Power supply system scheduling method and power supply system thereof

Abstract

The invention relates to the technical field of power dispatching, and discloses a power supply system dispatching method and a power supply system thereof, and the power supply system comprises a monitoring module, a dispatching module, a protection module, a data prediction module, an energy storage module, a remote control and communication module, an artificial intelligence and adaptive dispatching module, a power quality monitoring module and an intelligent load management module. The power quality monitoring module collects harmonic wave and voltage fluctuation data in real time, gives an alarm immediately when the harmonic wave and voltage fluctuation data exceed the standard, and pushes governance suggestions; and the scheduling module is matched to adjust the non-linear load operation time, so that quality problems are reduced from the source. Scene features are identified through deep learning, and an emergency strategy is automatically generated; and the scheduling module receives feedback in real time and corrects the strategy. The protection module accurately locates a fault point through multi-sensor cross validation and automatically cuts off a fault loop; the remote control module pushes fault information to the operation and maintenance terminal in real time, remote diagnosis is supported, and the on-site inspection frequency is reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power dispatching, in particular to a power supply system dispatching method and a power supply system thereof. BACKGROUND

[0002] Power dispatching is the core link of ensuring the safe, stable and efficient operation of the power system. Simply put, it is to ensure that the power generated at the power generation end can accurately match the power demand at the user end through "overall allocation, dynamic adjustment and emergency treatment", while taking into account economy and reliability.

[0003] For example, in the prior art, a power supply system dispatching method and a power supply system thereof (publication number CN118920599A) are disclosed. The system includes: an application monitoring module for monitoring the power consumption equipment and obtaining power consumption; based on the data of the monitoring module, a data prediction module processes data through an RBF radial basis neural network regression prediction model; based on the monitoring module and the data prediction module, a dispatching module dispatches the power supply system scheme; after the dispatching scheme based on the prediction data is used, the early warning module is based on the real-time monitoring of the change of the power consumption of the power consumption equipment, and in the case that the power supply of the factory area cannot be met, the early warning module performs early warning and uses energy storage equipment to supply power. The present application realizes the dispatching of a complex power supply system containing multiple distributed power sources, and the dispatching method is more reasonable and effective, which can effectively control the power generation cost.

[0004] However, the traditional system relies on manual experience dispatching, which is difficult to cope with sudden load fluctuations, and the power quality is out of standard, which damages precision equipment. Manual inspection, fault positioning is difficult, recovery time is long, and the impact of the accident is easy to expand. Therefore, the present application provides a power supply system dispatching method and a power supply system thereof to solve the problems in the above background art. SUMMARY

[0005] The present application aims to provide a power supply system dispatching method and a power supply system thereof to solve the problems in the above background art.

[0006] To achieve the above-mentioned purpose, the present application provides the following technical solutions: A power supply system, comprising: The monitoring module is responsible for real-time tracking of each link of the power supply system, including current, voltage, frequency, load, temperature and other parameters. Through data acquisition equipment, the monitoring module can comprehensively monitor the system state in real time. The dispatching module is responsible for controlling and optimizing the allocation of power to ensure that power is reasonably dispatched between various loads to avoid overload or power shortage. A protection module that automatically starts cutting off or isolating the faulty part when an overload, short circuit, overvoltage, or other abnormal situation occurs in the power system, preventing damage to equipment or causing more serious accidents; A data prediction module that predicts future power demand and load conditions based on historical data and real-time monitoring data through algorithms; An energy storage module that stores excess power, especially when power demand is low or power supply is excessive, to ensure the stability of power supply; A remote control and communication module for remote monitoring, diagnosis, and control of the power supply system through wireless or wired networks; An artificial intelligence and adaptive scheduling module that automatically optimizes power scheduling through artificial intelligence and big data analysis, improves the adaptive ability of the system, and responds to load fluctuations and external environmental changes in a timely manner; A power quality monitoring module responsible for monitoring power quality issues in the power grid, such as voltage fluctuations, harmonics, frequency deviations, and flicker; An intelligent load management module that balances and optimizes the use of various loads through data analysis and intelligent scheduling to ensure efficient operation of the system.

[0007] As a further scheme of the present application: the monitoring module includes: Data collection: determine the key parameters to be monitored according to the system architecture, and deploy corresponding sensors at key locations such as power generation end, transmission line, distribution node, terminal load, and energy storage equipment, collect current, voltage, frequency, load, temperature, and other data through smart meters, sensors, and other devices; Data transmission: real-time transmission of collected data to the central monitoring system or cloud platform; Data processing: real-time processing and analysis of data to detect abnormalities such as high or low voltage, current fluctuations, and trigger immediate warnings; Data storage and recording: all monitoring data will be saved regularly for subsequent analysis and backtracking; Real-time state visualization: the cloud platform or monitoring terminal converts data into intuitive charts to display the running state of each link of the system in real time for operation and maintenance personnel to monitor.

