Intelligent power supply management system based on multi-energy complementation

Abstract

The invention provides an intelligent power supply management system based on multi-energy complementation, which relates to the technical field of intelligent power supply management and comprises a multi-energy supply module, a data acquisition module, a multi-source data fusion module, a scheduling optimization module, a demand prediction and response module, a supply safety evaluation module, an energy storage and load balancing module and an energy scheduling module. The multi-energy supply module comprises a solar energy supply sub-module, a wind energy supply sub-module, an energy storage system and a traditional power grid access sub-module, and by combining multi-source data and an advanced algorithm, the precision of power load prediction is remarkably improved. Compared with a traditional prediction method based on historical data, the method comprehensively considers the influence of various external factors such as climate change, holidays and festivals, and can more accurately predict the power demand fluctuation. The system can dynamically adjust the load prediction model in real time, improves the reliability of power supply, and can effectively cope with challenges caused by demand changes, environment changes and emergencies.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent power supply management, and specifically to an intelligent power supply management system based on multi-energy complementarity. Background Art

[0002] Existing power supply management solutions usually adopt static scheduling methods based on traditional power grid structures, relying on fixed load forecasting models for power distribution. However, these methods have significant limitations in coping with power demand fluctuations, environmental changes, and emergencies, and cannot meet the requirements of modern power systems for dynamics, intelligence, and efficiency.

[0003] According to an intelligent power supply management system with the Chinese patent number CN 116452367 B. It includes a balance area matching module and a balance time point planning module. The invention determines the high and low electricity consumption in different regions at different time points through the balance area matching module, selects the matching area, and at the same time, through the balance time point planning module, according to the matching area and the matching time point, plans the balance time point, and takes the balance time point as the centralized power supply time point: timely disconnects the non-essential power supply areas, concentrates the power supply to the areas with high electricity consumption needs, and reasonably distributes the power supply within different times to achieve intelligent power supply management.

[0004] Although the existing intelligent power supply management system can reasonably distribute the power supply within different times and achieve intelligent power supply management, there are still the following problems: Problem 1: The traditional power supply management system formulates power supply plans based on historical load data and simple time period division methods. However, the existing load forecasting technology usually calculates based on single historical electricity consumption data, fails to effectively consider the impact of external factors such as climate change and holidays on power demand. At the same time, the existing power supply management system usually adopts a fixed-time scheduling method, making it difficult to respond to changes in power demand in real time; Problem 2: Traditional power supply management systems are mostly regional scheduling, and fail to conduct differential management according to the personalized needs of each electricity user. Especially in large-scale cities and industrial areas, the electricity consumption patterns of users vary significantly. The traditional power supply management system cannot accurately match various demands, resulting in power supply surplus for some users and power supply shortage for some users; Problem 3: Although modern smart grids have adopted real-time data monitoring in some fields, most systems still fail to make full use of real-time data for efficient processing, resulting in a lag in the prediction of power demand, affecting the real-time performance and accuracy of scheduling. At the same time, when the existing intelligent power supply system conducts load forecasting, although it sometimes considers weather factors, it fails to comprehensively analyze various meteorological factors such as temperature, humidity, and wind speed in environmental changes. Especially the sudden impact of extreme weather events on power demand is often ignored; Question 4: Although many current intelligent power supply systems have introduced machine learning and artificial intelligence algorithms, they mostly focus on static load forecasting and single optimization scheduling. The system often fails to combine multiple algorithms for comprehensive optimization, thus failing to achieve more efficient load balancing and resource allocation. Summary of the Invention

[0005] In view of the deficiencies of the prior art, the present invention provides an intelligent power supply management system based on multi-energy complementarity, which solves the problem that the function management in the prior art has a scheduling lag behind the load demand.

[0006] To achieve the above objectives, the intelligent power supply management system based on multi-energy complementarity of the present invention is realized through the following technical solutions: Multi-energy supply module, which selects the energy supply source according to the real-time environmental conditions and load demands; Data acquisition module, which acquires the real-time data of the multi-energy supply module, and acquires meteorological data, load demand data, and equipment status data; Multi-source data fusion module, which fuses the data acquired by the data acquisition module; Energy scheduling module, which schedules energy resources according to the fused real-time load demand data and environmental change data; Demand prediction and response module, which predicts the future energy load demand according to the fused data, combined with weather forecasts and historical load data, and adjusts the load response strategy according to the prediction results; Supply security assessment module, which assesses the security of the current energy supply according to the fused data, and predicts potential supply interruption or power fluctuation risks; Energy storage and load balancing module, which dynamically manages the energy storage nodes according to the fused data, and balances the load and energy storage; Scheduling optimization module, which dynamically adjusts the energy supply constraint conditions and optimization objectives according to the current energy resource scheduling of the energy scheduling module, the output of each energy source, and the system operation indicators during real-time operation according to environmental changes and load demands.

