Energy management method and device based on closed-loop feedback mechanism, equipment and storage medium

Abstract

The invention discloses an energy management method and device based on a closed-loop feedback mechanism, equipment and a storage medium, and relates to the technical field of energy management, and the method comprises the steps: obtaining energy related data, carrying out the feature extraction according to the energy related data, obtaining energy feature information, inputting the energy feature information into a preset energy consumption prediction model, and carrying out the prediction of energy consumption. The method comprises the steps of obtaining an energy consumption prediction result, performing energy distribution on the energy consumption prediction result according to an energy use strategy to obtain an energy distribution result, and optimizing the energy distribution result according to an evaluation optimization strategy to realize energy management based on a closed-loop feedback mechanism. By collecting the energy data, then inputting the energy data into the preset energy consumption prediction model to obtain the prediction result, then formulating the energy use strategy and performing distribution, and finally optimizing the distribution result through a closed-loop feedback mechanism, intelligent energy management is realized, and the efficiency and accuracy of energy use are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of energy management, and particularly to an energy management method, device, equipment and storage medium based on a closed-loop feedback mechanism. Background Art

[0002] Traditional energy management systems mainly rely on manual monitoring and simple automation control. These systems usually include basic data collection, analysis, and simple operations according to preset rules. In modern industrial and commercial environments, there are already some energy management systems based on static models attempting to solve these problems by using historical data to predict future energy consumption trends and formulating corresponding energy management strategies accordingly. However, these systems have limited capabilities in processing large-scale data sets, responding to market changes in real time, and adapting to dynamic changes in energy demand. Therefore, an efficient and intelligent energy optimization management method is needed, which is of great significance for addressing energy challenges and achieving sustainable development goals. Summary of the Invention

[0003] The main purpose of this application is to provide an energy management method, device, equipment and storage medium based on a closed-loop feedback mechanism, aiming to solve the technical problem of how to improve the automation and efficiency of energy management.

[0004] To achieve the above object, this application proposes an energy management method based on a closed-loop feedback mechanism, the method comprising: Performing feature extraction according to the energy-related data to obtain energy feature information; Inputting the energy feature information into a preset energy consumption prediction model to obtain an energy consumption prediction result, the preset energy consumption model being a long short-term memory network model; Performing energy allocation on the energy consumption prediction result according to an energy usage strategy to obtain an energy allocation result; Optimizing the energy allocation result according to an evaluation and optimization strategy to achieve energy management based on a closed-loop feedback mechanism; The step of optimizing the energy allocation result according to an evaluation and optimization strategy to achieve energy management based on a closed-loop feedback mechanism comprises: Performing an environmental impact assessment on the energy allocation result to obtain an environmental impact index value; Performing a cost and efficiency assessment on the energy allocation result to obtain an energy usage cost value and an efficiency index value; Combining the environmental impact index value, the energy usage cost value and the efficiency index value and performing energy allocation according to the target optimization function to obtain an updated energy allocation result; Evaluate according to the updated energy distribution result until the preset energy distribution requirement is met to form a closed-loop feedback mechanism.

[0005] In one embodiment, the step of performing energy distribution on the energy consumption prediction result according to the energy usage strategy to obtain an energy distribution result includes: Judge according to the energy consumption prediction result to obtain the energy demand status; Analyze the energy demand status according to the energy usage strategy to obtain the corresponding energy supply methods for different time periods; Based on the energy supply method, adjust the energy distribution in real time to obtain the energy distribution result.

[0006] In one embodiment, before the step of combining the environmental impact index value, the energy usage cost value, and the efficiency index value and performing energy distribution according to the target optimization function to obtain an updated energy distribution result, it further includes: Construct a cost function, an efficiency function, and an environmental impact function. The cost function is specifically expressed as: Wherein, represents the energy consumption in the represents the real-time energy price in the represents the reward or penalty for demand response in the represents the time period, The efficiency function is specifically expressed as: Wherein, represents the actual output in the The specific expression of the environmental impact function is: Wherein, represents the emission factor in the According to the cost function, the efficiency function, and the environmental impact function, obtain the corresponding energy usage cost value, energy usage cost value, and environmental impact index value. The energy usage cost value, the energy usage cost value, and the environmental impact index value are all constant values; According to the energy usage cost value, energy usage cost value, and environmental impact index value, comprehensively construct a target optimization function, which is specifically expressed as: Wherein, Denote the weights of energy usage cost, energy usage efficiency, and environmental impact indicators. The weight of energy usage cost is dynamically adjusted according to the user's budget and market energy price fluctuations. The weight of energy usage efficiency is set based on the equipment operation efficiency. The weight of environmental impact indicators is determined according to the carbon emission target and environmental protection requirements.

[0007] In one embodiment, the step of combining the environmental impact indicator value, the energy usage cost value, and the efficiency indicator value to perform energy allocation according to the target optimization function and obtain an updated energy allocation result includes: Combining the environmental impact indicator value, the energy usage cost value, and the efficiency indicator value to generate candidate solutions through genetic algorithm operations according to the target optimization function; Performing local optimization on the candidate solutions through particle swarm optimization operations to obtain optimized solutions; Performing global optimization on the optimized solutions through simulated annealing operations to obtain globally optimal solutions; Taking the globally optimal solution as the updated energy allocation result.

[0008] In one embodiment, before the step of inputting the energy characteristic information into a preset energy consumption prediction model to obtain an energy consumption prediction result, it includes: Dividing the energy-related data into a training set, a test set, and a validation set; Obtaining the energy data of the training set and establishing an initial energy consumption prediction model; Optimizing the initial energy consumption prediction model according to the energy data of the training set to obtain a preset energy consumption prediction model.