[0008] As a further scheme of the present application: the scheduling module includes: Load analysis: receive real-time power data of each terminal load uploaded by the monitoring module, classify and count by load priority and type, form a "load state matrix" containing real-time power, priority, adjustability, and health of each load, providing a basis for subsequent allocation; Load forecasting: Based on historical data and real-time information, future load changes are predicted. Short-term forecasting uses the LSTM (Long Short-Term Memory) model, while medium-term forecasting combines the ARIMA model with a rule base. Power allocation: A two-tier model based on priority and capacity constraints. The first tier guarantees a fixed quota for primary loads, while the second tier allocates the remaining capacity to secondary and tertiary loads according to the predicted demand multiplied by the adjustable coefficient. Load balancing: By using intelligent interconnection switches, part of the load on overloaded lines is transferred to lightly loaded lines, and power is distributed between different areas and equipment by phase sequence switching switches to avoid overload of a single link; Optimized scheduling aims to minimize line losses, reduce electricity purchase costs, and maximize the absorption rate of distributed energy resources. The optimal solution is obtained by using the particle swarm optimization algorithm to minimize energy consumption and improve system efficiency. Electricity price peak-valley difference period: During off-peak hours, the energy storage module is instructed to charge at full power, and during peak hours, the stored power is released first to reduce the cost of electricity purchase; During periods of peak solar power generation: When solar power output exceeds local load, excess electricity is prioritized for storage to increase self-consumption rate and reduce curtailment losses.

[0009] As a further embodiment of the present invention: the protection module includes: Fault feature library establishment: Preset characteristic parameters of typical faults to form a fault judgment threshold library; Fault detection: Real-time monitoring of parameters such as current, voltage, and frequency in the power system, and real-time comparison with the fault feature database to detect whether overload, short circuit, overvoltage, or other problems have occurred; Fault location and isolation: Through cross-verification of multi-sensor data, the module or line where the fault occurs is accurately located, and the circuit breaker is immediately triggered to trip and the fuse is blown to cut off the fault circuit. At the same time, the fault area is isolated to prevent the fault from spreading. Fault Recovery Record: After troubleshooting, determine whether the system is ready for recovery through manual confirmation or automatic detection. Restore power supply remotely or locally, and record the fault type, occurrence time, handling process, and equipment status. Store the records in the fault database to provide a basis for subsequent operation and maintenance.

[0010] As a further embodiment of the present invention: the energy storage module includes: Energy storage status monitoring: Real-time collection of key parameters of energy storage equipment, such as remaining capacity, charging and discharging current, individual battery voltage, temperature, and cycle count, to determine the health status of the equipment and store excess power in batteries or other energy storage devices when power demand is low or power supply is excessive. Energy release: When power demand is high, electricity is released from the energy storage device to supplement the system power supply; when power demand is low, the charging mode can be activated. Energy storage monitoring: During charging, the charging current is controlled by the charger to avoid large current surges; if an abnormal voltage of a single battery cell is detected, the automatic balancing circuit is activated to balance the state of each battery. During discharge, the inverter converts the DC power into AC power synchronized with the power grid, controls the discharge power, and maintains stable output voltage and frequency.

[0011] As a further embodiment of the present invention: the remote control and communication module includes: Communication network setup and protocol adaptation: terminal devices and edge nodes use short-distance communication, edge nodes and cloud platforms use long-distance communication, sensor data uses the MQTT protocol, and control commands use Modbus or DL / T645. Remote monitoring and transmission: The edge node packages and encrypts the real-time data and device status of the monitoring module, and uploads them to the remote monitoring platform through the communication network to ensure data security; Remote diagnostics: The platform collects data and logs to perform fault analysis and diagnosis, automatically diagnoses potential equipment faults, and pushes early warning information to maintenance personnel; Remote control: Maintenance personnel initiate control commands through the monitoring platform. The commands are encrypted and sent to the control unit of the corresponding device. After receiving the command, the device first verifies the legality of the command, then executes the operation, and feeds back the execution result to the platform, forming a closed loop.

[0012] As a further aspect of the present invention: the artificial intelligence and adaptive scheduling module includes: Big data feature learning and data acquisition: Collect and process real-time data from the system, construct datasets, and learn the correlation patterns between data through deep learning models; Model building: Real-time data from the monitoring and prediction modules are accessed, and the current operating scenario is identified and its characteristics are determined through an AI model; Adaptive scheduling strategy: Based on the current scenario and learned patterns, the AI ​​model automatically generates an optimized scheduling scheme; Adaptive scheduling optimization: The generated scheduling instructions are sent to the scheduling module and energy storage module for execution, while the execution effect is monitored in real time. If the actual result deviates from the expected result by more than a threshold, the model adjusts the strategy parameters through a reinforcement learning feedback mechanism to achieve self-optimization.

[0013] As a further embodiment of the present invention: the power quality monitoring module includes: Power quality data acquisition: Real-time acquisition of data such as voltage fluctuations, frequency changes, and harmonics in the power grid, and recording of events and their durations; Quality assessment: Based on the collected data, it is compared with national standards to determine whether the standards are exceeded; Anomaly detection: When power quality is abnormal, such as excessive voltage fluctuations or harmonic exceedances, an alarm will be triggered immediately and corresponding measures will be taken. Quality Reporting and Traceability: Generates power quality reports for operators to analyze and optimize, and pushes governance suggestions to the scheduling module, and tracks the governance effects.