[0007] The present invention has the following beneficial effects: 1. By combining multi-source data with advanced algorithms, the present invention significantly improves the accuracy of power load forecasting. Compared with the traditional prediction method based on historical data, the present invention comprehensively considers the influence of various external factors such as climate change and holidays, and can more accurately predict the power demand fluctuations. The system can dynamically adjust the load prediction model in real time, improve the reliability of power supply, and can effectively cope with the challenges brought by demand changes, environmental changes, and emergencies.

[0008] 2. The present invention introduces a scheduling optimization algorithm based on real-time data, which can achieve real-time dynamic adjustment of power supply. When the power demand changes, the system can respond quickly, avoiding the lag problem of traditional scheduling methods, and ensuring that during peak load periods or emergencies, power scheduling can balance the load more flexibly and accurately, optimizing power supply distribution. This real-time scheduling ability improves the emergency response and scheduling efficiency of the power supply system.

[0009] 3. The present invention further strengthens the matching ability for personalized electricity consumption needs. By precisely analyzing the electricity consumption patterns and needs of each user, the system can supply power differentially according to specific situations, avoiding the problems of over-supply or under-supply caused by "uniform" distribution in traditional regional scheduling. This personalized scheduling method not only improves the utilization efficiency of power resources but also effectively enhances the electricity consumption satisfaction of users, meeting the actual needs of different regions and user groups.

[0010] 4. The present invention collects and analyzes multi-dimensional environmental data (such as temperature, humidity, wind speed, etc.), and fully considers the impact of environmental factors on power demand in load forecasting and scheduling. Especially during extreme weather events, the system can predict in advance and adjust the power supply strategy, thereby enhancing the adaptability of the power grid to climate change and sudden weather events. In addition, by combining multiple intelligent algorithms (such as reinforcement learning, particle swarm optimization, etc.), the present invention improves the overall scheduling efficiency of the power system, realizes more efficient resource allocation and energy utilization, and promotes the development of the power grid towards intelligence and automation. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 is the overall system composition diagram of the present invention; Figure 2 is the simulation diagram of the energy supply security assessment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0012] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention. Specific Embodiment 1 As Figure 1-2 shown, the intelligent power supply management system based on multi-energy complementarity includes: The multi - energy supply module selects the energy supply source according to the real - time environmental conditions and load demands. The multi - energy supply module includes a solar energy supply sub - module, a wind energy supply sub - module, an energy storage system, a traditional power grid access sub - module, and an intelligent switching sub - module. During operation, it performs priority power supply switching among solar energy, wind energy, the energy storage system, and the traditional power grid access based on the real - time output power, energy cost, and system stability requirements of the solar energy supply sub - module and the wind energy supply sub - module.

[0014] The data acquisition module adopts a distributed multi - sensor fusion architecture, automatically switches to standby sensors when a sensor fails, acquires the real - time data of the multi - energy supply module, including the real - time data of solar energy, wind energy, and the energy storage situation, and also acquires meteorological data, load demand data, and equipment status data.

[0015] The multi - source data fusion module fuses the data acquired by the data acquisition module. The multi - source data fusion module is built - in with a data quality assessment unit, which uses a fuzzy logic algorithm to perform real - time assessment on the quality of the acquired data. For low - quality data, a data repair mechanism is adopted.

[0016] The energy scheduling module performs energy resource scheduling according to the fused real - time load demand data and environmental change data. The energy scheduling module accelerates the learning process of energy scheduling through a quantum reinforcement learning algorithm combined with quantum computing, simulates complex and non - linear energy demand fluctuations through a chaotic system and optimizes the scheduling strategy, and optimizes the scheduling strategy and parameters through a deep neural network algorithm to improve the system response accuracy. The energy scheduling module has a simulation function and can simulate the system operation status under each season, weather, and electricity load scenario in a virtual environment.

[0017] The formula of the reinforcement learning algorithm is as follows: ; where is the Q - value of taking action in the current state , representing the expected return after executing action from state ; is the learning rate, representing the acceptance of new information and controlling the balance between new learning and old knowledge; is the immediate reward at the current time t, representing the feedback after executing the action; is the discount factor, representing the weight of future rewards; is the maximum Q - value of all possible actions in the next state , representing the maximum return that can be obtained in the future state.

[0018] The demand forecasting and response module predicts future energy load demands based on the fused data, combined with weather forecasts and historical load data, and adjusts the load response strategy according to the prediction results. The demand forecasting and response module captures the spatio-temporal dependence characteristics of energy demands through spatio-temporal deep learning algorithms, and adjusts the load response strategy according to the prediction results through adaptive deep reinforcement learning algorithms. Precise state estimation is carried out in complex environments through the extended Kalman filter algorithm to improve the prediction accuracy. The demand forecasting and response module can be connected to the intelligent home appliance control system, and send electricity regulation instructions to intelligent home appliances in advance according to the predicted load demands to achieve off-peak electricity consumption for home appliances.

