)). Then, subtract this probability from 1 to get the probability that the system failure event occurs.

[0187] In real-world complex fault trees, there may be multiple AND gates, OR gates, and combinations of basic events at different levels. By using Boolean algebra rules and probability calculation rules, the probability of system fault events can be calculated step by step.

[0188] Implementation steps:

[0189] Sp1: Fault Tree Construction: Identify system fault events (top-level events). In a grid-connected photovoltaic power generation and energy storage equipment safety monitoring system, these events may include the complete shutdown of the energy storage equipment or the occurrence of serious safety accidents (such as fires, explosions, etc.). Identify the direct causes of system faults as intermediate events and further decompose them into basic events. Basic events are usually indivisible equipment faults, human errors, environmental factors, etc. For example, battery short circuits, cooling system faults, and sensor false alarms can all be used as basic events. Determine the logical relationships between events and use AND gates (meaning that the output event occurs only when all input events occur simultaneously) and OR gates (meaning that the output event occurs as long as one input event occurs) to connect the various events and construct the fault tree.

[0190] S2: Probability Assignment: For basic events, the probability of their occurrence is determined through historical data statistics, equipment reliability manuals, expert experience, etc. For example, if according to historical data, the probability of a certain type of battery short-circuiting under specific operating conditions is 0.01, then this probability is assigned to the corresponding basic event.

[0191] SP3: Fault Probability Calculation: Starting from the basic events at the bottom layer, the probability of system fault events is calculated step by step upwards according to the logical relationships between events in the fault tree (AND gates or OR gates) and the above probability calculation formula is followed.

[0192] In the adaptive energy shield fault isolation system module, when a fault occurs, the entire system operates with the goal of protecting the equipment and reducing the damage caused by the fault. First, it quickly detects fault signals, which is achieved by setting the normal operating parameter range of the equipment and comparing real-time monitoring data. Once a fault is detected, the energy shield system immediately activates, using electromagnetic shielding to prevent the spread of electromagnetic interference generated by the fault. At the same time, it processes excess energy through energy absorption and conversion devices to prevent energy accumulation from leading to more serious fault consequences, such as fire or further damage to the equipment. Deploying the adaptive energy shield fault isolation system around the energy storage equipment, when a fault is detected, it quickly isolates the fault area, prevents the fault from spreading, and converts and stores or safely releases excess energy, reducing the damage caused by the fault and the risk of system downtime. This helps protect other normal components of the energy storage equipment, reduces the scope and severity of the fault's impact on the entire system, maintains some functions of the system, and improves the system's reliability and stability. By handling abnormal energy fluctuations caused by faults in a timely and effective manner, it avoids potential safety accidents, such as explosions caused by thermal runaway or electrical fires caused by short circuits, ensuring the safe operation of the grid-connected photovoltaic power generation system. Specific Implementation Example 4

[0194] Based on the structure of the machine learning algorithm in the above system, a visualization experiment was conducted on the algorithm, and the data is as follows: Figure 3-10As shown, it specifically includes the following:

[0195] like Figure 3 As shown in the diagram, the Kalman filter algorithm is visualized. The graph clearly shows the effect of this feedback. The observed values ​​(green dots) are noisy, relatively discrete, and deviate from the true state (blue line) due to the interference of observation noise. However, the estimated values ​​after Kalman filtering (red line) are continuously adjusted using the feedback of the observed values ​​as the time step progresses, gradually approaching the true state value. This demonstrates the algorithm's ability to effectively integrate information, reduce uncertainty, and improve estimation accuracy through the feedback mechanism.

[0196] like Figure 4 The diagram shows the network topology and shortest path. From the network topology diagram, we can roughly understand this feedback: when the shortest path of a certain node (such as node A) is found, its neighboring nodes (such as nodes B, C, etc.) will re-evaluate whether their distance to the source node needs to be updated based on the link cost with node A and the newly determined shortest distance from node A to the source node. If it is updated, it will affect the distance evaluation of other nodes adjacent to these nodes. This process is repeated iteratively, with continuous feedback and adjustment, until the shortest distance of all nodes no longer changes (reaching convergence).

[0197] like Figure 5 As shown in the diagram, the comparison between the base selection and the measurement results shows the randomness of the sender's base selection and the distribution of the receiver's measurement results. Only with feedback information based on the consistency of the base selections of both parties can valid key bits that can be used for encryption be extracted from numerous measurement results, which reflects the feature of ensuring the security of key generation based on information feedback.

[0198] like Figure 6 As shown in the figure, the line graphs of the data from each sensor and the fused data are assumed to be dynamically adjusted according to certain judgment conditions (such as the deviation between the data and the true value, data stability, etc.) in subsequent actual operation. Then, the fused data is recalculated and the line graph is drawn. It can be seen that the fused data line (black line) will be closer to the data line of the sensor with increased weight as the weight changes (for example, after the weight of the sensor corresponding to the red line increases, the fused data line will be closer to the sensor data line represented by the red line). This intuitively reflects the impact of weight adjustment (feedback mechanism) on the fusion result.