[0014] As a further aspect of the present invention: the intelligent load management module includes: Load identification: Real-time monitoring of the power demand of each load, identifying peak and off-peak periods of load usage; Load operation monitoring: Combines data from monitoring modules to track the operating status of each load in real time and identify inefficiently operating loads; Load optimization: Based on the principle of "efficiency first + demand matching", an optimization plan is formulated. For adjustable loads: Automatically adjusts according to ambient temperature; Interruptible loads: postpone operation during periods of power shortage to stagger power consumption; Inefficient load: Push replacement suggestions.

[0015] Energy-saving control: Optimization strategies are executed through intelligent controllers to remotely adjust load operating parameters, evaluate the effectiveness of the strategies, and continuously iterate and optimize the solution. A power supply system dispatching method includes the following steps: S1: The monitoring module collects power supply parameters in real time through sensors deployed at the power generation end, transmission lines, load nodes, etc. After data preprocessing, it is uploaded to the central monitoring platform through the remote control and communication module, and simultaneously synchronized to the scheduling module, data prediction module, and artificial intelligence and adaptive scheduling module. S2: Based on real-time power data, analyze and classify the loads of each terminal, and generate a load status matrix according to the load's priority, type, adjustability and other characteristics to provide a basis for subsequent power allocation; S3: Based on historical data and real-time monitoring data, the LSTM model is used for short-term load forecasting, and the ARIMA model is combined with the rule base for medium-term forecasting. Power is allocated according to the forecast results and priorities. The first-level load is guaranteed its fixed quota, and the remaining capacity is allocated to the second and third-level loads according to the forecast demand. S4: Through intelligent scheduling, some of the load on overloaded lines is transferred to lightly loaded lines to achieve load balancing; By employing optimization algorithms such as particle swarm optimization, line losses can be reduced, power procurement costs can be optimized, and the overall system efficiency can be improved. By combining factors such as peak and off-peak electricity prices and photovoltaic power output, the optimal scheduling of energy storage and power distribution can be achieved; S5: The data prediction module combines historical data, real-time monitoring data, and related factors, and uses an LSTM model to make short-term predictions, outputting minute-level load fluctuation ranges. The ARIMA model and rule base are used for mid-term prediction, and the load peak value is output in different time periods. The prediction results are synchronized to the scheduling module and the artificial intelligence and adaptive scheduling module. S6: The artificial intelligence and adaptive scheduling module accesses monitoring data, prediction results, and historical optimization cases, and identifies the current scenario through a deep learning model; During periods of high solar power generation: Prioritize storing excess electricity in energy storage systems to increase self-consumption rate to over 90%. During peak electricity price periods: Maximize energy storage discharge to reduce electricity purchases from the grid; When the load fluctuates: it is recommended to dynamically adjust the range of adjustable loads (such as air conditioners); S7: The scheduling module performs allocation based on a two-layer model of priority and capacity constraints. First layer: Guarantee the fixed quota for primary load; Second tier: Allocate the second and third tier loads within the remaining capacity according to the predicted demand multiplied by the adjustable factor; S8: The protection module compares the monitoring data with the fault feature database in real time. Once an overload, short circuit or other abnormality is detected, the circuit breaker is immediately triggered to trip, the fault area is isolated, and a fault signal is sent through the remote control module. After troubleshooting, the protection module confirms that the insulation is qualified, the dispatch module gradually restores the power supply to the load, and the protection module records the fault details to the database for subsequent operation and maintenance optimization. S9: The scheduling module evaluates the scheduling effect every hour: calculates indicators such as line loss rate (target ≤ 5%), energy storage utilization rate (target ≥ 70%), and load balance (target ≥ 85%), and compares them with the optimization targets; If the deviation exceeds the threshold (e.g., absorption rate < 80%), it is fed back to the artificial intelligence and adaptive scheduling module, and the model parameters are corrected through reinforcement learning; The intelligent load management module iteratively optimizes strategies to continuously improve system efficiency. S10: The remote control and communication module transmits real-time scheduling data, load status, and fault information to the operation and maintenance terminal via 5G / fiber optics. The data is then visualized in the form of current curves, load heat maps, etc., enabling the formulation of optimized scheduling strategies, the execution of energy-saving control, and the optimization of load utilization efficiency.

[0016] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention tracks the line load rate in real time through a monitoring module. Once the load approaches a threshold, the scheduling module transfers the overloaded load via an intelligent interconnection switch. If a power supply shortage occurs, the energy storage system starts discharging to replenish the power and prevent power outages. 2. The power quality monitoring module of this invention collects harmonic and voltage fluctuation data in real time, and immediately alarms and pushes governance suggestions when the data exceeds the standard; the scheduling module works together to adjust the running time of nonlinear loads to reduce quality problems from the source.

[0017] 3. The AI ​​module of this invention identifies scene features through deep learning and automatically generates emergency strategies; the scheduling module receives feedback in real time and corrects the strategies, such as dynamically adjusting the adjustment range of adjustable load when the load fluctuates to maintain system balance.

