Method and device for improving clean energy consumption efficiency through load allocation and storage medium

Abstract

The invention relates to the field of clean energy, and provides a method for improving clean energy consumption efficiency through intelligent power load management, and the method comprises the steps: collecting power grid operation data and clean energy power generation data through an intelligent monitoring system; using the data analysis prediction model to predict the power grid load and the clean energy generating capacity; according to the prediction result and the current state of the power grid, dynamically adjusting power grid operation parameters, and optimizing power load distribution; during the peak period of the power grid load, reducing the load demand through a demand response strategy, and preferentially using clean energy; in the off-peak period of the power grid load, the energy storage system is used for storing excess clean energy for use in the peak period; and monitoring a power grid operation state and a clean energy consumption condition in real time, and adjusting a load management strategy according to actual operation data. Therefore, the absorption efficiency of clean energy is improved.

Description

Technical Field

[0001] The present invention relates to the field of clean energy, and in particular to a method, a device and a storage medium for improving the clean energy consumption efficiency. Background Art

[0002] In the power industry, accurate prediction of power load is crucial to improving the utilization rate of clean energy. In-depth analysis of historical data, combined with changes in economic development trends and climate conditions, and the use of efficient forecasting technology can significantly improve the accuracy of forecasts of future power consumption patterns. Such forecasting capabilities are key to clean energy power generation companies in formulating power generation plans and optimizing resource allocation.

[0003] Furthermore, by real-time monitoring of clean energy generation and combining it with the current operating status of the power grid, a more efficient dispatching strategy can be designed. This optimized dispatching can not only reduce the waste of wind and solar power generation, but also significantly improve the overall utilization efficiency of clean energy.

[0004] The introduction of smart grid technology has brought new features of automation, informatization and interactivity to the power system. The integration of these technologies has significantly improved the operating efficiency of the power grid and enhanced the grid's ability to accept and process clean energy.

[0005] Demand-side management and its response measures effectively balance the load on the grid by encouraging users to use more electricity during periods of low electricity demand. This management approach not only reduces the pressure on the grid during peak hours, but also creates more opportunities for clean energy consumption.

[0006] In addition, the reasonable layout and configuration of energy storage systems have played an indispensable role in balancing the load of the power grid and improving the clean energy consumption capacity. Storing energy during peak power generation and releasing energy during peak demand effectively alleviates the contradiction between supply and demand of the power grid.

[0007] However, my country is still facing some challenges in absorbing new energy power. The main problems include the irrationality of the power structure, the mismatch between the planning and layout of new energy resources, and the imperfection of the power market mechanism, which have restricted the effective absorption and optimal allocation of new energy power. Summary of the invention

[0008] The purpose of the present invention is to improve the consumption efficiency of clean energy.

[0009] At least in part for the above purpose, the present invention provides a method for improving clean energy consumption efficiency through load dispatching, comprising the following steps:

[0010] S1. Collect grid operation data and clean energy power generation data through intelligent monitoring systems;

[0011] S2. Use data analysis and prediction models to predict grid load and clean energy generation;

[0012] S3. Dynamically adjust the grid operation parameters and optimize the power load distribution according to the prediction results and the current state of the grid;

[0013] S4. During peak hours of grid load, reduce load demand through demand response strategies and give priority to the use of clean energy;

[0014] S5. During the off-peak period of the power grid load, the energy storage system is used to store excess clean energy for use during peak hours;

[0015] S6. Monitor the grid operation status and clean energy consumption in real time, and adjust the load management strategy based on actual operation data.

[0016] Optionally, in step S2, the prediction model includes an algorithm of time series analysis, neural network or support vector machine.

[0017] Optionally, step S3 further comprises:

[0018] Dynamically adjust the grid operating parameters, and establish charging and discharging models to adjust power generation plans, optimize power flows, and control energy storage systems.

[0019] Optionally, in step S4, the demand response strategy includes encouraging users to reduce electricity consumption or shift electricity consumption time during peak hours; and the charging and discharging model includes:

[0020]

[0021] Among them, the constraints of the charge and discharge model are:

[0022]

[0023] E min ≤E t+1 ≤E max ,as well as,

[0024] P charge : Charging power.

[0025] P discharge : Discharge power.

[0026] P charge,max : Maximum charging power.

[0027] P discharge,max : Maximum discharge power.

[0028] S charge: Charge status, which may refer to the battery's current charge level or energy storage state.

[0029] S discharge : Discharge state, which may refer to the battery's current discharge level or energy storage state.

[0030] S set : The set energy storage state, which may be a target or reference level.

[0031] μ charge and μ discharge : Controls the exponential rate of charge and discharge.

[0032] E t : Energy level of the energy storage system at time t.

[0033] E t+1 : Energy level of the energy storage system at time t+1.

[0034] η charge : Charging efficiency.

[0035] η discharge : Discharge efficiency.

[0036] Δt: time step.

[0037] E min : Minimum energy level of the energy storage system.

[0038] E max : Maximum energy level of the energy storage system.

[0039] λ charge and λ discharge : Weight coefficient of charging and discharging.

[0040] P charge,ref and P discharge,ref : Reference power for charging and discharging.

[0041] P grid,max : Maximum transmission power of the power grid.

[0042] α: exponential coefficient related to the end effect of the energy storage system.

[0043] λ storage : Final energy storage state weight coefficient.

[0044] E ref : Reference value of energy storage status.

[0045] E T : Energy level of the energy storage system at the final time T.

[0046] E final,ref: Reference value of the final energy storage state.

[0047] In another alternative embodiment, a charge and discharge model is provided, assuming that P charge (t) is the charging power at time t, P discharge (t) is the discharge power at time t. The charge and discharge model can be defined as follows:

[0048] 1. When grid load demand is low and clean energy generation is high (i.e. there is excess electricity):

[0049] P charge (t) = k1 × (P clean (t)-L(t))

[0050] Among them, P charge (t) is the clean energy power generation at time t.

[0051] L(t) is the grid load at time t.

[0052] k1 is a positive proportional coefficient used to adjust the charging rate and can be set according to actual needs.

[0053] 2. When the grid load demand is high and the energy storage system has enough energy to release:

[0054] P discharge (t) = k2 × (L (t) - P clean (t))

[0055] Among them, k2 is also a positive proportional coefficient used to adjust the discharge rate.

[0056] 3. Energy level update of energy storage system:

[0057]

[0058] in,

[0059] E(t) is the energy level of the energy storage system at time t.

[0060] ·η charge and η discharge They are charging efficiency and discharging efficiency respectively.

[0061] Δt is the time step.

[0062] 11. Boundary conditions:

[0063] Ensure that the energy level E(t) of the energy storage system does not exceed its maximum capacity E max or below the minimum capacity E min Therefore, after each update, perform the following checks:

[0064] If E(t+1)>E max , then let E(t+1)=E max .

[0065] If E(t+1) <E min , then let E(t+1)=E min .

[0066] This charging and discharging model determines the charging and discharging behavior of the energy storage system by directly linking clean energy generation to grid load, while maintaining basic control over the energy level of the energy storage system, thereby providing a clear and intuitive approach to handling common energy storage operation scenarios.