[0199] like Figure 7 As shown, a line graph of the original time series and the predicted series;

[0200] like Figure 8 The image shows a heatmap visualization of the action value function Q(s,a).

[0201] like Figure 9 As shown in the line graph, the impact of changes in the probability of basic events on the probability of top-level events is clearly visible. The graph shows that when the probability of a basic event (e.g., E1) changes, the probability of the top-level event (T) changes accordingly. For example, as the probability of basic event E1 increases (along the positive x-axis), the probability of top-level event T (y-axis value) also increases. This indicates that changes in the probability of basic events are fed back into the overall system failure status (top-level events) through the logical structure of the fault tree (combinations of AND gates, OR gates, etc.). This helps us analyze the impact of different basic events on system reliability. Furthermore, in emergency response mechanisms, this feedback information allows us to focus on the components or factors corresponding to critical basic events that have a significant impact on system failure, and to formulate more targeted emergency strategies and maintenance plans.

[0202] like Figure 10 As shown in the line graph, the shield dynamically adjusts the remaining energy based on the real-time monitored fault energy level through energy absorption and conversion technology. The line graph clearly shows that the fault energy (red line) fluctuates randomly, while the remaining energy (blue line) is generally lower than the fault energy after being processed by the energy shield. This is because the shield processes the fault energy according to its set absorption and conversion efficiency, absorbing and converting a portion of the energy to maintain the remaining energy at a relatively low level, preventing excessive energy from causing more serious damage to the system.

[0203] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, 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 a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising a reference structure" does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0204] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A safety monitoring system for grid-connected photovoltaic power generation and energy storage equipment, characterized in that: The data acquisition layer is used to acquire operational status data, microscopic physical quantity data, and multi-field parameter data from multiple sources, and to perform preliminary processing on the acquired data. The data transmission layer is used for quantum encryption of the acquired and processed data and to optimize the data transmission path. The data storage and management layer is used to store the data collected above and integrate the multi-source data transmitted above. The data analysis and decision-making layer is used to analyze the physical correlations of the above multi-source data, predict trends, and make decisions.

2. The safety monitoring system for grid-connected photovoltaic power generation and energy storage equipment according to claim 1, characterized in that: The data acquisition layer includes a quantum sensing module, a multiphysics sensor network module, and an IoT edge computing node module. The quantum sensing module is used to deploy quantum magnetometers and quantum gyroscopes to collect information on the internal current distribution, magnetic field changes, and device attitude of the energy storage device. The multiphysics sensor network module is used to monitor multiphysics parameters such as electric field, magnetic field, thermal field, and stress field. The IoT edge computing node module uses a Kalman filter algorithm to perform preliminary screening, cleaning, and analysis on the raw data collected above, and to make preliminary judgments and marks on abnormal data.

3. The safety monitoring system for grid-connected photovoltaic power generation and energy storage equipment according to claim 1, characterized in that: The data transmission layer includes a quantum encrypted communication link module and a distributed data transmission network module. The quantum encrypted communication link module is used to establish a quantum encrypted communication channel from the data acquisition end, i.e., the edge computing node, to the data processing center, and uses quantum encryption to encrypt the transmitted data. The distributed data transmission network module is used to build a distributed data transmission network by combining wired and wireless networks, and uses the OSPF (Open Shortest Path First) routing algorithm to optimize the data transmission path, dynamically selecting the best data transmission path based on the network topology and link status information.

4. The safety monitoring system for grid-connected photovoltaic power generation and energy storage equipment according to claim 1, characterized in that: The data storage and management layer includes a blockchain data storage module and a data fusion and management center module. The blockchain data storage module is used to encrypt the collected energy storage device operation data and store it in the blockchain network, with each data node storing a complete copy of the data record. The data fusion and management center module is used to receive data from the data transmission layer and integrate and manage it. It uses the weighted average method in the data fusion algorithm to fuse quantum sensing data, multiphysics field data, and data from other sources, assigning corresponding weights according to the reliability and importance of different sensor data, storing it according to a unified data format and standard, and establishing a data index and query mechanism.

5. The safety monitoring system for grid-connected photovoltaic power generation and energy storage equipment according to claim 1, characterized in that: The data analysis and decision-making layer includes a cross-scale multiphysics coupling analysis module, a spatiotemporal dimension expansion analysis engine module, and a predictive maintenance decision-making system module. The cross-scale multiphysics coupling analysis module uses finite element analysis algorithms to perform cross-scale analysis on the fused multiphysics data, discretizing the complex physical system into a finite number of units for solution. By establishing a model linking microscopic particle behavior and macroscopic equipment performance, it uncovers the physical laws of the data and predicts in advance the risk of macroscopic performance degradation or failure due to changes in microstructure. The spatiotemporal dimension expansion analysis engine module integrates historical data, current real-time data, and future data predicted based on external information to construct a spatiotemporal dimension expansion analysis model. It uses the ARIMA autoregressive moving average algorithm of time series analysis to analyze historical data, extracting data trends and seasonal characteristics. Combined with current data, it performs short-term predictions and uses external environmental data for long-term trend correction, predicting the operating trend of equipment under different environmental conditions in the future and adjusting equipment operating parameters and maintenance plans in advance. The predictive maintenance decision-making system module uses Q-learning reinforcement learning algorithms to intelligently formulate predictive maintenance decision strategies based on the above analysis results.