[0019] The formula of the spatio-temporal deep learning algorithm is as follows: ; where is the output of the hidden layer at the current time representing the state of the network; is the activation function; is the weight matrix at the current time, representing the mapping relationship between the input data and the network state; is the hidden layer state at the previous time used to model the temporal dependence; is the input data at the current time.

[0020] The supply security assessment module evaluates the security of the current energy supply based on the fused data, and predicts potential supply interruption or power fluctuation risks. The game process between energies is simulated through the generative adversarial network algorithm, potential risks are warned according to the fluctuations of energy outputs, the collaborative and load distribution capabilities between each energy are improved through the swarm intelligence optimization algorithm, and the competition and cooperation strategies in energy allocation are further optimized through the adversarial learning algorithm. The supply security assessment module establishes an energy security risk warning model, and based on big data analysis and machine learning algorithms, identifies potential energy supply interruptions and power fluctuation risks 30 minutes in advance.

[0021] The energy storage and load balancing module dynamically manages energy storage nodes based on the fused data to balance the load and energy storage. The energy storage and load balancing module extracts high-dimensional features from environmental and energy data through the deep convolutional generative adversarial network algorithm, adjusts the energy storage strategy in real time according to external environmental changes through the adaptive environment perception algorithm, and processes data locally in real time and executes scheduling decisions through the edge computing optimization algorithm. The energy storage and load balancing module supports the access of distributed energy storage nodes, and realizes trusted data sharing and collaborative control between energy storage nodes through blockchain technology to ensure the efficient allocation of energy storage resources within the region.

[0022] The formula of the generative adversarial network is as follows: ; where is a sample output by the generator, and the generator generates data according to the latent variable ; is the judgment value of the discriminator for the sample , representing the probability that the sample is real; is the distribution of the latent space ; is the data distribution, representing the distribution of real samples.

[0023] The scheduling optimization module adjusts the energy supply constraint conditions and optimization objectives dynamically according to the environmental changes and load demands during real-time operation based on the current energy resource scheduling of the energy scheduling module, the output of each energy source, and the system operation indexes. The scheduling optimization module searches for the optimal scheduling strategy through the quantum genetic algorithm, represents and updates the relationship between each energy source through the dynamic knowledge graph, and adjusts the energy supply constraint conditions and optimization objectives dynamically according to the environmental changes and load demands during real-time operation through the adaptive network optimization algorithm; the scheduling optimization module has a visual interaction interface, and the operator can intuitively view the current scheduling strategy, the output of each energy source, and the system operation indexes through this interface, and can manually input the constraint conditions and optimization objectives at the same time.

[0024] The formula of the quantum genetic algorithm is as follows: ; where is the fitness of the individual , measuring the quality of the solution. The higher the fitness, the better the individual; is the -th component in the decision variable , representing a specific parameter in the solution; is the weight of the -th decision variable, representing the importance of this variable in the objective function; is the number of variables in the solution. Specific Embodiment 2 As Figure 1-2 shown, in the intelligent power supply management system based on multi-energy complementarity, in addition to the above-mentioned various modules and functions, the details of its actual application and system implementation are further refined, and the specific content is as follows: Intelligent Energy Switching and Dynamic Scheduling: In actual operation, the multi-energy supply module automatically selects the best energy supply source according to real-time weather conditions, energy demand, and system stability. For example, when sunlight is strong, the system will give priority to solar power supply; when the wind speed is high, the wind energy module may be preferred; in the case of insufficient sunlight and wind energy, the energy storage system will provide support according to the load demand or supplement the insufficient power through connection to the traditional power grid. The system will also adjust the priorities among solar energy, wind energy, energy storage, and grid connection in real time according to the change of load, so as to achieve efficient and reliable energy supply.

[0026] Intelligent Load Forecasting and Response: The demand forecasting and response module can predict the change of energy demand under daily or seasonal load fluctuations and make an early response by integrating a spatio-temporal depth adaptive reinforcement learning algorithm and combining meteorological forecasts and historical load data. For example, the system can predict that the energy demand will increase during a certain period and send scheduling instructions to intelligent household appliances, commercial equipment, etc. in advance to achieve load shifting and reduce the pressure during peak hours.

[0027] Environmental Adaptability and Energy Storage Optimization: The energy storage and load balancing module can analyze external environmental changes in real time and dynamically optimize the energy storage strategy according to factors such as weather changes and energy supply conditions through a deep convolutional generative adversarial network (DCGAN) and edge computing technology. When the battery storage is charging and discharging, the system will adjust the charging and discharging efficiency according to the health status of the energy storage system and environmental conditions, while reducing the negative impact that overcharging and over-discharging may have on the energy storage equipment. Through edge computing, the system can quickly process and respond to data locally to ensure the execution of fast and efficient decisions.