[0067] The technical effect analysis of the above-mentioned charging and discharging model is as follows:

[0068] 1. Real-time responsiveness

[0069] Rapid response to grid changes: Since the model uses a direct linear relationship, it can respond more quickly to changes in grid load and clean energy generation. This is particularly effective for handling instantaneous fluctuations in power demand.

[0070] Flexible adjustment of charge and discharge power: Through simple proportional coefficients k1 and k2, the charge and discharge rate can be flexibly adjusted according to actual conditions to adapt to different grid operating conditions.

[0071] 2. Enhance system robustness

[0072] Reduce the risk of overfitting: A complex model may perform well on a specific data set, but may have overfitting problems in actual applications. The charge-discharge model in this embodiment has a smaller number of parameters, which reduces the risk of overfitting and improves the generalization ability of the model.

[0073] Improve fault tolerance: The simple charge and discharge model in this embodiment is usually less susceptible to data noise, so it can still maintain good performance when facing inaccurate data.

[0074] 3. Optimize energy utilization

[0075] Improve clean energy consumption: When clean energy generation exceeds the grid load, the model can quickly start charging operations and store excess clean energy, thereby reducing wind and solar power abandonment.

[0076] Balancing grid load: During peak hours of grid load, the model can release energy from the energy storage system in a timely manner, smooth the load curve, and improve the stability and reliability of the grid.

[0077] 4. Reduce operating costs

[0078] Reduce equipment wear: The charge and discharge model in this embodiment avoids frequent charge and discharge operations, which helps to extend the service life of the energy storage system and reduce maintenance costs.

[0079] Optimize the use of energy storage systems: Through reasonable charging and discharging strategies, the model can maximize the use of the capacity of the energy storage system and reduce unnecessary charging and discharging times, thereby reducing operating costs.

[0080] 5. Easy to expand and integrate

[0081] Modular design: The charging and discharging model in this embodiment can be easily integrated with other system modules (such as prediction module, optimization scheduling module) to form a complete intelligent power load management system.

[0082] Support multi-scenario applications: The model has a clear design concept and can adapt to energy storage systems of different sizes and types by adjusting the proportional coefficients k1 and k2, making it suitable for a variety of application scenarios.

[0083] Optionally, in step S5, the energy storage system includes but is not limited to a battery energy storage system, a pumped storage system or other forms of energy storage systems.

[0084] Optionally, in step S6, real-time monitoring is achieved through Internet of Things technology to ensure the real-time and accuracy of the data.

[0085] Optionally, the prediction model can be defined as:

[0086]

[0087] Among them, α, β, γ represent the weight coefficients of time series analysis, neural network and support vector machine respectively, and α+β+γ=1. They represent the weight of the prediction value within each model, and N and M represent the number of prediction values ​​for time series analysis, neural network, and support vector machine, respectively.

[0088] Optionally, the charge and discharge model includes one or more of the following:

[0089] A) Charging model during low-peak hours:

[0090] E battery (t+1)=E battery (t)+η charge,battery (Eb attery (t))·min(P surplus (t), P charge，battery )·Δt

[0091] E pump (t+1)=E pump (t)+ηcharge，pump (E pump (t))·min(P surplus (t)-P charge，battery (t), P charge，pump )·Δt

[0092] Among them, P surplus (t) = P clean (t)-L(t)

[0093] B) Discharge model during peak hours:

[0094]

[0095] Among them, P demand (t) = L(t) - P clean (t)

[0096] C) Generation plan adjustment model:

[0097] P gen (t) = P demand (t)-P discharge，battery (t)-P discharge，pump (t)

[0098] Among them, P gen (t) = max(P demand (t)-P discharge，battery (t)-P discharge，pump (t), 0)

[0099] D) Power flow optimization model:

[0100] P flow (t) = min(P gen (t), P max，grid )-max(P surplus (t)-P charge，battery (t)-P charge，pump (t), 0)

[0101] Among them, P max，grid is the maximum carrying power of the power grid;

[0102] E) Efficiency Model:

[0103] η charge，battery (E battery ) = η charge，battery，max -log(1+exp(-λ b ·(E battery -E center，battery )))

[0104] η charge，pump (E pump ) = ηcharge，pump，max -log(1+exp(-λ p ·(E purnp -E center，pump )))

[0105] η discharge，battery (E batterry ) = η discharge，battery，max +log(1+exp(λ b ·(E battery -E center，battery )))

[0106] η discharge，pump (E pump ) = η discharge，pump，max +log(1+exp(λ p ·(E pump -E center，pump ))).

[0107] In other embodiments, an electronic device is also proposed, including a processor, which is operable to execute the method for improving clean energy consumption efficiency through load dispatching as described in any other embodiment of the present application.

[0108] In other embodiments, a non-volatile storage medium is also proposed, which stores a program that can be activated to execute the method for improving the clean energy consumption efficiency through load allocation described in any other embodiment.

[0109] In some other embodiments, a device for improving clean energy consumption efficiency through load dispatching is also proposed, including:

[0110] Processor module;

[0111] A data collection unit, operable to obtain analog signals of actual active output of wind power, actual active power of electric heating load, grid connection point voltage, and grid connection point current;

[0112] A prediction unit, operable to convert the data obtained by the data acquisition module into parallel data through an analog-to-digital converter and transmit the parallel data to the processor module;

[0113] An optimization dispatching unit operable to optimize the power load according to the real-time collected power load data, clean energy generation data and energy storage system status;

[0114] A real-time monitoring module, which can be operated to monitor the power information of the power grid and the operating status information of the energy storage system in real time, and control the charging and discharging status of the energy storage system;

[0115] The energy storage module can be operated to configure the energy storage system according to the power load status information, improve the grid's ability to accept renewable energy power generation, and promote the consumption of renewable energy.

[0116] Optionally, the data collection unit further comprises:

[0117] A first sliding rheostat is operable to input a wind power actual active output analog signal;

[0118] A second sliding rheostat is operable to input an actual active power simulation signal of the electric heating load;

[0119] A third sliding rheostat operable to input a grid connection point voltage analog signal;

[0120] a fourth sliding rheostat operable to input a grid-connected point current analog signal;

[0121] Wherein, the prediction unit further includes:

[0122] an analog-to-digital converter operable to convert the analog signal into a digital signal;

[0123] a processor module operable to process data from the analog-to-digital converter and perform calculations;

[0124] A display operable to display the calculation results of the processor module; and the optimization scheduling unit further comprises:

[0125] A load stability evaluation value calculation unit, operable to obtain a regional load stability evaluation value through processing according to the power load at each monitoring time point in the region within a historical time period;

[0126] A new energy output stability evaluation value calculation unit is operable to analyze and obtain a new energy output stability evaluation value based on the wind power output and photovoltaic output at each monitoring time point in the region within a historical time period;

[0127] The energy storage configuration demand index calculation unit can be operated to analyze and obtain the regional energy storage configuration demand index according to the regional load stability evaluation value and the new energy output stability evaluation value.