[0018] 4. The intelligent load management module of this invention dynamically adjusts the temperature of adjustable loads and staggers the operation of interruptible loads; at the same time, it pushes suggestions for replacing inefficient loads to reduce energy waste from the load side.

[0019] 5. The protection module of this invention accurately locates the fault point through cross-verification by multiple sensors and automatically cuts off the fault circuit; the remote control module pushes fault information to the operation and maintenance terminal in real time, supports remote diagnosis, and reduces the frequency of on-site inspections. Detailed Implementation

[0020] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. 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 of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0021] Please refer to the embodiments of the present invention, a power supply system scheduling method and a power supply system thereof, comprising: The monitoring module is responsible for tracking all aspects of the power supply system in real time, including parameters such as current, voltage, frequency, load, and temperature. Through data acquisition devices (such as smart meters and sensors), the monitoring module can perform comprehensive real-time monitoring of the system status. The scheduling module is responsible for controlling and optimizing power distribution, ensuring that power is rationally distributed among various loads to avoid overload or insufficient power supply. The protection module is used to detect abnormal conditions such as overload, short circuit, and overvoltage in the power system and automatically disconnect or isolate the faulty part to avoid damage to equipment or cause more serious accidents. The data prediction module, based on historical data and real-time monitoring data, uses algorithms (such as machine learning, time series analysis, etc.) to predict future electricity demand and load conditions. Energy storage modules are used to store excess electricity, especially when electricity demand is low or there is a surplus of supply, to ensure the stability of the power supply. The remote control and communication module is used for remote monitoring, diagnosis and control of the power supply system via wireless or wired networks; The Artificial Intelligence and Adaptive Scheduling Module is used to automatically optimize power dispatch through artificial intelligence and big data analysis, improve the system's adaptability, and respond promptly to load fluctuations and changes in the external environment. The power quality monitoring module is responsible for monitoring power quality issues in the power grid, such as voltage fluctuations, harmonics, frequency deviations, flicker, etc. The intelligent load management module uses data analysis and intelligent scheduling to balance and optimize the use of various loads, ensuring the efficient operation of the system.

[0022] Specifically, the monitoring module of this invention includes: Data Acquisition: Based on the system architecture, determine the key parameters to be monitored (current, voltage, frequency, load power, equipment temperature, ambient humidity, etc.), and deploy corresponding sensors (such as Hall current sensors, voltage transformers, temperature transmitters, smart meters, etc.) at key locations such as the power generation end, transmission lines, distribution nodes, terminal loads, and energy storage devices. Collect data such as current, voltage, frequency, load, and temperature through smart meters, sensors, and other devices. Data transmission: The collected data is transmitted to the central monitoring system or cloud platform in real time; Data processing: Real-time processing and analysis of data to detect anomalies, such as excessively high or low voltage, current fluctuations, etc., and trigger immediate warnings; Data storage and recording: All monitoring data will be saved regularly for subsequent analysis and retrospection; Real-time status visualization: The cloud platform or monitoring terminal transforms the data into intuitive charts (such as current curves, load distribution diagrams, and temperature heat maps) to display the real-time operating status of each part of the system for operation and maintenance personnel to monitor.

[0023] Specifically, the scheduling module of this invention includes: Load analysis: Receive real-time power data of each terminal load uploaded by the monitoring module, classify and statistically analyze the load according to load priority and type (resistive, inductive, capacitive) to form a "load status matrix", which includes the real-time power, priority, adjustability, and health status (such as the failure rate of aging equipment) of each load, providing a basis for subsequent allocation; Load forecasting: Based on historical data and real-time information, future load changes are predicted. Short-term forecasting uses the LSTM (Long Short-Term Memory) model, while medium-term forecasting combines the ARIMA model with a rule base. Power allocation: A two-tier model based on priority and capacity constraints—the first tier guarantees a fixed quota for primary loads, and the second tier allocates the remaining capacity to secondary and tertiary loads according to the predicted demand multiplied by the adjustable coefficient; Load balancing: By using intelligent interconnection switches, part of the load on overloaded lines is transferred to lightly loaded lines, and power is distributed between different areas and equipment by phase sequence switching switches to avoid overload of a single link; Optimized scheduling: With the goals of "minimizing line loss", "minimizing electricity purchase cost" and "maximizing distributed energy consumption rate", the optimal solution is obtained by using particle swarm optimization (PSO) to minimize energy consumption and improve system efficiency. Electricity price peak-valley difference period: During off-peak hours (0:00-6:00), the energy storage module is instructed to charge at full power (using low-priced electricity), and during peak hours (8:00-22:00), the energy storage power is released first to reduce the cost of purchasing electricity; During periods of peak solar power generation: When solar power output exceeds local load, excess electricity is prioritized for storage in energy storage to increase self-consumption rate (e.g., from 60% to 90%) and reduce curtailment losses.

