Coordination control method for smart power grid

Abstract

The invention relates to the technical field of power grid coordination control, and provides a coordination control method for a smart power grid, which comprises the following steps: S1, collecting voltage, current, frequency and load data in real time by deploying sensors and monitoring equipment, S2, processing the data by using a big data analysis technology, extracting load trend and fault early warning information, and carrying out early warning on the load trend and the fault early warning information. Real-time data collection is realized by deploying sensors and monitoring equipment, data analysis and load prediction are performed by using big data and a machine learning technology, an optimized objective function is constructed, fine coordination and a global optimal solution are realized by using a multi-agent system and a distributed optimization algorithm, and meanwhile, energy privacy is protected. The reliability of the power grid is improved through intelligent fault detection and a quick response mechanism; according to the technical scheme, interaction between the power grid and users is enhanced, demand side management is achieved, an efficient safety monitoring system is established, multi-level protection measures are taken, safe and stable operation of the power grid is ensured, and the problem that optimization and coordination are not comprehensive in the prior art is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of coordinated control of power grids, and in particular to a coordinated control method for a smart grid. Background Art

[0002] At present, the coordinated control of the power grid is a comprehensive process, which involves the coordinated regulation of multiple links and equipment in the power grid to ensure the safe, stable and efficient operation of the power grid.

[0003] However, existing technologies often have the following shortcomings: they are not comprehensive and timely enough in real-time data collection, and lack efficient deployment of sensors and monitoring equipment; in data analysis and load forecasting, they fail to fully utilize the advantages of big data and machine learning technologies, resulting in limited prediction accuracy and optimization effects; in coordinated control, there is a lack of application of multi-agent systems and distributed optimization algorithms, making it difficult to achieve fine coordination and global optimal solutions, and energy privacy protection may be ignored; fault detection and response mechanisms may not be smart and rapid enough, affecting the reliability of the power grid; there is insufficient interaction with users, and demand-side management methods are limited; at the same time, the safety monitoring system may not be perfect, lacking multi-level protection measures, making it difficult to ensure the safe and stable operation of the power grid Summary of the invention

[0004] The present invention proposes a coordinated control method for a smart grid, which solves the problem of incomplete optimization and coordination in related technologies.

[0005] The technical solution of the present invention is as follows: A coordinated control method of a smart grid, comprising: S1, collecting voltage, current, frequency, and load data in real time by deploying sensors and monitoring equipment;

[0006] S2. Use big data analysis technology to process these data and extract load trends and fault warning information;

[0007] S3. Use big data analysis technology to predict the power grid load and build an optimization objective function based on the prediction results;

[0008] S4. Treat wind power, photovoltaic power, and energy storage as independent intelligent entities, and achieve fine coordination through a multi-agent system. The intelligent entities perform autonomous optimization based on local information and respond to global instructions to ensure consistent goals. A distributed optimization algorithm is introduced to protect energy privacy while finding the global optimal solution.

[0009] S5. Use machine learning algorithms to make short-term and long-term forecasts of power grid loads, generate load forecast curves, and formulate load response strategies based on load forecast results and real-time power grid status;

[0010] S6. Through real-time monitoring of the grid operation status, using intelligent fault detection to quickly locate the fault point and establish a rapid response mechanism;

[0011] S7. Strengthen the interaction between the power grid and users, collect user electricity consumption information through smart meters, smart homes and other devices, and realize demand-side management;

[0012] S8. Establish an efficient security monitoring system to monitor potential safety hazards in the power grid in real time, and adopt multi-level security protection measures to protect key equipment and systems from attacks.

[0013] As a preferred solution of the present invention, the specific steps of deploying sensors and monitoring equipment to collect voltage, current, frequency and load data in real time in step S1 are as follows:

[0014] S11. Clarify the specific requirements for data collection, including the grid area to be monitored, monitoring parameters, data accuracy requirements, and real-time requirements;

[0015] S12. Based on the demand analysis results, plan the deployment plan of sensors and monitoring equipment, including the number, type, installation location and data transmission method of the equipment;

[0016] S13, start the sensors and monitoring equipment to collect real-time grid operation data such as voltage, current, frequency and load. The equipment converts the collected data into digital signals and performs preliminary processing;

[0017] S14, transmitting the collected data to the data processing center and the cloud platform via wired or wireless means.

[0018] As a preferred solution of the present invention, the specific steps of processing these data using big data analysis technology in step S2 to extract load trend and fault warning information are as follows:

[0019] S21. Remove outliers and noise data to ensure data accuracy and completeness, integrate data from different sources and formats into a unified format, and perform appropriate conversion and normalization on the data according to analysis requirements;

[0020] S22, extracting features reflecting the load variation pattern (daily load peak value, valley value, average load, load rate and load rate variation rate) from the load data,

[0021] S23. Use time series analysis methods (ARIMA model is used, which can simultaneously consider the autoregressive, sliding average and differential stationarity of data, and is suitable for situations where load changes are complex, seasonal fluctuations and random fluctuations exist) to conduct in-depth analysis of load data and extract load trend information;

[0022] S24. Present the results of load trend analysis to relevant personnel in the form of charts, reports, etc. When the fault warning (the warning model adopts the exponential smoothing model, which can find out the random factors that cause its instability based on long-term recorded data, and find out the most appropriate probability model based on the probability statistics method, so as to estimate the main change trend of the risk value and express it in the form of a smoothed value. When the smoothed value is greater than a certain threshold, the system will issue a warning signal and will also perform self-correction based on future recorded data) model issues a warning signal, promptly notify relevant personnel to take corresponding measures to deal with it.