[0128] Optionally, the real-time monitoring module further includes:

[0129] A control time setting unit operable to set a number of control times;

[0130] An environmental information collection unit, operable to collect environmental information of the new energy station at each control moment;

[0131] A prediction value calculation unit, operable to obtain a prediction value of the output of the new energy source at each control moment and a load prediction value at each control moment through comprehensive analysis based on a new energy output prediction model;

[0132] The energy storage configuration state adjustment evaluation value calculation unit is operable to obtain the power grid power information and the energy storage system operation state information at each control moment, and obtain the energy storage configuration state adjustment evaluation value at each control moment through comprehensive analysis based on the predicted value of the new energy output and the load predicted value at each control moment;

[0133] Wherein, the energy storage module further comprises:

[0134] An energy storage device screening unit operable to screen energy storage devices corresponding to energy storage configuration types in a statistical area from energy storage devices corresponding to energy storage configuration types stored in a power grid database;

[0135] An integration unit operable to integrate the energy storage devices in the screened configurations as the energy storage system in a regional configuration;

[0136] A charge and discharge control unit operable to control the energy storage system to enter a charging state when the energy storage configuration state adjustment evaluation value is lower than a state adjustment evaluation critical value, and to control the energy storage system to enter a discharging state when the energy storage configuration state adjustment evaluation value is higher than the state adjustment evaluation critical value;

[0137] a dust removal mechanism operable to clean dust inside the storage device;

[0138] The heat dissipation mechanism can be operated to ensure the normal discharge of heat dissipated by the absorption platform body during high-power operation.

[0139] Optionally, the device for improving the clean energy consumption efficiency through load dispatch in some embodiments further includes a memory for storing executable instructions of the processor module; wherein the processor is configured to execute the method for improving the clean energy consumption efficiency through load dispatch described in any other embodiment of the present application by executing the executable instructions.

[0140] Through some embodiments of the present application, the grid operation data and clean energy power generation data collected by the intelligent monitoring system can more accurately predict the grid load and clean energy power generation, providing reliable data support for subsequent load management. Thus, by accurately predicting clean energy power generation and grid load, the grid operation parameters can be adjusted in advance, and the power load distribution can be optimized, thereby reducing the abandonment of clean energy and improving the utilization rate of clean energy. Furthermore, by collecting grid operation data and clean energy power generation data through the intelligent monitoring system, the grid operation status can be grasped in real time, the grid operation parameters can be adjusted dynamically, and the operation efficiency of the grid can be improved. Thus, by giving priority to the use of clean energy and using the energy storage system to store excess clean energy, clean energy can be absorbed and utilized to the maximum extent, reducing dependence on traditional energy.

[0141] In a further embodiment, multiple related models are fused: the prediction results of the three models are integrated using stacking generalization technology, and different weights are assigned according to their respective prediction performances. The consumption characteristics of clean energy are characterized from more dimensions, which leads to more accurate prediction results and greater improvement in the consumption efficiency of clean energy. BRIEF DESCRIPTION OF THE DRAWINGS

[0142] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0143] Figure 1 The figure is a schematic diagram of signal flow in a vehicle decision-making method based on a PFC model according to an embodiment of the present invention.

[0144] Figure 2 A schematic diagram of the structure of an optimization scheduling device for an integrated smart energy system based on fuzzy logic according to an embodiment of the present application is shown;

[0145] Figure 3 A schematic diagram showing the structure of a computing device according to an embodiment of the present application is shown;

[0146] Figure 4 A schematic diagram of the structure of a clean energy consumption efficiency improvement device according to an embodiment of the present application is shown;

[0147] Figure 5 An external structural diagram and an internal circuit module schematic diagram of a clean energy consumption efficiency improvement device according to an embodiment of the present application are shown. DETAILED DESCRIPTION

[0148] In one embodiment of the present application, Figure 1 As shown in the figure, a method for improving the clean energy consumption efficiency through intelligent power load management is proposed, including the following steps:

[0149] S1. Collect power grid operation data and clean energy power generation data through intelligent monitoring system;

[0150] S2. Use data analysis and prediction models to predict grid load and clean energy generation;

[0151] S3. Dynamically adjust grid operation parameters and optimize power load distribution based on prediction results and current grid status;

[0152] S4. During peak hours of the power grid, reduce load demand through demand response strategies and give priority to the use of clean energy;

[0153] S5. During the off-peak period of the power grid load, the energy storage system is used to store excess clean energy for use during peak hours;

[0154] S6. Monitor the operation status of the power grid and the consumption of clean energy, and adjust the load management strategy according to the actual operation data.

[0155] In some embodiments, a fusion model is constructed, and one of the formulas will express the fusion model in detail. In some embodiments of the present application, time series analysis, neural networks and support vector machines (SVM) are used as basic prediction models, and their prediction results are integrated through a complex fusion model. The following are detailed mathematical formulas and algorithm steps:

[0156] Data preprocessing

[0157] 1. Logarithmic transformation:

[0158] and

[0159] The input data X is logarithmically transformed to make the data distribution more even and reduce the impact of extreme values.

[0160] The target variable Y is logarithmically transformed to improve the data distribution.

[0161] 2. Standardization processing:

[0162] and

[0163] Among them, μ X and σ X Respectively The mean and standard deviation of Y and σ Y Respectively The mean and standard deviation of the input data after standardization, where μ X and σ XRepresent the original logarithmic transformed data The mean and standard deviation of .

[0164] Model Fusion

[0165] Next, some embodiments of the present application will use time series analysis, neural networks and support vector machines (SVM) to predict future values ​​and blend them together.

[0166] 1. Time Series Analysis:

[0167]

[0168] The autoregressive integrated moving average (ARIMA) model is used to normalize the input data. Make forecasts and get the predicted values ​​for time series analysis.

[0169] 2. Neural Networks:

[0170]

[0171] Use the neural network model to normalize the input data Make predictions and get the predicted values ​​of the neural network.

[0172] 3. Support Vector Machine (SVM):

[0173]

[0174] Use the support vector regression (SVR) model to normalize the input data Make predictions and get the predicted values ​​of the support vector machine.

[0175] 4. Model Fusion:

[0176]

[0177] Use stacking technology to take the prediction results of time series analysis, neural network and support vector machine as input features, and then train an integrated model (here is another neural network NN stack ) to make the final prediction.

[0178] Weighted average calculation

[0179] 1. Weight coefficient:

[0180] oα, β, γ: represent the weight coefficients of time series analysis, neural network and support vector machine respectively, and α+β+γ=1.

[0181] 2. Internal prediction value weight:

[0182] o Represent the weight of the prediction value within each model.

[0183] 3. Number of predicted values:

[0184] o N: represents the number of prediction values ​​within each model.

[0185] o M: represents the number of components of each predicted value within each model.

[0186] The fusion model is expressed as:

[0187]

[0188] 1. Final prediction value calculation:

[0189] o The final forecast value is obtained by integrating the results of time series analysis, neural network and support vector machine.

[0190] 2. Weighted average calculation:

[0191]

[0192] Compute the weighted mean prediction for each model.

[0193] 3. Logarithmic and exponential operations:

[0194] Each forecast value for time series analysis Perform the exponential operation, sum it, and then take the logarithm.

[0195] For each prediction value of the neural network Perform the exponential operation, sum it, and then take the logarithm.

[0196] For each prediction value of the support vector machine Perform the exponential operation, sum it, and then take the logarithm.

[0197] Practical Application

[0198] 1. Data preprocessing: logarithmic transformation and standardization of data.