[0024] Specifically, the protection module of this invention includes: Fault feature database establishment: Preset characteristic parameters of typical faults (such as current increase of more than 10 times during short circuit, voltage exceeding 120% of rated value during overvoltage, and zero-sequence current greater than 30mA during leakage) to form a fault judgment threshold database; Fault detection: Real-time monitoring of parameters such as current, voltage, and frequency in the power system, and real-time comparison with the fault feature database to detect whether overload, short circuit, overvoltage, or other problems have occurred; Fault location and isolation: Through cross-verification of multi-sensor data, the module or line where the fault occurs is accurately located, and the circuit breaker is immediately triggered to trip and the fuse is blown to cut off the fault circuit. At the same time, the fault area is isolated to prevent the fault from spreading. Fault Recovery Record: After troubleshooting, determine whether the system is ready for recovery through manual confirmation or automatic detection. Restore power supply remotely or locally, and record the fault type, occurrence time, handling process, and equipment status. Store the records in the fault database to provide a basis for subsequent operation and maintenance.

[0025] Specifically, the energy storage module of this invention includes: Energy storage status monitoring: Real-time collection of key parameters of energy storage equipment, such as remaining capacity, charging and discharging current, individual battery voltage, temperature, and cycle count, to determine the health status of the equipment and store excess power in batteries or other energy storage devices when power demand is low or power supply is excessive. Energy release: When power demand is high, electricity is released from the energy storage device to supplement the system power supply; when power demand is low, the charging mode can be activated. Energy storage monitoring: During charging, the charging current is controlled by the charger (such as constant current-constant voltage mode) to avoid large current surges; if an abnormal voltage of a single battery cell is detected, the automatic balancing circuit is activated to balance the state of each battery. During discharge, the inverter converts the DC power into AC power synchronized with the power grid, controls the discharge power, and maintains stable output voltage and frequency.

[0026] Specifically, the remote control and communication module of this invention includes: Communication network setup and protocol adaptation: Terminal devices (sensors, switches) and edge nodes (smart distribution boxes) use short-range communication (LoRa, ZigBee), edge nodes and cloud platforms use long-range communication (5G, fiber optic Ethernet), sensor data uses the MQTT protocol, and control commands use Modbus or DL / T645. Remote monitoring and transmission: Edge nodes package and encrypt real-time data from monitoring modules and equipment status (such as switch on / off position, energy storage SOC) (e.g., AES encryption), and upload them to the remote monitoring platform via the communication network to ensure data security; Remote diagnostics: The platform collects data and logs to perform fault analysis and diagnosis, automatically diagnoses potential equipment faults, and pushes early warning information to maintenance personnel; Remote control: Maintenance personnel initiate control commands through the monitoring platform. The commands are encrypted and sent to the control unit of the corresponding device. After receiving the command, the device first verifies the legality of the command, then executes the operation, and feeds back the execution result to the platform, forming a closed loop.

[0027] Specifically, the artificial intelligence and adaptive scheduling module of this invention includes: Big data feature learning and data acquisition: Collect and process real-time data from the system, construct datasets, and learn the correlation patterns between data through deep learning models; Model building: Real-time data from the monitoring and prediction modules are accessed, and the current operating scenario is identified and its characteristics are determined through an AI model; Adaptive scheduling strategy: Based on the current scenario and learned patterns, the AI ​​model automatically generates an optimized scheduling scheme; Adaptive scheduling optimization: The generated scheduling instructions are sent to the scheduling module and energy storage module for execution, while the execution effect is monitored in real time. If the actual result deviates from the expected result by more than a threshold, the model adjusts the strategy parameters through a reinforcement learning feedback mechanism to achieve self-optimization.

[0028] Specifically, the power quality monitoring module of this invention includes: Power quality data acquisition: Real-time acquisition of data such as voltage fluctuations, frequency changes, and harmonics in the power grid, and recording of events and their durations; Quality assessment: Based on the collected data, it is compared with national standards to determine whether the standards are exceeded; Anomaly detection: When power quality is abnormal, such as excessive voltage fluctuations or harmonic exceedances, an alarm will be triggered immediately and corresponding measures will be taken. Quality Reporting and Traceability: Generates power quality reports for operators to analyze and optimize, and pushes governance suggestions to the scheduling module, and tracks the governance effects.

[0029] Specifically, the intelligent load management module of this invention includes: Load identification: Real-time monitoring of the power demand of each load, identifying peak and off-peak periods of load usage; Load operation monitoring: Combines data from monitoring modules to track the operating status of each load in real time and identify inefficiently operating loads; Load optimization: Based on the principle of "efficiency first + demand matching", an optimization plan is formulated. For adjustable loads (such as air conditioning temperature): automatically adjusts according to ambient temperature; Interruptible loads (such as charging stations): can postpone operation during periods of power shortage and utilize electricity during off-peak hours; Inefficient load: Push replacement suggestions.

[0030] Energy-saving control: Optimization strategies are executed through intelligent controllers (such as air conditioner thermostats and charging pile management systems) to remotely adjust load operating parameters, evaluate the effectiveness of the strategies, and continuously iterate and optimize the solution. A power supply system scheduling method for implementing a power supply system according to any one of claims 1-9, comprising the following steps: S1: The monitoring module collects power supply parameters in real time through sensors deployed at the generator end, transmission lines, load nodes, etc. After data preprocessing, it is uploaded to the central monitoring platform through the remote control and communication module, and simultaneously synchronized to the scheduling module, data prediction module, and artificial intelligence and adaptive scheduling module.