[0023] As a preferred solution of the present invention, the step S3 uses big data analysis technology to predict the power grid load, and based on the prediction results, the specific steps of constructing the optimization objective function are as follows:

[0024] S31. Collect historical load data, weather data, holiday information, and economic indicator-related factor data, and clean, integrate, convert, and normalize the data to ensure data quality;

[0025] S32, extracting features reflecting the load variation law from the preprocessed data, and constructing an optimization objective function for the features that conform to the variation law to achieve optimal dispatching and planning of the power grid;

[0026] The optimization objective function is: Let Pt be the predicted load in time period t, Gt be the generator output in time period t, Ct(Gt) be the power generation cost function in time period t, Dt be the demand response amount in time period t, and Rt(Dt) be the demand response cost function in time period t. The optimization objective function can be expressed as: \text{Minimize}\sum_{t=1}^{T}\left[C_t(G_t)+R_t(D_t)\right]\]####Constraints 1. **Load balance constraint**: [G_t+D_t=P_t,\quad\forallt\] Ensure that the sum of the generator output and the demand response amount in each time period is equal to the predicted load, 2. **Generator output constraint**: [G_{\text{min}}\leq G_t\leq G_{\text{max}},\quad\forall t\] Ensure that the generator output is within the allowable range, 3. **Demand response constraint**: [D_{\text{min}}\leq D_t\leq D_{\text{max}},\quad\forall t\] Ensure that the demand response volume is within the allowable range, 4. **Other constraints** (such as ramp rate constraints, spare capacity constraints, etc.): [text{Other related constraints}\]###3. Function representation (pseudocode) The following is the load forecasting and optimization objective function construction process represented by Python pseudocode: ```python#Load forecasting function defload_forecast(historical_data,features,model):#Data preprocessing preprocessed_data=preprocess(historical_data,features)#Model training model.fit(preprocessed_data)#Load forecasting forecasted_load=model.predict(future_features)returnforecasted_load#Optimization objective function def optimize_objective(forecasted_load,generator_costs,demand_response_costs):T=len(forecasted_load)total_cost=0for t in range(T):#Initialize generator output and demand response amount G_t=initialize_generator_output(t)D_t=initialize_demand_response(t)#Load balancing constraintwhile G_t+D_t!=forecasted_load[t]:#Adjust generator output and demand response amount to meet load balanceadjust_output(G_t,D_t,forecasted_load[t])#Calculate costcost_t=generator_costs[t](G_t)+demand_response_costs[t](D_t)total_cost+=cost_t return total_cost#Main function defmain():#Load historical data and features historical_data=load_historical_data() features=extract_features(historical_data)#Select and initialize the model model=initialize_model()#Perform load forecasting forecasted_load=load_forecast(historical_data,features,model)#Define generator cost and demand response cost functions generator_costs=define_generator_costs() demand_response_costs=define_demand_response_costs()#Optimize the objective function and output the result optimal_cost=optimize_objective(forecasted_load,generator_costs,demand_response_costs) print(f"Optimal cost:{optimal_cost}")#Execute the main function if__name__=="__main__":main()```。 .

[0027] As a preferred solution of the present invention, in step S4, wind power, photovoltaic power and energy storage are regarded as independent intelligent agents, and fine coordination is achieved through a multi-agent system. The intelligent agents perform autonomous optimization based on local information and respond to global instructions to ensure consistent goals. A distributed optimization algorithm is introduced to protect energy privacy while finding the global optimal solution. The specific steps are as follows:

[0028] S41: Define three agents: wind power (W), photovoltaic (PV) and energy storage (ES). Each agent has its own state variables, objective functions and constraints. Initialize the state variables and optimization parameters of each agent (the initialization algorithm formula is: for each agent i (i∈{W, PV, ES}), initialize the state variable xi and optimization parameters (such as learning rate α), and the global coordinator initializes the global variable z and the number of iterations k);

[0029] S42, each agent performs autonomous optimization based on local information and its own objective function, and uses an optimization algorithm to solve the local optimization problem of each agent (the local optimization algorithm formula is: calculate the local gradient Where fi is the objective function of agent i, and updates the local state variables:

[0030] S43. Design a global coordinator (GC) to issue global instructions and coordinate the behaviors among agents. The global coordinator generates instructions based on global goals and status information (such as total energy demand and grid stability) and sends them to each agent.

[0031] S44. Establish communication mechanisms between agents, allowing them to exchange necessary information, and design collaborative strategies so that agents can respond to global instructions and adjust their own optimization strategies to maintain consistency;

[0032] S45. Select an appropriate distributed optimization algorithm to decompose the global optimization problem into multiple sub-problems. Each sub-problem is solved by a corresponding intelligent agent, and the global optimal solution is gradually approached through an iterative process (global variable: zk+1=N1∑i=1Nxik+1, where N is the number of intelligent agents);

[0033] S46. Set the number of iterations or convergence conditions, and repeat until the preset number of iterations and convergence conditions are reached. When the algorithm converges, output the global optimal solution (iteration formula: |||zk+1-zk|||<∈, optimal solution output formula: x*=zK).

[0034] As a preferred solution of the present invention, in step S5, a machine learning algorithm is used to perform short-term and long-term forecasts on the power grid load, a load forecast curve is generated, and a load response strategy is formulated according to the load forecast results and the real-time state of the power grid. The specific steps are as follows:

[0035] S51. Real-time monitoring of the operation status of the power grid, including current load, power generation capacity, and energy storage status;

[0036] S52. Adjust the electricity price strategy based on the load forecast results and real-time status to guide users to reduce electricity load during peak hours.