[0028] Supply Security Assessment and Optimization: The supply security assessment module can evaluate the security of the current energy supply and predict possible supply interruption or overload risks through a generative adversarial network (GAN) and swarm intelligence algorithm. For example, the system can monitor the operating status of each energy module in real time, evaluate the reliability of the current supply, and warn of potential risks according to the fluctuations of energy output. In addition, the swarm intelligence algorithm optimizes the coordination and load distribution among energies to improve the overall stability of the system.

[0029] Multi-source Data Fusion and Precise Scheduling: With the support of the multi-source data fusion module, the system can collect and analyze data from various aspects such as solar energy, wind energy, meteorology, energy storage, and load in real time. Through fuzzy logic and weighted fusion algorithms, the system can ensure data quality and consistency, thus improving the accuracy of scheduling decisions. The data quality assessment mechanism will monitor the integrity and accuracy of data collection in real time and automatically activate the data repair mechanism to ensure that the scheduling system is not affected by data errors.

[0030] Smart Grid and Regional Energy Management: In practical applications, the system can not only optimize the energy supply within a single building or area, but also exchange data and collaborate with the smart grid system or regional energy dispatch center. Through this collaborative work, the system can interact with the energy supply chains and power networks in other regions, improving the load balancing and emergency response capabilities of the regional power grid. For example, based on the changes in local load demand, the system can feedback the current energy storage capacity and power generation capacity to the grid, helping the grid adjust the access of external power sources and load distribution.

[0031] Intelligent Scheduling and User Participation: The scheduling optimization module of the system includes a visual interaction interface, through which operators can intuitively view the current scheduling strategies and the output of each energy source. When needed, operators can also manually input constraint conditions or optimization objectives, such as giving priority to ensuring a certain critical load or selecting the optimal energy scheduling strategy based on economic benefits. In addition, the system supports user participation and can perform customized scheduling based on the electricity consumption habits of household or commercial users, achieving personalized energy management.

[0032] Security and Transparency: In terms of energy storage and load balancing, the system supports the access of distributed energy storage nodes and realizes trusted data sharing and collaborative control among energy storage nodes through blockchain technology. Blockchain technology ensures the immutability and transparency of data in the energy management process, which not only helps the efficient allocation of energy storage resources but also prevents malicious interference and ensures the security of the system.

[0033] Application Examples: Household Smart Energy Management: In a smart home system, the system can, through interaction with the home appliance control system, adjust the usage time of home appliances such as air conditioners and water heaters according to the load demand prediction results, optimize the household electricity load, and reduce electricity bills.

[0034] Industrial Park Energy Management: In large industrial parks, the system optimizes the electricity usage strategies of factory production equipment, reduces the energy demand during peak hours, and stores electricity during off-peak hours through an energy storage system to provide stable power supply for the production process.

[0035] Off-grid Energy Supply: In some remote off-grid areas, the system can rely on the combined operation of solar and wind power generation and energy storage systems to ensure power supply within the region and adjust the priority supply of energy according to environmental changes to ensure the efficient use of energy.

[0036] Through these modules and applications, the intelligent power supply management system based on multi-energy complementarity can achieve intelligent scheduling and dynamic optimization of energy, thus significantly improving energy use efficiency, reducing carbon emissions, ensuring the stability and security of the system, and ultimately providing users with a more reliable, economical, and green power supply solution. Specific Embodiment 3 The following is a detailed analysis of the key algorithms mentioned in Example 1, including their core mathematical formulas and explanations: 1. Quantum Genetic Algorithm, Fitness Function: The fitness function is used to measure the quality of an individual (solution). In the quantum genetic algorithm, the fitness function helps determine which solutions are selected for crossover and mutation operations.

[0038] ; where is the fitness of the individual , measuring the quality of the solution. The higher the fitness, the better the individual; is the decision variable in the -th component, representing a specific parameter in the solution; is the weight of the -th decision variable, indicating the importance of this variable in the objective function; is the number of variables in the solution.

[0039] 2. Spatiotemporal Depth Adaptive Reinforcement Learning Load Demand Prediction Optimization Algorithm, Spatiotemporal Deep Learning Model is used to capture the spatiotemporal dependence of load demand and processes the input data through a neural network: Formula: ; where is the output of the hidden layer at the current time , representing the state of the network; is the activation function, such as ReLU or Sigmoid, applied to the output of the network; is the weight matrix at the current time, representing the mapping relationship between the input data and the network state; is the hidden layer state at the previous time , used to model the temporal dependence; is the input data at the current time.