[0031] S2: Based on real-time power data, analyze and classify the loads of each terminal, and generate a load status matrix according to the load's priority, type, adjustability and other characteristics to provide a basis for subsequent power allocation.

[0032] S3: Based on historical data and real-time monitoring data, the LSTM model is used for short-term load forecasting, and the ARIMA model is combined with the rule base for medium-term forecasting. Power is allocated according to the forecast results and priorities. The first-level load is guaranteed its fixed quota, and the remaining capacity is allocated to the second and third-level loads according to the forecast demand.

[0033] S4: Through intelligent scheduling, some of the load on overloaded lines is transferred to lightly loaded lines to achieve load balancing; By employing optimization algorithms such as particle swarm optimization, line losses can be reduced, power procurement costs can be optimized, and the overall system efficiency can be improved. By combining factors such as peak and off-peak electricity prices and photovoltaic power output, the optimal scheduling of energy storage and power distribution can be achieved.

[0034] S5: The data prediction module combines historical data, real-time monitoring data, and related factors, and uses an LSTM model to make short-term predictions, outputting minute-level load fluctuation ranges. The ARIMA model and rule base are used for mid-term prediction, and the load peak values ​​are output in different time periods. The prediction results are synchronized to the scheduling module and the artificial intelligence and adaptive scheduling module.

[0035] S6: The artificial intelligence and adaptive scheduling module integrates monitoring data, prediction results, and historical optimization cases, and identifies the current scenario through a deep learning model. During periods of high solar power generation: Prioritize storing excess electricity in energy storage systems to increase self-consumption rate to over 90%. During peak electricity price periods: Maximize energy storage discharge to reduce electricity purchases from the grid; When the load fluctuates: it is recommended to dynamically adjust the range of adjustable loads (such as air conditioners).

[0036] S7: The scheduling module performs allocation based on a two-layer model of priority and capacity constraints. First layer: Guarantee the fixed quota for primary load; Second tier: Allocate the second and third tier loads within the remaining capacity according to the predicted demand multiplied by the adjustable factor.

[0037] S8: The protection module compares the monitoring data with the fault feature database in real time. Once an overload, short circuit or other abnormality is detected, the circuit breaker is immediately triggered to trip, the fault area is isolated, and a fault signal is sent through the remote control module. After troubleshooting, the protection module confirmed that the insulation was qualified, the dispatch module gradually restored the power supply to the load, and the protection module recorded the fault details to the database for subsequent operation and maintenance optimization.

[0038] S9: The scheduling module evaluates the scheduling effect every hour: calculates indicators such as line loss rate (target ≤ 5%), energy storage utilization rate (target ≥ 70%), and load balance (target ≥ 85%), and compares them with the optimization targets; If the deviation exceeds the threshold (e.g., absorption rate < 80%), it is fed back to the artificial intelligence and adaptive scheduling module, and the model parameters are corrected through reinforcement learning; The intelligent load management module iteratively optimizes strategies to continuously improve system efficiency.

[0039] S10: The remote control and communication module transmits real-time scheduling data, load status, and fault information to the operation and maintenance terminal via 5G / fiber optics. The data is then visualized in the form of current curves, load heat maps, etc., enabling the formulation of optimized scheduling strategies, the execution of energy-saving control, and the optimization of load utilization efficiency.

[0040] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, it is intended that all variations falling within the meaning and scope of equivalents of the claims be included within the present invention.

[0041] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A power supply system, characterized in that, include: It includes a monitoring module, which is responsible for tracking all aspects of the power supply system in real time, including parameters such as current, voltage, frequency, load, and temperature. Through data acquisition equipment, the monitoring module can perform comprehensive real-time monitoring of the system status. The scheduling module is responsible for controlling and optimizing power distribution, ensuring that power is rationally distributed among various loads to avoid overload or insufficient power supply. The protection module is used to detect abnormal conditions such as overload, short circuit, and overvoltage in the power system and automatically disconnect or isolate the faulty part to avoid damage to equipment or cause more serious accidents. The data prediction module uses algorithms to predict future electricity demand and load based on historical and real-time monitoring data. Energy storage modules are used to store excess electricity, especially when electricity demand is low or there is a surplus of supply, to ensure the stability of the power supply. The remote control and communication module is used for remote monitoring, diagnosis and control of the power supply system via wireless or wired networks; The Artificial Intelligence and Adaptive Scheduling Module is used to automatically optimize power dispatch through artificial intelligence and big data analysis, improve the system's adaptability, and respond promptly to load fluctuations and changes in the external environment. The power quality monitoring module is responsible for monitoring power quality issues in the power grid, such as voltage fluctuations, harmonics, frequency deviations, flicker, etc. The intelligent load management module uses data analysis and intelligent scheduling to balance and optimize the use of various loads, ensuring the efficient operation of the system.