[0037] S53, store excess power during low load periods and release the stored energy during peak hours to balance the grid load;

[0038] S54. Implement the formulated load response strategy and monitor the implementation effect;

[0039] Load forecasting algorithm model: Let X be the feature matrix, y be the target load vector, and use SVR for regression: minimize 21∥w∥2+C∑i=1nξi+ξi*y_i-(w\cdot\phi(x_i))-b\leq\epsilon+\xi_i\\(w\cdot

[0040] \phi(x_i))+b-y_i\leq\epsilon+\xi_i^*\\\xi_i,\xi_i^*\geq 0\end{cases}$$where $w$ is the weight vector, $b$ is the bias term, $\phi(x_i)$ is the function that maps the input $x_i$ to a high-dimensional space, $\epsilon$ is the allowed bias, $C$ is the regularization parameter, and $\xi_i,\xi_i^*$ are slack variables. ####3. Load response strategy - electricity price adjustment Assume $p_t$ is the electricity price in the $t$th period, and adjust the electricity price according to the load forecast results and real-time status: _t=p_0+\alpha\left(\frac{L_t{\text{avg}}}{\sigma_L}\right)$$Wherein, p0 is the basic electricity price, α is the adjustment coefficient, Ltpred is the predicted load value, Ltavg is the historical average load value, and σL is the standard deviation of the load.

[0041] As a preferred solution of the present invention, in step S6, by real-time monitoring of the grid operation status, using intelligent fault detection to quickly locate the fault point, and establishing a rapid response mechanism, the specific steps are as follows:

[0042] S61. Use multiple fault detection principles to determine whether there is a fault by comparing the difference in electrical quantities on both sides of the line or analyzing the propagation characteristics of the traveling wave signal. Use machine learning and deep learning artificial intelligence technology to learn and analyze the historical fault data and real-time data of the power grid;

[0043] S62, after the fault is detected, the fault location method is started to determine the specific location of the fault point, and the fault point is quickly located through analysis in combination with the topological structure of the power grid and real-time data;

[0044] S63. When a fault is detected and located, an alarm signal is immediately sent to the control center, and the fault information is transmitted to the dispatcher and maintenance team through the communication network;

[0045] S64. After confirming the fault point, quickly isolate the fault area to prevent the fault from spreading and affecting other non-fault areas. Start the backup equipment and lines to restore power supply to the affected area to reduce power outage time and losses. At the same time, formulate a fault handling plan based on the type and severity of the fault and organize a maintenance team to carry out emergency repairs;

[0046] S65. After the fault handling is completed, conduct an in-depth analysis of the fault data to find out the cause of the fault and potential risks. According to the analysis results, formulate corresponding preventive measures and improvement plans, summarize and evaluate the entire fault handling process, and continuously improve and optimize the rapid response mechanism.

[0047] As a preferred solution of the present invention, the specific steps of strengthening the interaction between the power grid and users in step S7, collecting user electricity consumption information through smart meters, smart homes and other devices, and realizing demand side management are as follows:

[0048] S71, using smart meters and smart home devices to collect users' electricity consumption data in real time, including power consumption, voltage, current and power factor, and transmitting the collected data to the power grid data center through power line carrier communication and wireless communication;

[0049] S72. Use big data analysis and artificial intelligence technology to conduct in-depth mining and analysis of user electricity consumption data. Identify users' electricity consumption patterns, load characteristics, and energy-saving potential, and formulate demand-side management strategies based on the analysis results;

[0050] S73. Establish an interactive platform between the power grid and users, release electricity price information, load status and energy-saving suggestions to users in real time through SMS and APP push, and encourage users to participate in demand response plans to balance the power grid load and improve energy efficiency by adjusting electricity consumption behavior;

[0051] S74. Regularly evaluate and provide feedback on the effectiveness of demand-side management, collect user opinions and suggestions, and continuously improve and optimize demand-side management strategies and measures to ensure the continued and effective advancement of demand-side management and continuously enhance user satisfaction and sense of gain.

[0052] As a preferred solution of the present invention, the specific steps of establishing an efficient security monitoring system in step S8, monitoring the potential safety hazards in the power grid in real time, taking multi-level security protection measures, and protecting key equipment and systems from attacks are as follows:

[0053] S81. Ensure that grid equipment and monitoring equipment are placed in a secure environment to prevent physical damage and unauthorized access, and implement strict physical access control measures;

[0054] S82. Deploy firewalls and intrusion detection and prevention systems (IDS / IPS) network security devices between the power grid and the Internet to prevent external network attacks, use encryption technology to protect the security of data transmission, ensure that sensitive data is not intercepted or tampered with, and regularly scan network devices for security vulnerabilities and update patches to repair potential security vulnerabilities.

[0055] S83. Conduct regular security audits and penetration tests on power grid control systems and monitoring platforms to identify and promptly repair security vulnerabilities, adopt access control policies to limit user permissions to ensure that only authorized users can access key equipment and systems, and deploy reliable antivirus software and security update mechanisms to prevent malware intrusion and damage;

[0056] S84. Encrypt sensitive data in the power grid for storage and transmission, ensure data security during storage and transmission, implement data backup and recovery strategies, regularly back up key data and store them in offline or remote secure locations, and establish a comprehensive data access control mechanism to prevent unauthorized access and leakage;

[0057] S85. Develop a detailed emergency response plan, clarify the handling process and division of responsibilities when a security incident occurs, conduct emergency drills and training regularly to enhance the ability of operation and maintenance personnel to respond to emergencies, and establish a rapid recovery mechanism to ensure that the normal operation of the power grid can be quickly restored after an attack or failure.