6. The safety monitoring system for grid-connected photovoltaic power generation and energy storage equipment according to claim 1, characterized in that: It also includes a security protection and emergency response layer, which includes an adaptive energy shield fault isolation system module and an emergency response mechanism module; The adaptive energy shield fault isolation system module is used to deploy an adaptive energy shield fault isolation system around the energy storage device. When a fault is detected, it uses electromagnetic shielding, energy absorption and conversion to isolate the fault area, prevent the fault from spreading, and convert and store or safely release excess energy. The emergency response mechanism module is used to establish an emergency response mechanism. When the data analysis layer predicts a fault or safety risk, it promptly triggers corresponding emergency measures, uses fault tree analysis (FTA) algorithm to analyze the emergency situation, constructs a fault tree model, determines the possible causes and propagation paths of the fault, and formulates emergency response strategies.

7. A monitoring method for a safety monitoring system for grid-connected photovoltaic power generation and energy storage equipment according to any one of claims 1-6, characterized in that... The steps are as follows: The system acquires operational status data, microscopic physical quantity data, and multi-field parameter data from multiple sources and performs preliminary processing on the acquired data. The collected and processed data is quantum encrypted and the data transmission path is optimized. Store the collected data and integrate the multi-source data transmitted above; Analyze the physical correlations of the above multi-source data and predict trends to make decisions.

8. The safety monitoring method for grid-connected photovoltaic power generation energy storage equipment according to claim 7, characterized in that: In the multi-source data acquisition step, the following data are collected: the internal current distribution, magnetic field changes, and device attitude information of the energy storage device; multi-physics field parameters such as electric field, magnetic field, thermal field, and stress field are collected; and the Kalman filter algorithm is used to perform preliminary screening, cleaning, and analysis of the raw data collected above, and to make preliminary judgments and mark abnormal data. In the data transmission steps described above, a quantum-encrypted communication channel is established from the data acquisition end to the data processing center, and quantum encryption is used to encrypt the transmitted data; a distributed data transmission network is constructed by combining wired and wireless networks, and the OSPF (Open Shortest Path First) routing algorithm is used to optimize the data transmission path, dynamically selecting the best data transmission path based on the network topology and link status information; In the data storage and integration management steps, the collected energy storage device operation data is encrypted and stored in the blockchain network, and each data node saves a complete copy of the data record. It receives data from the data transmission layer, integrates and manages it, and uses the weighted average method in the data fusion algorithm to fuse quantum sensing data, multiphysics field data and data from other sources. It assigns corresponding weights according to the reliability and importance of different sensor data, stores it according to a unified data format and standard, and establishes a data index and query mechanism. In the data analysis and decision-making steps, the finite element analysis algorithm is used to perform cross-scale analysis on the fused multiphysics data. The complex physical system is discretized into a finite number of units for solution. By establishing a model linking microscopic particle behavior and macroscopic equipment performance, the physical laws of the data are explored, and the risk of macroscopic performance degradation or failure due to changes in microstructure is predicted in advance. Historical data, current real-time data, and future data predicted based on external information are integrated to construct an analysis model with expanded spatiotemporal dimensions. The ARIMA autoregressive moving average algorithm of time series analysis is used to analyze historical data, extract data trends and seasonal characteristics, and make short-term predictions in combination with current data. At the same time, long-term trend correction is made using external environmental data to predict the operating trend of the equipment under different environmental conditions in the future, and adjust the equipment operating parameters and maintenance plans in advance. Based on the above analysis results, the Q-learning reinforcement learning algorithm is used to make intelligent decisions and formulate predictive maintenance decision strategies.

9. A safety monitoring method for grid-connected photovoltaic power generation and energy storage equipment according to claim 7, characterized in that: It also includes safety protection and emergency response steps, such as isolating energy and developing strategies in the event of a failure.

10. A safety monitoring method for grid-connected photovoltaic power generation and energy storage equipment according to claim 9, characterized in that: In the aforementioned safety protection and emergency response steps, an adaptive energy shield fault isolation system is deployed around the energy storage device. When a fault is detected, the fault area is isolated by electromagnetic shielding, energy absorption and conversion to prevent the fault from spreading and to convert and store or safely release excess energy. Establish an emergency response mechanism. When the data analysis layer predicts a fault or security risk, trigger corresponding emergency measures in a timely manner. Use the Fault Tree Analysis (FTA) algorithm to analyze the emergency situation, construct a fault tree model, determine the possible causes and propagation paths of the fault, and formulate emergency response strategies.

Images