[0199] 2. Single model prediction: prediction is performed using time series analysis, neural network, and support vector machine respectively.

[0200] 3. Model fusion: Use stacking generalization technology to integrate the prediction results of the three models and assign different weights according to their respective prediction performance.

[0201] 4. Denormalization and exponential transformation: restore the prediction results of the fusion model to the original data scale.

[0202] 5. Application of prediction results: Dynamically adjust grid operation parameters according to prediction results to optimize power load distribution.

[0203] Summary This formula uses logarithmic and exponential operations to process data and combines the prediction results of time series analysis, neural network and support vector machine by weighted average, aiming to improve the accuracy of prediction. Through the above detailed explanation, we have a clear understanding of each variable in the formula.

[0204] In some embodiments of the present application, stacking technology is used to take the prediction results of time series analysis, neural network and support vector machine as input features, and then train an integrated model to perform the final prediction.

[0205] The ensemble model can be another neural network, a decision tree, etc.

[0206] In some embodiments of the present application, the concept of weighted average is also introduced so that each model can be given different weights according to its prediction performance.

[0207] Fusion model expressed in mathematical formula

[0208] In order to construct and fuse the model, some embodiments of the present application will use a neural network containing multiple layers as an integrated model. The following is the mathematical formula of this complex model:

[0209]

[0210] In this formula, α, β, and γ represent the weight coefficients of time series analysis, neural network, and support vector machine, respectively, and α+β+γ=1. They represent the weight of the prediction value within each model, and N and M represent the number of prediction values ​​for time series analysis, neural network, and support vector machine, respectively.

[0211] Denormalization and Exponential Transformation

[0212] 1. Anti-standardization:

[0213]

[0214] 2. Antilogarithmic transformation:

[0215]

[0216] Practical Application

[0217] In practical application, you can follow the steps below:

[0218] 1. Data preprocessing: logarithmic transformation and standardization of data.

[0219] 2. Single model prediction: prediction is performed using time series analysis, neural network, and support vector machine respectively.

[0220] 3. Model fusion: Use stacking generalization technology to integrate the prediction results of the three models and assign different weights according to their respective prediction performance.

[0221] 4. Denormalization and exponential transformation: restore the prediction results of the fusion model to the original data scale.

[0222] 5. Application of prediction results: Dynamically adjust grid operation parameters according to prediction results to optimize power load distribution.

[0223] Please note that the above formula and steps are a conceptual and simplified example. In actual applications, the model will be more complex and may require more details and parameter adjustments. In addition, in order to obtain the best prediction performance, the model parameters need to be carefully tuned, and it may be necessary to use methods such as cross-validation to evaluate the performance of the model.

[0224] In order to better illustrate how to substitute specific application scenarios and detailed parameters into the formula model proposed above, and to prove the technical effect brought by the fusion model, a specific example will be constructed in some embodiments of the present application.

[0225] Application Scenario

[0226] Assume that in some embodiments of the present application, a power load forecasting system is being developed, the goal of which is to predict future power loads and clean energy generation to optimize the operation of the power grid and improve the clean energy absorption capacity. In some embodiments of the present application, time series analysis, neural networks, and support vector machines (SVMs) are used as basic forecasting models, and their forecasting results are integrated through a complex fusion model.

[0227] Parameter settings

[0228] 1. Dataset: Historical power load data and clean energy power generation data, covering one year of data records.

[0229] 2. Data preprocessing:

[0230] oLogarithmic transformation: and

[0231] oStandardization: and

[0232] 3. Model parameters:

[0233] oTime series analysis (ARIMA): p=2, d=1, q=2.

[0234] oNeural network (LSTM): The number of hidden layer nodes is 64, and the activation function is ReLU.

[0235] oSupport Vector Machine (SVM): The kernel function is Radial Basis Function (RBF),

[0236] C=1.0,γ=0.1.

[0237] 4. Fusion model parameters:

[0238] oWeight coefficients: α=0.4, β=0.3, γ=0.3.

[0239] oEnsemble model: A multi-layer perceptron (MLP) neural network is used with 128 hidden layer nodes and ReLU activation function.

[0240] 5. Other parameters:

[0241] o N represents the number of predicted values, which is set to 10.

[0242] o M represents the number of predictions within each prediction value, which is set to 5.

[0243] o Represent the weights of the prediction values ​​within each model, assuming a uniform distribution, that is, ωi=1 / N.

[0244] Calculation of the fusion model expressed in formula

[0245] Next, in some embodiments of the present application, specific numerical values ​​will be used to demonstrate the calculation process of the fusion model. For the sake of simplicity, in some embodiments of the present application, fictitious data points will be used.

[0246] Specific values

[0247] Assume that in some embodiments of the present application, there are the following predicted values ​​(these values ​​are values ​​after logarithmic transformation and standardization):

[0248] Predicted value of time series analysis:

[0249] Prediction value of neural network:

[0250] Support vector machine predictions:

[0251] Calculation process

[0252] 1. Calculate the weighted average prediction value of each model:

[0253]

[0254] Specific calculation example

[0255] In order to simplify the calculation, it is assumed that the number of predictions M within each prediction value is 5, and some embodiments of the present application only show the calculation process of the first prediction value. For i=1, some embodiments of the present application have:

[0256] Internal forecast values ​​of time series analysis:

[0257] Inside the neural network’s predictions:

[0258] The prediction value inside the support vector machine:

[0259] 1. Calculate the weighted average forecast value for time series analysis:

[0260]

[0261] 2. Calculate the weighted average prediction value of the neural network:

[0262]

[0263] Compute the weighted average prediction of the support vector machine:

[0264]

[0265] Calculation process in the prior art

[0266] In the prior art, the prediction results of the three models are simply averaged as the final prediction result.

[0267] 1. Calculate the average predicted value:

[0268]

[0269] Comparison of technical effects By using the above complex formula, some embodiments of the present application can synthesize information from multiple prediction models to obtain more accurate prediction results. Specifically, this method can:

[0270] 1. Improve prediction accuracy: By integrating the advantages of multiple models, the bias and variance in the prediction of a single model can be reduced.

[0271] 2. Enhanced robustness: Even if the prediction effect of some models is poor, the overall prediction results remain stable through the supplementation of other models.

[0272] 3. Strong adaptability: This method can flexibly adapt to different types of prediction tasks, and the weight of the model can be adjusted according to actual conditions.

[0273] In contrast, the simple average method in the prior art is simple to implement, but when faced with complex problems, it may not be able to fully utilize the information of each model and cannot characterize the consumption characteristics of clean energy from more dimensions, which leads to inaccurate prediction results. Through specific numerical calculations and parameter settings, some embodiments of this application demonstrate the application of mathematical models expressed by complex formulas in power load forecasting and their technical effects relative to the prior art.

[0274] If the three models are operated independently, the process is as follows:

[0275] 1. Time series analysis forecast value:

[0276]

[0277] 2. Neural network prediction value:

[0278]

[0279] 3. Support vector machine prediction value:

[0280]

[0281] Comparison of technical effects By using the above complex formula, some embodiments of the present application can synthesize information from multiple prediction models to obtain more accurate prediction results. Specifically, this method can:

[0282] 1. Improve prediction accuracy: By integrating the advantages of multiple models, the bias and variance in the prediction of a single model can be reduced.