2. The power supply system according to claim 1, characterized in that, The monitoring module includes: Data Acquisition: Based on the system architecture, determine the key parameters that need to be monitored, and deploy corresponding sensors at key locations such as the power generation end, transmission lines, distribution nodes, terminal loads, and energy storage devices. Collect data such as current, voltage, frequency, load, and temperature through smart meters, sensors, and other devices. Data transmission: The collected data is transmitted to the central monitoring system or cloud platform in real time; Data processing: Real-time processing and analysis of data to detect anomalies, such as excessively high or low voltage, current fluctuations, etc., and trigger immediate warnings; Data storage and recording: All monitoring data will be saved regularly for subsequent analysis and retrospection; Real-time status visualization: The cloud platform or monitoring terminal transforms the data into intuitive charts, displaying the real-time operating status of each part of the system for operation and maintenance personnel to monitor.

3. A power supply system according to claim 1, characterized in that, The scheduling module includes: Load analysis: Receive real-time power data of each terminal load uploaded by the monitoring module, classify and statistically analyze the load according to load priority and type, and form a load status matrix, including the real-time power, priority, adjustability, health status, etc. of each load, to provide a basis for subsequent allocation; Load forecasting: Predicts future load changes based on historical data and real-time information. Short-term forecasting uses an LSTM model, while medium-term forecasting combines an ARIMA model with a rule base. Power allocation: A two-tier model based on priority and capacity constraints. The first tier guarantees a fixed quota for primary loads, while the second tier allocates the remaining capacity to secondary and tertiary loads according to the predicted demand multiplied by the adjustable coefficient. Load balancing: By using intelligent interconnection switches, part of the load on overloaded lines is transferred to lightly loaded lines, and power is distributed between different areas and equipment by phase sequence switching switches to avoid overload of a single link; Optimized scheduling: Solving for the optimal solution using the particle swarm optimization algorithm to minimize energy consumption and improve system efficiency; Electricity price peak-valley difference period: During off-peak hours, the energy storage module is instructed to charge at full power, and during peak hours, the stored power is released first to reduce the cost of electricity purchase; During periods of peak solar power generation: When solar power output exceeds local load, excess electricity is prioritized for storage to increase self-consumption rate and reduce curtailment losses.

4. A power supply system according to claim 1, characterized in that, The protection module includes: Fault feature library establishment: Preset characteristic parameters of typical faults to form a fault judgment threshold library; Fault detection: Real-time monitoring of parameters such as current, voltage, and frequency in the power system, and real-time comparison with the fault feature database to detect whether overload, short circuit, overvoltage, or other problems have occurred; Fault location and isolation: Through cross-verification of multi-sensor data, the module or line where the fault occurs is accurately located, and the circuit breaker is immediately triggered to trip and the fuse is blown to cut off the fault circuit. At the same time, the fault area is isolated to prevent the fault from spreading. Fault Recovery Record: After troubleshooting, determine whether the system is ready for recovery through manual confirmation or automatic detection. Restore power supply remotely or locally, and record the fault type, occurrence time, handling process, and equipment status. Store the records in the fault database to provide a basis for subsequent operation and maintenance.

5. A power supply system according to claim 1, characterized in that, The energy storage module includes: Energy storage status monitoring: Real-time collection of key parameters of energy storage equipment, such as remaining capacity, charging and discharging current, individual battery voltage, temperature, and cycle count, to determine the health status of the equipment and store excess power in batteries or other energy storage devices when power demand is low or power supply is excessive. Energy release: When power demand is high, electricity is released from the energy storage device to supplement the system power supply; when power demand is low, the charging mode can be activated. Energy storage monitoring: During charging, the charging current is controlled by the charger to avoid large current surges; if an abnormal voltage of a single battery cell is detected, the automatic balancing circuit is activated to balance the state of each battery. During discharge, the inverter converts the DC power into AC power synchronized with the power grid, controls the discharge power, and maintains stable output voltage and frequency.

6. A power supply system according to claim 1, characterized in that, The remote control and communication module includes: Communication network setup and protocol adaptation: terminal devices and edge nodes use short-distance communication, edge nodes and cloud platforms use long-distance communication, sensor data uses the MQTT protocol, and control commands use Modbus or DL / T645. Remote monitoring and transmission: The edge node packages and encrypts the real-time data and device status of the monitoring module, and uploads them to the remote monitoring platform through the communication network to ensure data security; Remote diagnostics: The platform collects data and logs to perform fault analysis and diagnosis, automatically diagnoses potential equipment faults, and pushes early warning information to maintenance personnel; Remote control: Maintenance personnel initiate control commands through the monitoring platform. The commands are encrypted and sent to the control unit of the corresponding device. After receiving the command, the device first verifies the legality of the command, then executes the operation, and feeds back the execution result to the platform, forming a closed loop.

7. A power supply system according to claim 1, characterized in that, The artificial intelligence and adaptive scheduling module includes: Big data feature learning and data acquisition: Collect and process real-time data from the system, construct datasets, and learn the correlation patterns between data through deep learning models; Model building: Real-time data from the monitoring and prediction modules are accessed, and the current operating scenario is identified and its characteristics are determined through an AI model; Adaptive scheduling strategy: Based on the current scenario and learned patterns, the AI ​​model automatically generates an optimized scheduling scheme; Adaptive scheduling optimization: The generated scheduling instructions are sent to the scheduling module and energy storage module for execution, while the execution effect is monitored in real time. If the actual result deviates from the expected result by more than a threshold, the model adjusts the strategy parameters through a reinforcement learning feedback mechanism to achieve self-optimization.