[0040] 3. Reinforcement Learning Q-learning Update Formula, Q-learning is a method in reinforcement learning used to continuously adjust the decision-making strategy according to experience. Formula: ; where is the Q-value of taking action in the current state , representing the expected return after executing action from state ; ​is the learning rate, representing the acceptance of new information. It controls the balance between new learning and old knowledge; is the immediate reward at the current time t, representing the feedback after executing an action; is the discount factor, representing the weight of future rewards; is the next state is the maximum Q value of all possible actions in, representing the maximum return that can be obtained in the future state.

[0041] 4. Generative Adversarial Network (GAN), the Generative Adversarial Network (GAN) is trained adversarially by a generator and a discriminator. The generator generates samples, and the discriminator judges the authenticity of the samples.

[0042] ; where is the sample output by the generator. The generator generates data according to the latent variable ; is the judgment value of the discriminator for the sample , representing the probability that the sample is real; is the distribution of the latent space , usually a simple distribution (such as Gaussian distribution); is the data distribution, representing the distribution of real samples.

[0043] 5. Unsupervised Deep Convolutional Generative Adversarial Network (DCGAN), the convolutional layer is used to extract features from data and is widely used in the Generative Adversarial Network (DCGAN). Formula: ; where Y is the output feature map of the convolutional layer; W is the convolutional kernel, defining the feature extraction rule for the local area; X is the input data (such as an image); b is the bias term, used to adjust the convolutional output; f is the activation function (such as ReLU or LeakyReLU), used to increase non-linearity.

[0044] 6. Chaos System Optimization Algorithm, the Chaos System Optimization Algorithm performs optimization by simulating chaotic behavior and is applicable to complex and non-linear optimization problems.

[0045] Formula: ; where is the current state, representing a certain variable of the system; is the control parameter, affecting the intensity of chaotic behavior; is the next state.

[0046] 7. Multi - energy complementary scheduling optimization algorithm of adaptive quantum search and dynamic knowledge network. The quantum search optimization algorithm uses quantum superposition and quantum interference to efficiently search for the optimal solution, especially suitable for the scheduling optimization problem of multi - energy complementarity.

[0047] Formula: ; Where is the quantum state, representing the superposition state in the search space; is the coefficient, representing the probability amplitude of the th state in the quantum state; is the ground state, representing a possible solution in the search space; Through quantum interference, the probability amplitude will be continuously adjusted according to the fitness value to increase the probability of the optimal solution.

[0048] Quantum operation and interference formula: ; Where is the quantum gate operation, adjusting the state of the qubit; is the control coefficient, used to adjust the superposition ratio of the qubit; is the basic state of the qubit. Specific embodiment 4 As Figure 1-2 shown, the following is the specific application logic step description of each module and algorithm in the intelligent power supply management system based on multi - energy complementarity: 1. The multi - energy supply module is responsible for managing the energy sources in the system, including the access of solar energy, wind energy, energy storage system and traditional power grid. This module ensures that the most suitable energy supply source is preferentially selected according to real - time environmental conditions and demands.

[0050] Dynamic scheduling optimization algorithm: Based on the real - time output power of current solar energy and wind energy, energy cost and system stability requirements, this algorithm automatically judges whether to supply power from the energy storage system or traditional power grid. If the output of solar energy or wind energy is too low, the system will automatically switch to the energy storage system or grid power supply to ensure power supply stability.

[0051] Step description: Data input, obtain the output power of solar energy and wind energy, energy storage power, and grid load information.

[0052] Calculation and decision - making, use the scheduling algorithm to judge the priority power supply source based on energy output, cost and demand situation.

[0053] Control output, dynamically switch to the appropriate energy source to ensure continuous power supply.

[0054] 2. The data acquisition module monitors the operating status of solar energy, wind energy, and energy storage systems in real time through a multi-sensor system, and simultaneously collects meteorological data, demand data, and equipment status information. The system reliability is improved through a distributed architecture.

[0055] Algorithms: Kalman filter algorithm and fuzzy logic data repair; Application logic: When a sensor fails, the system can automatically switch to a backup sensor for data acquisition.

[0056] Step description: Data acquisition, where sensors collect data from each sub-module.

[0057] Fault detection, detecting sensor faults and switching to backup sensors.

[0058] 3. The multi-source data fusion module fuses data from multiple data sources (such as weather data, equipment status, production data, etc.) to generate comprehensive decision-making basis for energy management.

[0059] Algorithms: Fuzzy logic and weighted fusion algorithm.

[0060] Application logic: Using the fuzzy logic algorithm to evaluate the data quality in real time, and for low-quality data, adopting a data repair mechanism.

[0061] Step description: Data fusion, fusing the processed data with information from other modules to generate decision-making basis for energy scheduling.