[0283] 2. Enhanced robustness: Even if the prediction effect of some models is poor, the overall prediction results remain stable through the supplementation of other models.

[0284] 3. Strong adaptability: This method can flexibly adapt to different types of prediction tasks, and the weight of the model can be adjusted according to actual conditions.

[0285] In contrast, although the independent operation of the three models can provide diverse prediction results, it may not be able to fully utilize the information of each model when facing complex problems, resulting in inaccurate prediction results. Through specific numerical calculations and parameter settings, some embodiments of the present application demonstrate the application of complex formulas in power load forecasting and their technical effects relative to the independent operation of the three models.

[0286] By comparing the above calculation process and results, it can be seen that the fusion model expressed by the formula has a significant improvement in prediction accuracy compared to the independent operation of the three models. This is because the complex formula can more effectively integrate the information of multiple models by weighted averaging the prediction results of each model and using logarithmic and exponential operations to further smooth the prediction results. This makes the complex formula have better prediction ability and robustness in practical applications.

[0287] In summary, through specific numerical calculations and parameter settings, the application of complex models in power load forecasting and the technical effects they bring are demonstrated.

[0288] In order to express the "adjustment of power generation plan, optimization of power flow, control of charging and discharging operation of energy storage system" in the claims using a formulaic model including logarithmic calculation and exponential calculation, a comprehensive model will be constructed in some embodiments of the present application, which will take into account multiple aspects such as grid load, clean energy power generation, energy storage system status, and adjustment of power generation plan. The following is an exemplary model that takes into account the above requirements and uses logarithmic and exponential operations to simulate the nonlinear behavior of the system.

[0289] Defining parameters

[0290] 1. Clean energy power generation: Pclean(t) represents the clean energy power generation at time t.

[0291] 2. Grid load: L(t) represents the grid load at time t.

[0292] 3. Storage capacity of battery energy storage system: E battery (t) represents the storage capacity of the battery energy storage system at time tt.

[0293] 4. Storage capacity of pumped storage system: E pump (t) represents the storage capacity of the pumped storage system at time tt.

[0294] 5. Maximum charging power of battery energy storage system: P charge,battery .

[0295] 6. Maximum discharge power of battery energy storage system: P discharge,battery .

[0296] 7. Maximum charging power of pumped storage system: P charge,pump .

[0297] 8. Maximum discharge power of pumped storage system: P discharge,pump .

[0298] 9. Charging efficiency of battery energy storage system: η charge,battery .

[0299] 10. Discharge efficiency of battery energy storage system: η discharge,battery .

[0300] 11. Charging efficiency of pumped storage system: η charge,pump .

[0301] 12. Discharge efficiency of pumped storage system: η discharge,pump .

[0302] 13. Maximum storage capacity of battery energy storage system: E max,battery .

[0303] 14. Maximum storage capacity of pumped storage system: E max,pump .

[0304] 15. Minimum storage capacity of battery energy storage system: E min,battery .

[0305] 16. Minimum storage capacity of pumped storage system: E min，pump .

[0306] 17. The charging efficiency of the battery energy storage system varies as a function of storage capacity:

[0307] η charge,battery (E battery ) = η charge,battery，max -log(1+exp(-λ b ·(E tattery -E center，battery )))

[0308] 18. Charging efficiency of pumped storage system as a function of storage capacity:

[0309] η charge，pump (E pump ) = η charge，pump，max -log(1+exp(-λ P ·(E pump -E center，pump )))

[0310] 19. The discharge efficiency of the battery energy storage system varies as a function of storage capacity:

[0311] η discharge,battery (E battery ) = η discharge , battery，max +log(1+exp(λ b ·(E battery -E center，battery )))

[0312] 20. The discharge efficiency of the pumped storage system as a function of storage capacity:

[0313] η discharge,pump (E pump ) = η discharge，pump，max +log(1+exp(λ p ·(E pump -E center，pump )))Charge and discharge model expressed in formula

[0314] 1. Low period (charging):

[0315] E battery (t+1)=E battery (t)+η charge，battery (E battery (t))·min(P surplus (t), P charge，battery )·Δt

[0316] E pump (t+1)=E pump (t)+η charge，pump (E pump (t))·min(P surplus (t)-P charge，battery (t), P charge，pump )·Δt

[0317] Among them, P surplus (t) = P clean (t)-L(t)

[0318] 2. Peak hours (discharging):

[0319]

[0320] Among them, P demand (t) = L(t) - P clean (t)

[0321] 3. Adjustment of power generation plan:

[0322] P gen (t) = P demand (t)-P discharge,battery (t)-P discharge，pump (t)

[0323] Among them, P gen (t) = max(P demand (t)-P dscharge，battery (t)-P discharge，pump (t), 0)

[0324] 4. Power flow optimization:

[0325] P flow(t) = min(P gen (t), P max，grid )-max(P surplus (t)-P charge，battery (t)-P charge，pump (t), 0)

[0326] Where P max,grid The maximum load capacity of the power grid

[0327] 5. Efficiency function:

[0328] η charge，battery (E battery ) = η charge，battery，max -log(1+exp(-λ b ·(E battery -E charge，battery )))

[0329] η charge，pump (E pump ) = η charge，pump，max -log(1+exp(-λ p ·(E pump -E charge，pump )))

[0330] η discharg，battery (E battery ) = η discharge，battery，max +log(1+exp(λ b ·(E battery -E center，battery )))

[0331] η discharge，pump (E pump ) = η discharge，pump，max +log(1+exp(λ p ·(E pump -E center，pump )))

[0332] Next, in this embodiment, all the above parts are combined into one formula to express the mathematical model. Here, in some embodiments of this application, the focus will be on the charging process during the off-peak period, the discharging process during the peak period, the adjustment of the power generation plan, and the optimization of the power flow.

[0333] E battery (t+1)=E battey (t)+η charge，battery (E battery (t))·min(P clean (t)-L(t),P charge，batter )·Δt

[0334] E pump(t+1)=E pump (t)+η charge，pump (AND pump (t))·mi(P clean (t)-L(t)-P charge，battery (t),P charge，pump )·Δt

[0335] P gen (t)max(L(t)-P clean (t)-P discharge，battery (t)-P discharg，pump (t), 0)

[0336] P flow (t)=min(P gen (t),P max，grid )-max( P clean (t)-L(t)-P charge，battery (t)-P charge，pump (t), 0)

[0337] η charge，battery (AND battery )=η charge，battery，max -log(1+exp(-λ b ·(AND battery -AND center，battery )))

[0338] η charge，pump (AND pump )=η charge，pump，max -log(1+exp(-λ p ·(AND pump -AND center，pump )))

[0339] η discharge，battery (AND battery )=η discharge，pump，max +log(1+exp(λ b ·(AND battery -AND center，battery )))

[0340] η discharge，pump (AND pump )=η discharge，pump，max +log1+exp(λ p ·(AND pump -AND center，pump )))

[0341] This formula integrates the charging process during off-peak hours, the discharging process during peak hours, the adjustment of power generation plans, and the optimization of power flow, including the calculation of charging efficiency, discharge efficiency, and storage capacity update of battery energy storage systems and pumped storage systems. The formula includes logarithmic and exponential operations. To simulate the linear characteristics of energy storage system sex .