[0058] The working principle and beneficial effects of the present invention are:

[0059] 1. The present invention realizes real-time data collection by deploying sensors and monitoring equipment, uses big data and machine learning technology for data analysis and load forecasting, and constructs an optimization objective function; adopts a multi-agent system and a distributed optimization algorithm to achieve fine coordination and global optimal solutions while protecting energy privacy; improves grid reliability through intelligent fault detection and rapid response mechanisms; strengthens the interaction between the grid and users to achieve demand-side management; and establishes an efficient safety monitoring system and adopts multi-level protection measures to ensure the safe and stable operation of the grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0061] Figure 1 It is a schematic diagram of the process structure of the present invention. DETAILED DESCRIPTION

[0062] The following will be combined with the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0063] Example:

[0064] In this embodiment, a coordinated control method of a smart grid includes:

[0065] S1. Collect voltage, current, frequency and load data in real time by deploying sensors and monitoring equipment;

[0066] S2. Use big data analysis technology to process these data and extract load trends and fault warning information;

[0067] S3. Use big data analysis technology to predict the power grid load and build an optimization objective function based on the prediction results;

[0068] S4. Treat wind power, photovoltaic power, and energy storage as independent intelligent entities, and achieve fine coordination through a multi-agent system. The intelligent entities perform autonomous optimization based on local information and respond to global instructions to ensure consistent goals. Distributed optimization algorithms are introduced to protect energy privacy while finding the global optimal solution.

[0069] S5. Use machine learning algorithms to make short-term and long-term forecasts of power grid loads, generate load forecast curves, and formulate load response strategies based on load forecast results and real-time power grid status;

[0070] S6. Through real-time monitoring of the grid operation status, using intelligent fault detection to quickly locate the fault point and establish a rapid response mechanism;

[0071] S7. Strengthen the interaction between the power grid and users, collect user electricity consumption information through smart meters, smart homes and other devices, and realize demand-side management;

[0072] S8. Establish an efficient security monitoring system to monitor potential safety hazards in the power grid in real time, and adopt multi-level security protection measures to protect key equipment and systems from attacks.

[0073] Specifically, the specific steps of deploying sensors and monitoring equipment in step S1 to collect voltage, current, frequency, and load data in real time are as follows:

[0074] S11. Clarify the specific requirements for data collection, including the grid area to be monitored, monitoring parameters, data accuracy requirements, and real-time requirements;

[0075] S12. Based on the demand analysis results, plan the deployment plan of sensors and monitoring equipment, including the number, type, installation location and data transmission method of the equipment;

[0076] S13, start the sensors and monitoring equipment to collect real-time grid operation data such as voltage, current, frequency and load. The equipment converts the collected data into digital signals and performs preliminary processing;

[0077] S14, transmitting the collected data to the data processing center and the cloud platform via wired or wireless means.

[0078] In this embodiment, the specific steps of processing these data using big data analysis technology in step S2 to extract load trend and fault warning information are as follows:

[0079] S21. Remove outliers and noise data to ensure data accuracy and completeness, integrate data from different sources and formats into a unified format, and perform appropriate conversion and normalization on the data according to analysis requirements;

[0080] S22, extracting features reflecting the load variation pattern (daily load peak value, valley value, average load, load rate and load rate variation rate) from the load data,

[0081] S23. Use time series analysis methods (ARIMA model is used, which can simultaneously consider the autoregressive, sliding average and differential stationarity of data, and is suitable for situations where load changes are complex, seasonal fluctuations and random fluctuations exist) to conduct in-depth analysis of load data and extract load trend information;

[0082] S24. Present the results of load trend analysis to relevant personnel in the form of charts, reports, etc. When the fault warning (the warning model adopts the exponential smoothing model, which can find out the random factors that cause its instability based on long-term recorded data, and find out the most appropriate probability model based on the probability statistics method, so as to estimate the main change trend of the risk value and express it in the form of a smoothed value. When the smoothed value is greater than a certain threshold, the system will issue a warning signal and will also perform self-correction based on future recorded data) model issues a warning signal, promptly notify relevant personnel to take corresponding measures to deal with it.

[0083] Specifically, the steps of using big data analysis technology to predict the power grid load in step S3 and constructing an optimization objective function based on the prediction results are as follows:

[0084] S31. Collect historical load data, weather data, holiday information, and economic indicator-related factor data, and clean, integrate, convert, and normalize the data to ensure data quality;

[0085] S32, extracting features reflecting the load variation law from the preprocessed data, and constructing an optimization objective function for the features that conform to the variation law to achieve optimal dispatching and planning of the power grid;

[0086] The optimization objective function is: Let Pt be the predicted load in time period t, Gt be the generator output in time period t, Ct(Gt) be the power generation cost function in time period t, Dt be the demand response amount in time period t, and Rt(Dt) be the demand response cost function in time period t. The optimization objective function can be expressed as: \text{Minimize}\sum_{t=1}^{T}\left[C_t(G_t)+R_t(D_t)\right]\]####Constraints 1. **Load balance constraint**: [G_t+D_t=P_t,\quad\forallt\] Ensure that the sum of the generator output and the demand response amount in each time period is equal to the predicted load, 2. **Generator output constraint**: [G_{\text{min}}\leq G_t\leq G_{\text{max}},\quad\forall t\] Ensure that the generator output is within the allowable range, 3. **Demand response constraint**: [D_{\text{min}}\leq D_t\leq D_{\text{max}},\quad\forall t\] Ensure that the demand response volume is within the allowable range, 4. **Other constraints** (such as ramp rate constraints, spare capacity constraints, etc.): [text{Other related constraints}\]###3. Function representation (pseudocode) The following is the load forecasting and optimization objective function construction process represented by Python pseudocode: ```python#Load forecasting function defload_forecast(historical_data,features,model):#Data preprocessing preprocessed_data=preprocess(historical_data,features)#Model training model.fit(preprocessed_data)#Load forecasting forecasted_load=model.predict(future_features)returnforecasted_load#Optimization objective function def optimize_objective(forecasted_load,generator_costs,demand_response_costs):T=len(forecasted_load)total_cost=0for t in range(T):#Initialize generator output and demand response amount G_t=initialize_generator_output(t)D_t=initialize_demand_response(t)#Load balancing constraintwhile G_t+D_t!=forecasted_load[t]:#Adjust generator output and demand response amount to meet load balanceadjust_output(G_t,D_t,forecasted_load[t])#Calculate costcost_t=generator_costs[t](G_t)+demand_response_costs[t](D_t)total_cost+=cost_t return total_cost#Main function defmain():#Load historical data and features historical_data=load_historical_data() features=extract_features(historical_data)#Select and initialize the model model=initialize_model()#Perform load forecasting forecasted_load=load_forecast(historical_data,features,model)#Define generator cost and demand response cost functions generator_costs=define_generator_costs() demand_response_costs=define_demand_response_costs()#Optimize the objective function and output the result optimal_cost=optimize_objective(forecasted_load,generator_costs,demand_response_costs) print(f"Optimal cost:{optimal_cost}")#Execute the main function if__name__=="__main__":main()```。 .