8. A power supply system according to claim 1, characterized in that, The power quality monitoring module includes: Power quality data acquisition: Real-time acquisition of data such as voltage fluctuations, frequency changes, and harmonics in the power grid, and recording of events and their durations; Quality assessment: Based on the collected data, it is compared with national standards to determine whether the standards are exceeded; Anomaly detection: When power quality is abnormal, such as excessive voltage fluctuations or harmonic exceedances, an alarm will be triggered immediately and corresponding measures will be taken. Quality Reporting and Traceability: Generates power quality reports for operators to analyze and optimize, and pushes governance suggestions to the scheduling module, and tracks the governance effects.

9. A power supply system according to claim 1, characterized in that, The intelligent load management module includes: Load identification: Real-time monitoring of the power demand of each load, identifying peak and off-peak periods of load usage; Load operation monitoring: Combines data from monitoring modules to track the operating status of each load in real time and identify inefficiently operating loads; Load optimization: Based on the principles of prioritizing efficiency and matching demand, formulate optimization plans; For adjustable loads: Automatically adjusts according to ambient temperature; Interruptible loads: postpone operation during periods of power shortage to stagger power consumption; Inefficient workload: Push replacement suggestions; Energy-saving control: The system executes optimization strategies through intelligent controllers, remotely adjusts load operating parameters, evaluates the effectiveness of the strategies, and continuously iterates to optimize the solution.

10. A power supply system scheduling method for implementing a power supply system according to any one of claims 1-9, characterized in that, Includes the following steps: S1: The monitoring module collects power supply parameters in real time through sensors deployed at the power generation end, transmission lines, load nodes, etc. After data preprocessing, it is uploaded to the central monitoring platform through the remote control and communication module, and simultaneously synchronized to the scheduling module, data prediction module, and artificial intelligence and adaptive scheduling module. S2: Based on real-time power data, analyze and classify the loads of each terminal, and generate a load status matrix according to the load's priority, type, adjustability and other characteristics to provide a basis for subsequent power allocation; S3: Based on historical data and real-time monitoring data, the LSTM model is used for short-term load forecasting, and the ARIMA model is combined with the rule base for medium-term forecasting. Power is allocated according to the forecast results and priorities. The first-level load is guaranteed its fixed quota, and the remaining capacity is allocated to the second and third-level loads according to the forecast demand. S4: Through intelligent scheduling, some of the load on overloaded lines is transferred to lightly loaded lines to achieve load balancing; By employing optimization algorithms such as particle swarm optimization, line losses can be reduced, power procurement costs can be optimized, and the overall system efficiency can be improved. By combining factors such as peak and off-peak electricity prices and photovoltaic power output, the optimal scheduling of energy storage and power distribution can be achieved; S5: The data prediction module combines historical data, real-time monitoring data, and related factors, and uses an LSTM model to make short-term predictions, outputting minute-level load fluctuation ranges. The ARIMA model and rule base are used for mid-term prediction, and the load peak value is output in different time periods. The prediction results are synchronized to the scheduling module and the artificial intelligence and adaptive scheduling module. S6: The artificial intelligence and adaptive scheduling module accesses monitoring data, prediction results, and historical optimization cases, and identifies the current scenario through a deep learning model; During periods of high solar power generation: Prioritize storing excess electricity in energy storage systems to increase self-consumption rate to over 90%. During peak electricity price periods: Maximize energy storage discharge to reduce electricity purchases from the grid; When the load fluctuates: it is recommended to dynamically adjust the load range. S7: The scheduling module performs allocation based on a two-layer model of priority and capacity constraints. First layer: Guarantee the fixed quota for primary load; Second tier: Allocate the second and third tier loads within the remaining capacity according to the predicted demand multiplied by the adjustable factor; S8: The protection module compares the monitoring data with the fault feature database in real time. Once an overload, short circuit or other abnormality is detected, the circuit breaker is immediately triggered to trip, the fault area is isolated, and a fault signal is sent through the remote control module. After troubleshooting, the protection module confirms that the insulation is qualified, the dispatch module gradually restores the power supply to the load, and the protection module records the fault details to the database for subsequent operation and maintenance optimization. S9: The scheduling module evaluates the scheduling effect every hour: calculates indicators such as line loss rate (target ≤ 5%), energy storage utilization rate (target ≥ 70%), and load balance (target ≥ 85%), and compares them with the optimization targets; If the deviation exceeds the threshold (e.g., absorption rate < 80%), it is fed back to the artificial intelligence and adaptive scheduling module, and the model parameters are corrected through reinforcement learning; The intelligent load management module iteratively optimizes strategies to continuously improve system efficiency. S10: The remote control and communication module transmits real-time scheduling data, load status, and fault information to the operation and maintenance terminal via 5G / fiber optics. The data is then visualized in the form of current curves, load heat maps, etc., enabling the formulation of optimized scheduling strategies, the execution of energy-saving control, and the optimization of load utilization efficiency.