[0342] Through the above formula expression related fusion model, some embodiments of the present application successfully express the joint model of battery energy storage system and pumped storage system, as well as power generation plan adjustment, power flow optimization, etc. This model not only considers the basic charging and discharging process, but also considers the nonlinear relationship between storage capacity and charging efficiency, which helps to more accurately simulate the operating characteristics of the actual system.

[0343] In order to perform specific numerical calculations, some embodiments of the present application will use previously defined parameters and formulas and substitute a set of specific numerical values ​​to demonstrate the entire process. The following are specific numerical calculation steps.

[0344] Parameter definition:

[0345] 1. Clean energy power generation: P clean (t) = 1000kW (during the off-peak period).

[0346] 2. Grid load: L(t) = 800kW (during off-peak hours).

[0347] 3. Storage capacity of battery energy storage system: E battery (t) = 500 kWh.

[0348] 4. Storage capacity of pumped storage system: E pump (t) = 1000 kWh.

[0349] 5. Maximum charging power of battery energy storage system: P charge,battery =300kW.

[0350] 6. Maximum discharge power of battery energy storage system: P discharge,battery =300kW.

[0351] 7. Maximum charging power of pumped storage system: P charge,pump =500kW.

[0352] 8. Maximum discharge power of pumped storage system: P discharge,pump =500kW.

[0353] 9. Maximum charging efficiency of battery energy storage system: η chargebattery,max =0.9.

[0354] 10. Maximum discharge efficiency of battery energy storage system: η discharge，battery，max =0.85.

[0355] 11. Maximum charging efficiency of pumped storage system: η charge,pump，max =0.8.

[0356] 12. Maximum discharge efficiency of pumped storage system: η discharge,pump，max =0.75.

[0357] 13. Maximum storage capacity of battery energy storage system: E max,battery =1000kWh.

[0358] 14. Maximum storage capacity of pumped storage system: E max，pump =2000kWh.

[0359] 15. Minimum storage capacity of battery energy storage system: E min，battery =100kWh.

[0360] 16. Minimum storage capacity of pumped storage system: E min，pump =200kWh.

[0361] 17. The charging efficiency of the battery energy storage system varies as a function of storage capacity:

[0362] η charge，battery (E battery )=0.9-log(1+exp(-0.01·(E battery -500)))

[0363] 18. Charging efficiency of pumped storage system as a function of storage capacity:

[0364] η charge，pump (E pump )=0.8-log(1+exp(-0.005·(E pump -1000)))

[0365] 19. The discharge efficiency of the battery energy storage system varies as a function of storage capacity:

[0366] η discharge，battery (E battery )=0.85+log(1+exp(0.001·(E battery -500)))

[0367] 20. The discharge efficiency of the pumped storage system as a function of storage capacity:

[0368] η discharge，pump (E pump)=0.75+log(1+exp(0.005·(E pump -1000)))

[0369] 21. Time step: Δt = 1 hour.

[0370] Numerical calculations

[0371] Low period (charging)

[0372] 1. Remaining power:

[0373] P surplus (t) = P clean (t)-L(t)=1000-800=200kW

[0374] 2. Charging efficiency of battery energy storage system:

[0375] η charge，battery (E battery (t))=0.9-log(1+exp(-0.01·(500-500)))

[0376] =0.9-log(1+exp(0))

[0377] =0.9-log(2)

[0378] ≈0.9-0.6931

[0379] ≈0.2069

[0380] 3. Charging efficiency of pumped storage system:

[0381] η charge，pump (E pump (t))=0.8-log(1+exp(-0.005·(1000-1000)))

[0382] =0.8-log(1+exp(0))

[0383] =0.8-log(2)

[0384] ≈0.8-0.6931

[0385] ≈0.1069

[0386] 4. Update of charging capacity of battery energy storage system:

[0387] E battery (t+1)=E battery (t)+η charge，battery (E battery (t))·min(P surplus (t), Pcharge，battery )·Δt

[0388] =500+0.2069·min(200,300)·1

[0389] =500+0.2069·200

[0390] =500+41.38

[0391] =541.38kWh

[0392] 5. Update on charging capacity of pumped storage system:

[0393] E pump (t+1)=E pump (t)+η chcharge，pump (E pump (t))·min(P surplus (t)-P charge，battery (t,P charge，pump )·Δt

[0394] =1000+0.1069·min(200-200,500)·1

[0395] =1000+0.1069·0

[0396] =1000kWh

[0397] Peak Hour (Discharge)

[0398] Assume that during peak hours, the grid load increases to L(t) = 1200kW and the clean energy generation decreases to P clean (t) = 1000kW.

[0399] 1. Power required:

[0400] P demand (t) = L(t) - P clean (t) = 1200-1000 = 200kW

[0401] 2. Discharge efficiency of battery energy storage system:

[0402] η discharge，battery (E battery (t))=0.85+log(1+exp(0.01·(541.38-500)))

[0403] =0.85+log(1+exp(0.01·41.38))

[0404] =0.85+log(1+exp(0.4138))

[0405] ≈0.85+log(1+1.5142)

[0406] ≈0.85+log(2.5142)

[0407] ≈0.85+0.3998

[0408] ≈1.2498

[0409] 3. Discharge efficiency of pumped storage system:

[0410] η discharge，pump (E pump (t))=0.75+log(1+exp(0.005·(1000-1000)))

[0411] =0.75+log(1+exp(0))

[0412] =0.75+log(2)

[0413] ≈0.75+0.6931

[0414] ≈1.4431

[0415] 4. Update of discharge capacity of battery energy storage system:

[0416]

[0417] 5. Update of discharge capacity of pumped storage system:

[0418]

[0419] During off-peak hours, the battery storage system’s storage capacity increased from 500kWh to 541.38kWh, while the pumped hydro storage system’s storage capacity remained unchanged at 1000kWh.

[0420] During peak hours, the battery system's storage capacity was reduced from 541.38 kWh to 381.35 kWh, while the pumped hydro storage system's storage capacity was reduced from 1,000 kWh to 861.41 kWh.

[0421] Through the above numerical calculations, some embodiments of the present application show the specific process of charging the battery energy storage system and the pumped storage system during the off-peak period and discharging during the peak period. These calculation results show that the energy storage system can effectively store excess clean energy and release it when needed to support the grid load, thereby improving the utilization rate of clean energy. In addition, some embodiments of the present application also take into account the nonlinear characteristics of the energy storage system, which helps to more accurately simulate the operating characteristics of the actual system.

[0422] In order to construct a joint model for adjusting power generation plans, optimizing power flows, and controlling charging and discharging operations of energy storage systems, and using logarithmic and exponential operations in the model, the following optimization problem may be defined in some embodiments of the present application:

[0423] Joint Optimization Model:

[0424] 1. **Objective function** (Minimize Total System Cost):

[0425] min{C fuel ·P thermal +C operation ·(P grid，in +P grid，out )+C storage ·(P charge +P discharge )}

[0426] 2. **Generation plan adjustment** (considering fuel costs and scheduling):

[0427]

[0428] 3. **Power flow optimization** (considering transmission losses and logarithmic terms):

[0429]

[0430] Here, ∈ is a small positive number to avoid mathematical errors of logarithms.