[0087] In this embodiment, in step S4, wind power, photovoltaic power, and energy storage are regarded as independent agents, and fine coordination is achieved through a multi-agent system. The agents perform autonomous optimization based on local information and respond to global instructions to ensure consistent goals. A distributed optimization algorithm is introduced to protect energy privacy while finding the global optimal solution. The specific steps are as follows:

[0088] S41: Define three agents: wind power (W), photovoltaic (PV) and energy storage (ES). Each agent has its own state variables, objective functions and constraints. Initialize the state variables and optimization parameters of each agent (the initialization algorithm formula is: for each agent i (i∈{W, PV, ES}), initialize the state variable xi and optimization parameters (such as learning rate α), and the global coordinator initializes the global variable z and the number of iterations k);

[0089] S42, each agent performs autonomous optimization based on local information and its own objective function, and uses an optimization algorithm to solve the local optimization problem of each agent (the local optimization algorithm formula is: calculate the local gradient Where fi is the objective function of agent i, and updates the local state variables:

[0090] S43. Design a global coordinator (GC) to issue global instructions and coordinate the behaviors among agents. The global coordinator generates instructions based on global goals and status information (such as total energy demand and grid stability) and sends them to each agent.

[0091] S44. Establish communication mechanisms between agents, allowing them to exchange necessary information, and design collaborative strategies so that agents can respond to global instructions and adjust their own optimization strategies to maintain consistency;

[0092] S45. Select an appropriate distributed optimization algorithm to decompose the global optimization problem into multiple sub-problems. Each sub-problem is solved by a corresponding intelligent agent, and the global optimal solution is gradually approached through an iterative process (global variable: zk+1=N1∑i=1Nxik+1, where N is the number of intelligent agents);

[0093] S46. Set the number of iterations or convergence conditions, and repeat until the preset number of iterations and convergence conditions are reached. When the algorithm converges, output the global optimal solution (iteration formula: |||zk+1-zk|||<∈, optimal solution output formula: x*=zK).

[0094] Specifically, in step S5, the machine learning algorithm is used to perform short-term and long-term forecasts on the power grid load, generate a load forecast curve, and formulate a load response strategy according to the load forecast results and the real-time status of the power grid. The specific steps are as follows:

[0095] S51. Real-time monitoring of the operation status of the power grid, including current load, power generation capacity, and energy storage status;

[0096] S52. Adjust the electricity price strategy based on the load forecast results and real-time status to guide users to reduce electricity load during peak hours.

[0097] S53, store excess electricity during low load periods and release the stored energy during peak hours to balance the grid load;

[0098] S54. Implement the formulated load response strategy and monitor the implementation effect;

[0099] Load forecasting algorithm model: Let X be the feature matrix, y be the target load vector, and use SVR for regression: minimize 21∥w∥2+C∑i=1nξi+ξi*y_i-(w\cdot\phi(x_i))-b\leq\epsilon+\xi_i\\(w\cdot

[0100] \phi(x_i))+b-y_i\leq\epsilon+\xi_i^*\\\xi_i,\xi_i^*\geq 0\end{cases}$$where $w$ is the weight vector, $b$ is the bias term, $\phi(x_i)$ is the function that maps the input $x_i$ to a high-dimensional space, $\epsilon$ is the allowed bias, $C$ is the regularization parameter, and $\xi_i,\xi_i^*$ are slack variables. ####3. Load response strategy - electricity price adjustment Assume $p_t$ is the electricity price in the $t$th period, and adjust the electricity price according to the load forecast results and real-time status: _t=p_0+\alpha\left(\frac{L_t{\text{avg}}}{\sigma_L}\right)$$Wherein, p0 is the basic electricity price, α is the adjustment coefficient, Ltpred is the predicted load value, Ltavg is the historical average load value, and σL is the standard deviation of the load.

[0101] In this embodiment, in step S6, the operation status of the power grid is monitored in real time, the fault point is quickly located using intelligent fault detection, and a rapid response mechanism is established. The specific steps are as follows:

[0102] S61. Use multiple fault detection principles to determine whether there is a fault by comparing the difference in electrical quantities on both sides of the line or analyzing the propagation characteristics of the traveling wave signal. Use machine learning and deep learning artificial intelligence technology to learn and analyze the historical fault data and real-time data of the power grid;

[0103] S62, after the fault is detected, the fault location method is started to determine the specific location of the fault point, and the fault point is quickly located through analysis in combination with the topological structure of the power grid and real-time data;

[0104] S63. When a fault is detected and located, an alarm signal is immediately sent to the control center, and the fault information is transmitted to the dispatcher and maintenance team through the communication network;

[0105] S64. After confirming the fault point, quickly isolate the fault area to prevent the fault from spreading and affecting other non-fault areas. Start the backup equipment and lines to restore power supply to the affected area to reduce power outage time and losses. At the same time, formulate a fault handling plan based on the type and severity of the fault and organize a maintenance team to carry out emergency repairs;

[0106] S65. After the fault handling is completed, conduct an in-depth analysis of the fault data to find out the cause of the fault and potential risks. According to the analysis results, formulate corresponding preventive measures and improvement plans, summarize and evaluate the entire fault handling process, and continuously improve and optimize the rapid response mechanism.