[0062] Data quality evaluation, using the fuzzy logic algorithm to evaluate the data quality in real time, and starting the repair mechanism for low-quality data.

[0063] 4. The energy scheduling module is used to schedule various energy resources and make dynamic adjustments according to real-time load demands and environmental changes to optimize the response accuracy of the system.

[0064] Algorithms: Quantum reinforcement learning and chaotic system simulation; Application logic: The quantum reinforcement learning algorithm accelerates the energy scheduling process, while the chaotic system model can simulate complex energy demand fluctuations, thereby optimizing the scheduling strategy.

[0065] 5. The demand forecasting and response module is responsible for forecasting future energy load demands and making response adjustments according to the forecasting results, such as sending off-peak electricity usage instructions to smart home appliances.

[0066] Algorithms: Spatiotemporal deep learning and adaptive deep reinforcement learning; Application logic: Capturing the spatiotemporal dependence of energy demands through spatiotemporal deep learning, and combining with adaptive deep reinforcement learning to adjust response strategies and optimize load forecasting.

[0067] 6. The supply security assessment module assesses the security of energy supply, predicts the risks of potential supply interruptions or power fluctuations, and takes countermeasures in advance.

[0068] Algorithms: Generative Adversarial Network (GAN) and swarm intelligence optimization algorithm; Application logic: GAN is used to simulate the game process between energies and optimize the scheduling and cooperation between energies. The swarm intelligence algorithm enhances the collaborative ability between energy resources to ensure supply security.

[0069] 7. The energy storage and load balancing module is responsible for balancing the load and energy storage. Through the dynamic management of energy storage nodes, it ensures that the system can remain stable during load fluctuations.

[0070] Algorithms: Unsupervised Deep Convolutional Generative Adversarial Network (DCGAN) and edge computing algorithm; Application logic: Use DCGAN for high-dimensional feature extraction, and at the same time optimize real-time scheduling and data processing through edge computing to ensure the system response speed and accuracy.

[0071] 8. The scheduling optimization module is responsible for scheduling various energy resources in a multi-energy system to meet the load demand and optimize the energy use efficiency. Use the quantum genetic algorithm to search for the optimal scheduling strategy, and at the same time update the power relationship in combination with the dynamic knowledge graph.

[0072] Algorithms: Quantum genetic algorithm and dynamic knowledge graph; Application logic: The quantum genetic algorithm can efficiently search for the optimal scheduling strategy. Represent the relationship between energy resources through the dynamic knowledge graph and update the cooperation mode between power supplies in real time. Specific Embodiment 5 As Figure 1-2 shown, the following is the detailed hardware composition and hardware description of each module in Embodiment 1: 1. Hardware composition of the multi-energy supply module: Solar photovoltaic panels, used to collect solar energy and convert it into electrical energy. The solar photovoltaic panels are composed of multiple photovoltaic units, with high energy conversion efficiency, and are equipped with MPPT (Maximum Power Point Tracking) technology to maximize the power output.

[0074] Wind turbines, used to convert wind energy into electrical energy. The wind turbines include wind turbines, generator sets, and frequency converters. The rotation speed of the wind turbine is related to the wind speed, and the frequency converter adjusts the current output to match the grid demand.

[0075] Energy storage system (battery pack), used to store excess electrical energy. The energy storage system consists of a high-capacity lithium battery pack and is equipped with an intelligent battery management system (BMS) to monitor the battery's power, temperature, and health status, ensuring efficient and safe energy storage.

[0076] Grid connection device (inverter), used to connect to the traditional power grid. The inverter converts direct current (DC) into alternating current (AC) and controls the output of the current to match the power grid.

[0077] 2. Hardware composition of the data acquisition module: Multi-sensor module, used to collect various physical quantity data (such as temperature, humidity, voltage, current, wind speed, etc.). The multi-sensor module includes temperature and humidity sensors, light sensors, barometric pressure sensors, current and voltage sensors, etc., integrated on a data acquisition circuit board, supporting high-speed data acquisition and outputting to the central control system.

[0078] Data acquisition control unit (microcontroller), used to collect and process sensor data. Based on a high-performance microcontroller (such as the ARM Cortex series), it has real-time data processing capabilities, can collect and store sensor data, and simultaneously transmit it to the main control unit of the system.

[0079] Wireless communication module (such as Wi-Fi, ZigBee, LoRa), used for remote data transmission. Adopting a low-power wireless communication module, it is used to transmit the collected data to the cloud or local control center, supporting long-distance data transmission and wireless networking.

[0080] 3. Hardware composition of the multi-source data fusion module: Data fusion unit, used to receive and fuse data from different sensors and devices. This unit is equipped with a high-speed processor, such as an FPGA or a high-performance ARM chip, which can execute data fusion algorithms and process data from different sources in real time.