[0431] 4.**Energy storage system control** (considering exponential charging and discharging response):

[0432]

[0433] 5. **Grid balance and energy conservation of energy storage system** (complex constraints):

[0434] P balance (t) = P thermal (t)+P renewable (t)-P demand (t)P grid，loss (t)

[0435]

[0436] subject to E min ≤E t+1 ≤E max

[0437] 6. **Formulated model expression**:

[0438]

[0439] subject to:

[0440]

[0441] E min ≤E t+1 ≤E max

[0442] "Optimal Control": In mathematics and engineering, optimal control usually refers to finding a set of control inputs that optimizes a certain performance indicator, such as minimizing costs or maximizing benefits. In a given formula, "Optimal Control" refers to solving the optimal charging and discharging strategy.

[0443] "subject to": subject to. In mathematical optimization problems, "subject to" is used to introduce constraints to indicate the restrictions on the variables in the optimization problem. In the given formula, "subject to" is followed by the energy conservation and capacity constraints of the energy storage system.

[0444] The following are the meanings of the variables, superscripts, and subscripts used in the above formula:

[0445] ·P charge : Charging power.

[0446] ·P discharge : Discharge power.

[0447] ·P charge,max : Maximum charging power.

[0448] ·P discharge,max : Maximum discharge power.

[0449] ·S charge : Charge status, which may refer to the battery's current charge level or energy storage state.

[0450] ·S discharge : Discharge state, which may refer to the battery's current discharge level or energy storage state.

[0451] ·S set : The set energy storage state, which may be a target or reference level.

[0452] ·μ charge and μ discharge : Controls the exponential rate of charge and discharge.

[0453] ·E t : Energy level of the energy storage system at time t.

[0454] ·E t+1 : Energy level of the energy storage system at time t+1.

[0455] ·η charge : Charging efficiency.

[0456] ·η discharge : Discharge efficiency.

[0457] Δt: time step.

[0458] ·E min : Minimum energy level of the energy storage system.

[0459] ·E max : Maximum energy level of the energy storage system.

[0460] ·λ charge and λ discharge : Weight coefficient of charging and discharging.

[0461] ·P charge,ref and P discharge,ref : Reference power for charging and discharging.

[0462] ·P grid,max : Maximum transmission power of the power grid.

[0463] α: exponential coefficient related to the end effect of the energy storage system.

[0464] ·λ storage : Final energy storage state weight coefficient.

[0465] ·E ref : Reference value of energy storage status.

[0466] ·E T : Energy level of the energy storage system at the final time T.

[0467] ·E final,ref : Reference value of the final energy storage state.

[0468] The derivative symbol arg min in the formula represents finding the parameter combination that minimizes the objective function. The entire formula is an optimization problem, the goal is to minimize the costs and losses associated with charging and discharging operations while satisfying the energy conservation and capacity constraints of the energy storage system.

[0469] Therefore, the whole formula is to find the optimal charging and discharging strategy while satisfying the energy conservation and capacity limitations of the energy storage system.

[0470] In this formula, the optimal control problem is defined in some embodiments of the present application, the goal is to minimize the weighted square error of charging and discharging operations, the logarithmic term of the dynamic response of the power grid, the exponential term of the terminal effect of the energy storage system, and the deviation of the final energy storage state from the expected value. At the same time, energy conservation and capacity constraints of the energy storage system are added in some embodiments of the present application.

[0471] This formula combines the squared errors of charging and discharging, logarithmic terms to represent the dynamic response of the power grid, exponential functions to simulate the end effects of the energy storage system, and constraints on the final energy storage state.

[0472] Technical effect:

[0473] -**Optimization of generation schedules and power flows**: By adjusting generation schedules and optimizing power flows, the model can reduce operating costs and improve grid efficiency.

[0474] -**Effective control of energy storage system**: The model balances the grid load by controlling the charging and discharging operations and improves the system's adaptability to renewable energy.

[0475] -**Grid stability and reliability**: Optimized charging and discharging strategies help smooth the intermittency of renewable energy and enhance grid stability and reliability.

[0476] Please note that this model is a theoretical construction and needs to be adjusted and optimized in actual applications according to specific grid parameters, energy storage system characteristics and market mechanisms.

[0477] Comparison results:

[0478] Existing technologies may simply increase power generation during peak electricity demand periods without considering the dynamic characteristics of energy storage systems and the limitations of grid transmission capacity.

[0479] Models using complex formulas will take into account fluctuations in electricity demand, uncertainties in renewable energy, grid transmission capacity and the dynamic characteristics of energy storage systems, enabling more sophisticated and cost-effective electricity load management.

[0480] Technical effect proof:

[0481] By solving the above complex formula, the following technical effects can be obtained in some embodiments of the present application:

[0482] Cost minimization: By optimizing generation scheduling and storage system operation, grid operators can minimize operating costs.

[0483] Grid stability: The charging and discharging operations of energy storage systems help balance electricity supply and demand and improve grid stability.

[0484] Improved utilization of renewable energy: Excess renewable energy is stored in the energy storage system, reducing the phenomenon of wind and solar power abandonment.

[0485] Power flow optimization: Reduces grid transmission losses and improves power transmission efficiency.

[0486] Consideration of the end effect of energy storage system: Optimizes the operation of energy storage system when it is nearly fully charged or discharged, and improves the flexibility of system response.

[0487] For example, if the optimal charging power P at a certain time point t is calculated in some embodiments of the present application charge and discharge power P discharge In some embodiments of the present application, these values ​​can be analyzed to help balance the grid load, reduce peak-to-valley differences, and improve overall energy efficiency.

[0488] Please note that the actual calculation process will involve a large amount of data and complex mathematical operations, which usually requires professional optimization software to solve. The above parameters and calculation process are only examples and may need to be adjusted according to specific circumstances in actual applications.

[0489] By using models with complex formulas, grid operators can:

[0490] More accurately predict and respond to changes in electricity demand.

[0491] · Use renewable energy more efficiently and reduce the phenomenon of wind and solar power abandonment.

[0492] More optimal dispatch of energy storage resources to improve grid stability and reliability.

[0493] Reduce operating costs and improve energy efficiency.

[0494] Compared with existing technologies, this model shows stronger adaptability and predictive capabilities, especially in applications such as power grid load management that require extremely high accuracy and robustness.

[0495] In one possible design, Figure 3 The data annotation apparatus of the illustrated embodiment may be implemented as a computing device, which may include a storage component 71 and a processing component 72 as shown in the figure;

[0496] The storage component 71 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 72 .

[0497] The processing component 72 is used to: obtain a query object provided by a user; based on the query object, query multiple data to be labeled that match the query object in a pre-established data set to be labeled, and generate multiple initial segmentation results corresponding to the data to be labeled; perform fine segmentation on the initial segmentation results corresponding to the multiple data to be labeled to generate multiple target segmentation results of the data to be labeled; use a pre-created labeling tool to perform labeling operations on the target segmentation results of the selected data to be labeled, and apply the labeling operations in batches to the remaining multiple target segmentation results of the data to be labeled.

[0498] The processing component 72 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components to perform the above method.