[0107] Specifically, the steps of strengthening the interaction between the power grid and users in step S7, collecting user electricity usage information through smart meters, smart homes and other devices, and implementing demand side management are as follows:

[0108] S71, using smart meters and smart home devices to collect users' electricity consumption data in real time, including power consumption, voltage, current and power factor, and transmitting the collected data to the power grid data center through power line carrier communication and wireless communication;

[0109] S72. Use big data analysis and artificial intelligence technology to conduct in-depth mining and analysis of user electricity consumption data. Identify users' electricity consumption patterns, load characteristics, and energy-saving potential, and formulate demand-side management strategies based on the analysis results;

[0110] S73. Establish an interactive platform between the power grid and users, release electricity price information, load status and energy-saving suggestions to users in real time through SMS and APP push, and encourage users to participate in demand response plans to balance the power grid load and improve energy efficiency by adjusting electricity consumption behavior;

[0111] S74. Regularly evaluate and provide feedback on the effectiveness of demand-side management, collect user opinions and suggestions, and continuously improve and optimize demand-side management strategies and measures to ensure the continued and effective advancement of demand-side management and continuously enhance user satisfaction and sense of gain.

[0112] In this embodiment, the specific steps of establishing an efficient security monitoring system in step S8, monitoring the potential safety hazards in the power grid in real time, taking multi-level security protection measures, and protecting key equipment and systems from attacks are as follows:

[0113] S81. Ensure that grid equipment and monitoring equipment are placed in a secure environment to prevent physical damage and unauthorized access, and implement strict physical access control measures;

[0114] S82. Deploy firewalls and intrusion detection and prevention systems (IDS / IPS) network security devices between the power grid and the Internet to prevent external network attacks, use encryption technology to protect the security of data transmission, ensure that sensitive data is not intercepted or tampered with, and regularly scan network devices for security vulnerabilities and update patches to repair potential security vulnerabilities.

[0115] S83. Conduct regular security audits and penetration tests on power grid control systems and monitoring platforms to identify and promptly repair security vulnerabilities, adopt access control policies to limit user permissions to ensure that only authorized users can access key equipment and systems, and deploy reliable antivirus software and security update mechanisms to prevent malware intrusion and damage;

[0116] S84. Encrypt sensitive data in the power grid for storage and transmission, ensure data security during storage and transmission, implement data backup and recovery strategies, regularly back up key data and store them in offline or remote secure locations, and establish a comprehensive data access control mechanism to prevent unauthorized access and leakage;

[0117] S85. Develop a detailed emergency response plan, clarify the handling process and division of responsibilities when a security incident occurs, conduct emergency drills and training regularly to enhance the ability of operation and maintenance personnel to respond to emergencies, and establish a rapid recovery mechanism to ensure that the normal operation of the power grid can be quickly restored after an attack or failure.

[0118] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A coordinated control method for a smart grid, characterized in that: include: S1. Collect voltage, current, frequency and load data in real time by deploying sensors and monitoring equipment; S2. Use big data analysis technology to process these data and extract load trends and fault warning information; S3. Use big data analysis technology to predict the power grid load and build an optimization objective function based on the prediction results; S4. Treat wind power, photovoltaic power, and energy storage as independent intelligent entities, and achieve fine coordination through a multi-agent system. The intelligent entities perform autonomous optimization based on local information and respond to global instructions to ensure consistent goals. A distributed optimization algorithm is introduced to protect energy privacy while finding the global optimal solution. S5. Use machine learning algorithms to make short-term and long-term forecasts of power grid loads, generate load forecast curves, and formulate load response strategies based on load forecast results and real-time power grid status; S6. Through real-time monitoring of the grid operation status, using intelligent fault detection to quickly locate the fault point and establish a rapid response mechanism; S7. Strengthen the interaction between the power grid and users, collect user electricity consumption information through smart meters, smart homes and other devices, and realize demand-side management; S8. Establish an efficient security monitoring system to monitor potential safety hazards in the power grid in real time, and adopt multi-level security protection measures to protect key equipment and systems from attacks.

2. A coordinated control method for a smart grid according to claim 1, characterized in that: The specific steps of collecting voltage, current, frequency and load data in real time by deploying sensors and monitoring equipment in step S1 are as follows: S11. Clarify the specific requirements for data collection, including the grid area to be monitored, monitoring parameters, data accuracy requirements, and real-time requirements; S12. Based on the demand analysis results, plan the deployment plan of sensors and monitoring equipment, including the number, type, installation location and data transmission method of the equipment; S13, start the sensors and monitoring equipment to collect real-time grid operation data such as voltage, current, frequency and load. The equipment converts the collected data into digital signals and performs preliminary processing; S14, transmitting the collected data to the data processing center and the cloud platform via wired or wireless means.

3. A coordinated control method for a smart grid according to claim 2, characterized in that: The specific steps of using big data analysis technology to process these data and extract load trend and fault warning information in step S2 are as follows: S21. Remove outliers and noise data to ensure data accuracy and completeness, integrate data from different sources and formats into a unified format, and perform appropriate conversion and normalization on the data according to analysis requirements; S22, extracting features reflecting the load variation pattern (daily load peak value, valley value, average load, load rate and load rate variation rate) from the load data, S23. Use time series analysis methods (ARIMA model is used, which can simultaneously consider the autoregressive, sliding average and differential stationarity of data, and is suitable for situations where load changes are complex, seasonal fluctuations and random fluctuations exist) to conduct in-depth analysis of load data and extract load trend information; S24. Present the results of load trend analysis to relevant personnel in the form of charts, reports, etc. When the fault warning (the warning model adopts the exponential smoothing model, which can find out the random factors that cause its instability based on long-term recorded data, and find out the most appropriate probability model based on the probability statistics method, so as to estimate the main change trend of the risk value and express it in the form of a smoothed value. When the smoothed value is greater than a certain threshold, the system will issue a warning signal and will also perform self-correction based on future recorded data) model issues a warning signal, promptly notify relevant personnel to take corresponding measures to deal with it.