[0081] Storage device (hard disk / SSD), used to store raw data and processed data. The system is equipped with a high-capacity and high-speed storage device, such as a solid-state drive (SSD), to store raw data and fused data for subsequent analysis and processing.

[0082] Data quality assessment unit, used to evaluate and repair data quality. It includes a fuzzy logic module and a data repair algorithm, which can automatically identify and repair low-quality data.

[0083] 4. Hardware composition of the energy scheduling module: Scheduling optimization processing unit, used to execute energy scheduling algorithms. Equipped with a high-performance computing platform, such as a high-performance computer based on a multi-core processor or a cloud computing platform, it supports the efficient calculation of quantum reinforcement learning and chaos system optimization algorithms.

[0084] Simulation and emulation hardware, used to simulate the system operation status. This module is equipped with a virtual simulation server, supporting the simulation of energy scheduling under different seasons, weather, and electricity load scenarios.

[0085] Real-time monitoring unit, used to monitor the energy scheduling effect in real time. Equipped with a real-time data acquisition system and display device, it shows the system operation status and the output of each energy source, and supports the real-time display of scheduling results.

[0086] 5. Hardware composition of the demand forecasting and response module: Forecasting and response processing unit, used to execute load demand forecasting and response strategy optimization. Equipped with a deep learning acceleration card (such as NVIDIA Jetson) or cloud GPU, it can support the real-time execution of spatio-temporal deep learning and reinforcement learning algorithms.

[0087] Household appliance control unit, used to control the electricity consumption behavior of intelligent household appliances. The intelligent household appliance control unit includes intelligent sockets, switches, sensors, etc., and is linked with the forecasting module through a wireless communication protocol (such as ZigBee, Wi-Fi).

[0088] Wireless communication module, used to send response signals to household appliances. Equipped with a low-power wireless communication module (such as ZigBee, LoRa), it ensures real-time connection with intelligent household appliances and sends load regulation instructions.

[0089] 6. Hardware composition of the supply security assessment module: Security assessment processing unit, used to evaluate the security of energy supply. This unit uses a high-performance computing platform, such as dedicated hardware based on FPGA or ASIC, supporting the efficient calculation of generative adversarial networks (GAN) and swarm intelligence optimization algorithms.

[0090] Early warning system hardware, used to send energy security warnings to the control center. Equipped with an efficient alarm device and communication equipment (such as LED indicator lights, voice alarm systems, SMS or email notification modules) for early warning of possible security risks in the system.

[0091] 7. Hardware composition of the energy storage and load balancing module: Battery management system (BMS), used to manage and monitor the battery energy storage status. The battery management system includes battery units, temperature sensors, voltage / current sensors, BMS chips, etc., ensuring the efficient operation and safe charging and discharging of the battery.

[0092] Energy storage node controller, used to control the charging and discharging process of energy storage. This controller integrates an adjustable current source, control circuit, and communication interface, used to regulate the charging and discharging process of the energy storage system.

[0093] A load regulation control unit for regulating system load and power distribution. An integrated power dispatching control module that, through smart meters and smart switches, adjusts the load distribution in real time to ensure load balance.

[0094] 8. Hardware components of the scheduling optimization module: A processing unit (CPU / GPU) for executing scheduling optimization algorithms. This module is equipped with a high-performance CPU or GPU, such as the Intel Xeon series or NVIDIA Tesla GPU, with powerful parallel computing capabilities, supporting the efficient execution of algorithms such as quantum genetic algorithms and dynamic knowledge graphs.

[0095] A visualization interface display for showing the current scheduling strategy and energy output. Equipped with a touch screen display (such as a 21-inch high-definition display), it can display energy supply, demand prediction, and scheduling strategies in real time and provide a manual input interface.

[0096] A communication module for communicating with other modules. This module uses a high-speed network interface (such as Ethernet or fiber optic) to interact with modules such as data acquisition and demand prediction.

[0097] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variation thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising a reference structure" does not exclude the existence of additional identical elements in the process, method, article or device comprising the element.

[0098] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent power supply management system based on multi - energy complementarity, characterized in that: The multi - energy supply module selects the energy supply source according to the real - time environmental conditions and load demands. The data acquisition module acquires the real - time data of the multi - energy supply module, and also acquires meteorological data, load demand data, and equipment status data. The multi - source data fusion module fuses the data acquired by the data acquisition module. The energy scheduling module conducts energy resource scheduling according to the fused real - time load demand data and environmental change data. The demand prediction and response module predicts the future energy load demand based on the fused data, combined with weather forecasts and historical load data, and adjusts the load response strategy according to the prediction results. The supply security assessment module assesses the security of the current energy supply based on the fused data, and predicts potential supply interruption or power fluctuation risks. The energy storage and load balancing module dynamically manages the energy storage nodes according to the fused data, and balances the load and energy storage. The scheduling optimization module dynamically adjusts the energy supply constraint conditions and optimization objectives according to the current energy resource scheduling of the energy scheduling module, the output of each energy source, and the system operation indicators during real - time operation according to environmental changes and load demands.