[0499] The storage component 71 is configured to store various types of data to support operations at the terminal. The storage component can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0500] Of course, the computing device may also include other components, such as input / output interfaces, display components, communication components, etc.

[0501] The display component 73 can be an electroluminescent (EL) element, a liquid crystal display or a micro display with a similar structure, or a retinal direct display or a similar laser scanning display.

[0502] The input / output interface provides an interface between the processing component and the peripheral interface module, which may be an output device, an input device, etc.

[0503] The communication components are configured to facilitate wired or wireless communications between the computing device and other devices.

[0504] Among them, the computing device can be a physical device or an elastic computing host provided by a cloud computing platform, etc. In this case, the computing device can refer to a cloud server, and the above-mentioned processing components, storage components, etc. can be basic server resources rented or purchased from the cloud computing platform.

[0505] The present application also provides a computer storage medium storing a computer program, wherein the computer program can achieve the above-mentioned Figure 1 The data labeling method of the illustrated embodiment.

[0506] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0507] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

Claims

1. A method for improving clean energy consumption efficiency through load allocation, characterized in that: The following steps are involved: S1. Collect grid operation data and clean energy power generation data through intelligent monitoring systems; S2. Use data analysis and prediction models to predict grid load and clean energy generation; S3. Dynamically adjust the grid operation parameters and optimize the power load distribution according to the prediction results and the current state of the grid; S4. During peak hours of grid load, reduce load demand through demand response strategies and give priority to the use of clean energy; S5. During the off-peak period of the power grid load, the energy storage system is used to store excess clean energy for use during peak hours; S6. Monitor the grid operation status and clean energy consumption in real time, and adjust the load management strategy based on actual operation data.

2. The method according to claim 1, characterized in that In step S2, the prediction model includes an algorithm of time series analysis, neural network or support vector machine.

3. The method according to claim 1 or 2, characterized in that: The step S3 further comprises: Dynamically adjust the grid operating parameters, and establish charging and discharging models to adjust power generation plans, optimize power flows, and control energy storage systems.

4. The method according to claim 3, characterized in that In step S4, the demand response strategy includes encouraging users to reduce electricity consumption or shift electricity consumption time during peak hours; and the charging and discharging model is defined as: When grid load demand is low and clean energy generation is high, P charge (t)=k1×(P clean (t)-L(t)) Among them, P charge (t) is the clean energy power generation at time t; L(t) is the grid load at time t; k1 is a positive proportional coefficient used to adjust the charging rate and can be set according to actual needs; When the grid load demand is high and the energy storage system has enough energy to release, P discharge (t)=k2×(L(t)-P clean (t)) Among them, k2 is also a positive proportional coefficient used to adjust the discharge rate.

5. The method according to claim 4, characterized in that The step further includes: updating the energy level of the energy storage system according to the following formula: in, E(t) is the energy level of the energy storage system at time t; η charge and η discharge They are charging efficiency and discharging efficiency respectively; Δt is the time step.

6. The method according to claim 5, characterized in that Also includes: Ensure that the energy level E(t) of the energy storage system does not exceed its maximum capacity E max or below the minimum capacity E min Therefore, after each update, perform the following checks: If E(t+1)>E max , then let E(t+1)=E max ; If E(t+1) <E min , then let E(t+1)=E min .

7. A non-volatile storage medium, characterized in that: A program is stored which can be activated to execute the method for improving the clean energy consumption efficiency through intelligent power load management as described in any one of claims 1 to 6.

8. A device for improving clean energy consumption efficiency through load allocation, characterized in that: include: Processor module; A data collection unit, operable to obtain analog signals of actual active output of wind power, actual active power of electric heating load, grid connection point voltage, and grid connection point current; A prediction unit, operable to convert the data obtained by the data acquisition module into parallel data through an analog-to-digital converter and transmit the parallel data to the processor module; An optimization dispatching unit operable to optimize the power load according to the real-time collected power load data, clean energy generation data and energy storage system status; A real-time monitoring module, which can be operated to monitor the power information of the power grid and the operating status information of the energy storage system in real time, and control the charging and discharging status of the energy storage system; The energy storage module can be operated to configure the energy storage system according to the power load status information, improve the grid's ability to accept renewable energy power generation, and promote the consumption of renewable energy.

9. The device for improving clean energy consumption efficiency through load allocation according to claim 8, characterized in that: The data collection unit comprises: A first sliding rheostat is operable to input a wind power actual active output analog signal; A second sliding rheostat is operable to input an actual active power simulation signal of the electric heating load; A third sliding rheostat operable to input a grid connection point voltage analog signal; a fourth sliding rheostat operable to input a grid-connected point current analog signal; Wherein, the prediction unit further includes: an analog-to-digital converter operable to convert the analog signal into a digital signal; a processor module operable to process data from the analog-to-digital converter and perform calculations; a display operable to display the calculation results of the processor module; and the optimization scheduling unit further comprises: a load stability evaluation value calculation unit operable to obtain a regional load stability evaluation value through processing according to the power load at each monitoring time point in the region within the historical time period; A new energy output stability evaluation value calculation unit is operable to analyze and obtain a new energy output stability evaluation value based on the wind power output and photovoltaic output at each monitoring time point in the region within a historical time period; The energy storage configuration demand index calculation unit can be operated to analyze and obtain the regional energy storage configuration demand index according to the regional load stability evaluation value and the new energy output stability evaluation value.

10. The device for improving clean energy consumption efficiency through load allocation according to claim 9, characterized in that: The real-time monitoring module includes: A control time setting unit operable to set a number of control times; An environmental information collection unit, operable to collect environmental information of the new energy station at each control moment; A prediction value calculation unit, operable to obtain a prediction value of the output of the new energy source at each control moment and a load prediction value at each control moment through comprehensive analysis based on a new energy output prediction model; The energy storage configuration state adjustment evaluation value calculation unit can be operated to obtain the power grid power information and the energy storage system operation status information at each control moment, and comprehensively analyze the predicted value of the new energy output and the load predicted value at each control moment to obtain the energy storage configuration state adjustment evaluation value at each control moment.

11. The device for improving clean energy consumption efficiency through load allocation according to claim 10, characterized in that: The energy storage module also includes: An energy storage device screening unit operable to screen energy storage devices corresponding to energy storage configuration types in a statistical area from energy storage devices corresponding to energy storage configuration types stored in a power grid database; an integration unit operable to integrate the energy storage devices of each selected configuration into a regional configuration of the energy storage system; a charge and discharge control unit operable to control the energy storage system to enter a charging state when the energy storage configuration state adjustment evaluation value is lower than the state adjustment evaluation critical value, and to control the energy storage system to enter a discharging state when the energy storage configuration state adjustment evaluation value is higher than the state adjustment evaluation critical value; a dust removal mechanism operable to clean dust inside the storage device; The heat dissipation mechanism can be operated to ensure the normal discharge of heat dissipated by the absorption platform body during high-power operation.

12. The device for improving clean energy consumption efficiency through load allocation according to claim 11, characterized in that: It also includes a memory for storing executable instructions of the processor module; wherein the processor is configured to execute the method for improving the clean energy consumption efficiency through load dispatching as described in any one of claims 1 to 6 by executing the executable instructions.

Images