4. A coordinated control method for a smart grid according to claim 3, characterized in that: In step S3, the power grid load is predicted by using big data analysis technology, and based on the prediction results, the optimization objective function is constructed in the following specific steps: S31. Collect historical load data, weather data, holiday information, and economic indicator-related factor data, and clean, integrate, convert, and normalize the data to ensure data quality; S32, extracting features reflecting the load variation law from the preprocessed data, and constructing an optimization objective function for the features that conform to the variation law to achieve optimal dispatching and planning of the power grid; The optimization objective function is: Let Pt be the predicted load in time period t, Gt be the generator output in time period t, Ct(Gt) be the power generation cost function in time period t, Dt be the demand response amount in time period t, and Rt(Dt) be the demand response cost function in time period t. The optimization objective function can be expressed as: \text{Minimize}\sum_{t=1}^{T}\left[C_t(G_t)+R_t(D_t)\right]\]####Constraints 1. **Load balance constraint**: [G_t+D_t=P_t,\quad\forall t\] Ensure that the sum of the generator output and the demand response amount in each time period is equal to the predicted load, 2. **Generator output constraint**: [G_{\text{min}}\leq G_t\leq G_{\text{max}},\quad\forall t\] Ensure that the generator output is within the allowable range, 3. **Demand response constraint**: [D_{\text{min}}\leq D_t\leq D_{\text{max}},\quad\forallt\] Ensure that the demand response volume is within the allowable range, 4. **Other constraints** (such as ramp rate constraints, spare capacity constraints, etc.): [text{Other related constraints}\]###3. Function representation (pseudocode) The following is the load forecasting and optimization objective function construction process expressed in Python pseudocode: ```python#Load forecasting function def load_forecast(historical_data,features,model):#Data preprocessing preprocessed_data=preprocess(historical_data,features)#Model training model.fit(preprocessed_data)#Load forecasting forecasted_load=model.predict(future_features)returnforecasted_load#Optimization objective function def optimize_objective(forecasted_load,generator_costs,demand_response_costs):T=len(forecasted_load)total_cost=0 for t in range(T):#Initialize generator output and demand response amount G_t=initialize_generator_output(t)D_t=initialize_demand_response(t)#Load balancing constraintwhile G_t+D_t!=forecasted_load[t]:#Adjust generator output and demand response amount to meet load balanceadjust_output(G_t,D_t,forecasted_load[t])#Calculate costcost_t=generator_costs[t](G_t)+demand_response_costs[t](D_t)total_cost+=cost_t return total_cost#Main function defmain():#Load historical data and features historical_data=load_historical_data() features=extract_features(historical_data)#Select and initialize the model model=initialize_model()#Perform load forecasting forecasted_load=load_forecast(historical_data,features,model)#Define generator cost and demand response cost functions generator_costs=define_generator_costs() demand_response_costs=define_demand_response_costs()#Optimize the objective function and output the result optimal_cost=optimize_objective(forecasted_load,generator_costs,demand_response_costs) print(f"Optimal cost:{optimal_cost}")#Execute the main function if__name__=="__main__":main()```。 .

5. A coordinated control method for a smart grid according to claim 4, characterized in that: In step S4, wind power, photovoltaic power, and energy storage are regarded as independent intelligent agents, and fine coordination is achieved through a multi-agent system. The intelligent agents perform autonomous optimization based on local information and respond to global instructions to ensure consistent goals. A distributed optimization algorithm is introduced to protect energy privacy while finding the global optimal solution. The specific steps are as follows: S41: Define three agents: wind power (W), photovoltaic (PV) and energy storage (ES). Each agent has its own state variables, objective functions and constraints. Initialize the state variables and optimization parameters of each agent (the initialization algorithm formula is: for each agent i (i∈{W, PV, ES}), initialize the state variable xi and optimization parameters (such as learning rate α), and the global coordinator initializes the global variable z and the number of iterations k); S42, each agent performs autonomous optimization based on local information and its own objective function, and uses an optimization algorithm to solve the local optimization problem of each agent (the local optimization algorithm formula is: calculate the local gradient Where fi is the objective function of agent i, and updates the local state variables: ) S43. Design a global coordinator (GC) to issue global instructions and coordinate the behaviors among agents. The global coordinator generates instructions based on global goals and status information (such as total energy demand and grid stability) and sends them to each agent. S44. Establish communication mechanisms between agents, allowing them to exchange necessary information, and design collaborative strategies so that agents can respond to global instructions and adjust their own optimization strategies to maintain consistency; S45. Select an appropriate distributed optimization algorithm to decompose the global optimization problem into multiple sub-problems. Each sub-problem is solved by a corresponding intelligent agent, and the global optimal solution is gradually approached through an iterative process (global variable: zk+1=N1∑i=1Nxik+1, where N is the number of intelligent agents); S46. Set the number of iterations or convergence conditions, and repeat until the preset number of iterations and convergence conditions are reached. When the algorithm converges, output the global optimal solution (iteration formula: |||zk+1-zk|||<∈, optimal solution output formula: x*=zK).