2. The intelligent power supply management system based on multi-energy complementarity according to claim 1, characterized in that: The multi - energy supply module includes a solar energy supply sub - module, a wind energy supply sub - module, an energy storage system, a traditional power grid access sub - module, and an intelligent switching sub - module; during operation, it performs priority power supply switching among solar energy, wind energy, the energy storage system, and traditional power grid access according to the real - time output power, energy cost, and system stability requirements of the solar energy supply sub - module and the wind energy supply sub - module.

3. The intelligent power supply management system based on multi-energy complementarity according to claim 1, characterized in that: The data acquisition module adopts a distributed multi - sensor fusion architecture and acquires the real - time data of the multi - energy supply module, including the real - time data of solar energy, wind energy, and energy storage status.

4. The intelligent power supply management system based on multi-energy complementarity according to claim 1, wherein: The multi - source data fusion module is built - in with a data quality assessment unit, which uses a fuzzy logic algorithm to evaluate the data quality in real - time. For low - quality data, a data repair mechanism is adopted.

5. The intelligent power supply management system based on multi-energy complementarity according to claim 1, characterized in that: The energy scheduling module accelerates the learning process of energy scheduling through a quantum reinforcement learning algorithm combined with quantum computing, simulates complex and non - linear energy demand fluctuations through a chaotic system and optimizes the scheduling strategy, and optimizes the scheduling strategy and parameters through a deep neural network algorithm. The formula of the reinforcement learning algorithm is as follows: ； where is the current state under which an action is taken and represents the Q-value of taking an action from state by performing the action and the expected return afterwards; is the learning rate, representing the acceptance of new information and controlling the balance between new learning and old knowledge; is the immediate reward at the current time t, representing the feedback after performing the action; is the discount factor, representing the weight of future rewards; is the next state and the maximum Q-value of all possible actions in it, representing the maximum return that can be obtained in the future state.

6. The intelligent power supply management system based on multi-energy complementarity according to claim 1, characterized in that: The demand prediction and response module captures the spatio - temporal dependence characteristics of energy demand through a spatio - temporal deep learning algorithm, adjusts the load response strategy according to the prediction results through an adaptive deep reinforcement learning algorithm; and performs accurate state estimation in a complex environment through an extended Kalman filter algorithm. The formula of the spatio - temporal deep learning algorithm is as follows: ； wherein is the output of the hidden layer at the current moment, representing the state of the network; is the activation function; is the weight matrix at the current moment, representing the mapping relationship between the input data and the network state; is the hidden layer state at the previous moment, used to model temporal dependencies; is the previous moment ; and is the input data at the current moment.

7. The intelligent power supply management system based on multi-energy complementarity according to claim 1, characterized in that: The supply security assessment module simulates the game process among energies through a generative adversarial network algorithm, warns of potential risks according to the fluctuations of energy output, improves the coordination and load distribution capabilities among each energy through a swarm intelligence optimization algorithm, and further optimizes the competition and cooperation strategies in energy allocation through an adversarial learning algorithm.

8. The intelligent power supply management system based on multi-energy complementarity according to claim 1, characterized in that: The energy storage and load balancing module extracts high-dimensional features from environmental and energy data through the deep convolutional generative adversarial network algorithm, adjusts the energy storage strategy in real time according to external environmental changes through the adaptive environmental perception algorithm, and processes data locally in real time and executes scheduling decisions through the edge computing optimization algorithm; The generative adversarial network formula is as follows: ； wherein is a sample output by the generator, and the generator generates data according to the latent variable ; is the judgment value of the discriminator for the sample , representing the probability that the sample is real; is the distribution of the latent space ; is the data distribution, representing the distribution of real samples.

9. The intelligent power supply management system based on multi-energy complementarity according to claim 1, characterized in that: The scheduling optimization module searches for the optimal scheduling strategy through the quantum genetic algorithm, represents and updates the relationship between each type of energy through the dynamic knowledge graph, and dynamically adjusts the energy supply constraint conditions and optimization objectives according to environmental changes and load demands during real-time operation through the adaptive network optimization algorithm; The quantum genetic algorithm formula is as follows: ； where is the fitness of an individual , which measures the quality of the solution. The higher the fitness, the better the individual; is the -th component in the decision variable , representing a specific parameter in the solution; is the weight of the -th decision variable, indicating its importance in the objective function; is the number of variables in the solution.

10. The intelligent power supply management system based on multi-energy complementarity according to claim 1, characterized in that: It also includes a smart grid and regional energy management module for data exchange and collaboration with the smart grid system or regional energy dispatch center.

Images