6. A coordinated control method for a smart grid according to claim 5, characterized in that: In step S5, a machine learning algorithm is used to perform short-term and long-term forecasts on the power grid load, generate a load forecast curve, and formulate a load response strategy according to the load forecast results and the real-time status of the power grid. The specific steps are as follows: S51. Real-time monitoring of the operation status of the power grid, including current load, power generation capacity, and energy storage status; S52. Adjust the electricity price strategy based on the load forecast results and real-time status to guide users to reduce electricity load during peak hours. S53, store excess power during low load periods and release the stored energy during peak hours to balance the grid load; S54. Implement the formulated load response strategy and monitor the implementation effect; Load forecasting algorithm model: Let X be the feature matrix, y be the target load vector, and use SVR for regression: minimize 21∥w∥2+C∑i=1nξi+ξi*y_i-(w\cdot\phi(x_i))-b\leq\epsilon+\xi_i\\(w\cdot\phi(x_i))+b-y_i\leq\epsilon+\xi_i^*\\\xi_i,\xi_i^*\geq 0\end{cases}$$where $w$ is the weight vector, $b$ is the bias term, $\phi(x_i)$ is the function that maps the input $x_i$ to a high-dimensional space, $\epsilon$ is the allowed deviation, $C$ is the regularization parameter, and $\xi_i,\xi_i^*$ are slack variables. ####3. Load response strategy - electricity price adjustment Assume $p_t$ is the electricity price in the $t$th period, and adjust the electricity price according to the load forecast results and real-time status: _t=p_0+\alpha\left(\frac{L_t{\text{avg}}}{\sigma_L}\right)$$Wherein, p0 is the basic electricity price, α is the adjustment coefficient, Ltpred is the predicted load value, Ltavg is the historical average load value, and σL is the standard deviation of the load.

7. A coordinated control method for a smart grid according to claim 6, characterized in that: In step S6, the operation status of the power grid is monitored in real time, the fault point is quickly located using intelligent fault detection, and a rapid response mechanism is established. The specific steps are as follows: S61. Use multiple fault detection principles to determine whether there is a fault by comparing the difference in electrical quantities on both sides of the line or analyzing the propagation characteristics of the traveling wave signal. Use machine learning and deep learning artificial intelligence technology to learn and analyze the historical fault data and real-time data of the power grid; S62, after the fault is detected, the fault location method is started to determine the specific location of the fault point, and the fault point is quickly located through analysis in combination with the topological structure of the power grid and real-time data; S63. When a fault is detected and located, an alarm signal is immediately sent to the control center, and the fault information is transmitted to the dispatcher and maintenance team through the communication network; S64. After confirming the fault point, quickly isolate the fault area to prevent the fault from spreading and affecting other non-fault areas. Start the backup equipment and lines to restore power supply to the affected area to reduce power outage time and losses. At the same time, formulate a fault handling plan based on the type and severity of the fault and organize a maintenance team to carry out emergency repairs; S65. After the fault is handled, conduct an in-depth analysis of the fault data to identify the cause of the fault and potential risks. Based on the analysis results, formulate corresponding preventive measures and improvement plans, summarize and evaluate the entire fault handling process, and continuously improve and optimize the rapid response mechanism.

8. A coordinated control method for a smart grid according to claim 7, characterized in that: The specific steps of step S7 in strengthening the interaction between the power grid and users, collecting user electricity consumption information through smart meters, smart homes and other devices, and realizing demand side management are as follows: S71, using smart meters and smart home devices to collect users' electricity consumption data in real time, including power consumption, voltage, current and power factor, and transmitting the collected data to the power grid data center through power line carrier communication and wireless communication; S72. Use big data analysis and artificial intelligence technology to conduct in-depth mining and analysis of user electricity consumption data. Identify users' electricity consumption patterns, load characteristics, and energy-saving potential, and formulate demand-side management strategies based on the analysis results; S73. Establish an interactive platform between the power grid and users, release electricity price information, load status and energy-saving suggestions to users in real time through SMS and APP push, and encourage users to participate in demand response plans to balance the power grid load and improve energy efficiency by adjusting electricity consumption behavior; S74. Regularly evaluate and provide feedback on the effectiveness of demand-side management, collect user opinions and suggestions, and continuously improve and optimize demand-side management strategies and measures to ensure the continued and effective advancement of demand-side management and continuously enhance user satisfaction and sense of gain.

9. A coordinated control method for a smart grid according to claim 8, characterized in that: The specific steps of establishing an efficient security monitoring system in step S8, monitoring the potential safety hazards in the power grid in real time, and taking multi-level security protection measures to protect key equipment and systems from attacks are as follows: S81. Ensure that grid equipment and monitoring equipment are placed in a secure environment to prevent physical damage and unauthorized access, and implement strict physical access control measures; S82. Deploy firewalls and intrusion detection and prevention systems (IDS / IPS) network security devices between the power grid and the Internet to prevent external network attacks, use encryption technology to protect the security of data transmission, ensure that sensitive data is not intercepted or tampered with, and regularly scan network devices for security vulnerabilities and update patches to repair potential security vulnerabilities. S83. Conduct regular security audits and penetration tests on power grid control systems and monitoring platforms to identify and promptly repair security vulnerabilities, adopt access control policies to limit user permissions to ensure that only authorized users can access key equipment and systems, and deploy reliable antivirus software and security update mechanisms to prevent malware intrusion and damage; S84. Encrypt sensitive data in the power grid for storage and transmission, ensure data security during storage and transmission, implement data backup and recovery strategies, regularly back up key data and store them in offline or remote secure locations, and establish a comprehensive data access control mechanism to prevent unauthorized access and leakage; S85. Develop a detailed emergency response plan, clarify the handling process and division of responsibilities when a security incident occurs, conduct emergency drills and training regularly to enhance the ability of operation and maintenance personnel to respond to emergencies, and establish a rapid recovery mechanism to ensure that the normal operation of the power grid can be quickly restored after an attack or failure.

Images