[0009] In one embodiment, the step of optimizing the initial energy consumption prediction model according to the energy data of the training set to obtain a preset energy consumption prediction model includes: Obtaining the actual energy consumption value; Initializing the model parameters of the initial energy consumption prediction model; Performing normalization processing on the energy data of the training set to obtain processed data; Inputting the processed data into the initial energy consumption prediction model to obtain a predicted energy consumption value; Calculating according to the loss function to obtain the error value between the predicted energy consumption value and the actual energy consumption value; Obtaining the learning rate; Iteratively updating the model parameters according to the learning rate through an optimization algorithm until the maximum number of iterations is reached or the error value calculated by the loss function converges to a preset threshold to obtain a preset energy consumption prediction model.

[0010] In one embodiment, the step of extracting features from the energy-related data to obtain energy feature information includes: Clean the energy-related data to obtain the cleaned data; Process and replace outliers in the cleaned energy-related data to obtain the data after outlier processing; Perform standard normalization based on the data after outlier processing to obtain the normalized data; Perform Fourier transform according to the normalized data to obtain feature information; Screen the feature information through statistical tests to obtain energy feature information.

[0011] In addition, to achieve the above object, the present application also proposes an energy management device based on a closed-loop feedback mechanism. The energy management device based on the closed-loop feedback mechanism includes: An acquisition module for acquiring energy-related data, where the energy-related data includes real-time energy prices, device status information, and user behavior patterns; A feature extraction module for extracting features from the energy-related data to obtain energy feature information; A calculation module for inputting the energy feature information into a preset energy consumption prediction model to obtain an energy consumption prediction result, where the preset energy consumption model is a long short-term memory network model; An allocation module for performing energy allocation on the energy consumption prediction result according to an energy usage strategy to obtain an energy allocation result; An optimization module for optimizing the energy allocation result according to an evaluation and optimization strategy to achieve energy management based on a closed-loop feedback mechanism; and also for performing an environmental impact assessment on the energy allocation result to obtain an environmental impact index value; performing a cost and efficiency assessment on the energy allocation result to obtain an energy usage cost value and an efficiency index value; combining the environmental impact index value, the energy usage cost value, and the efficiency index value and performing energy allocation according to the target optimization function to obtain an updated energy allocation result; and evaluating according to the updated energy allocation result until a preset energy allocation requirement is met to form a closed-loop feedback mechanism.

[0012] In addition, to achieve the above object, the present application also proposes a storage medium. The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the energy management method based on the closed-loop feedback mechanism as described above are implemented.

[0013] In addition, to achieve the above object, the present application also provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, it realizes the steps of the energy management method based on a closed-loop feedback mechanism as described above.

[0014] In the present application, by obtaining energy-related data, extracting features according to the energy-related data to obtain energy feature information, inputting the energy feature information into a preset energy consumption prediction model to obtain an energy consumption prediction result, and performing energy allocation on the energy consumption prediction result according to an energy usage strategy to obtain an energy allocation result, and optimizing the energy allocation result according to an evaluation and optimization strategy to achieve energy management based on a closed-loop feedback mechanism. By collecting data such as real-time energy prices, device status, and user behavior patterns, extracting features and then inputting them into a preset energy consumption prediction model to obtain a prediction result, formulating an energy usage strategy based on this result and performing allocation, and then optimizing the allocation result through a closed-loop feedback mechanism, intelligent energy management is realized, the efficiency and accuracy of energy use are improved, and the adaptability and automation of the system are ensured through continuous optimization to meet the requirements of modern energy management. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0016] Figure 1 It is a schematic flowchart of the first embodiment of the energy management method based on a closed-loop feedback mechanism of the present application; Figure 2 It is a schematic flowchart of the second embodiment of the energy management method based on a closed-loop feedback mechanism of the present application; Figure 3 It is a schematic flowchart of the third embodiment of the energy management method based on a closed-loop feedback mechanism of the present application; Figure 4 It is a schematic module structure diagram of the energy management device based on a closed-loop feedback mechanism in an embodiment of the present application; Figure 5 It is a schematic device structure diagram of the hardware operating environment involved in the energy management method based on a closed-loop feedback mechanism in an embodiment of the present application.

[0017] The implementation, functional features, and advantages of the object of the present application will be further described with reference to the embodiments and the drawings. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0019] To better understand the technical solution of this application, the following will be described in detail in conjunction with the accompanying drawings of the specification and specific embodiments.

[0020] The main solution of the embodiment of this application is: by obtaining energy-related data, extracting features based on the energy-related data to obtain energy feature information, inputting the energy feature information into a preset energy consumption prediction model to obtain an energy consumption prediction result, performing energy allocation on the energy consumption prediction result according to an energy usage strategy to obtain an energy allocation result, and optimizing the energy allocation result according to an evaluation and optimization strategy to achieve energy management based on a closed-loop feedback mechanism.

[0021] Based on this, the embodiment of this application provides an energy management method based on a closed-loop feedback mechanism. Refer to Figure 1 , Figure 1 which is a schematic flowchart of the first embodiment of the energy management method based on a closed-loop feedback mechanism of this application.

[0022] In this embodiment, the energy management method based on a closed-loop feedback mechanism includes steps S10 to S50: Step S10, obtain energy-related data.

[0023] It should be noted that the above energy-related data includes real-time energy prices, device status information, and user behavior patterns. Specifically, the data on real-time energy prices usually comes from API interfaces provided by the electricity market or energy suppliers. It contains electricity price information for different time periods, which is crucial for formulating time-of-use electricity strategies. For example, storing electrical energy during the night when the electricity price is low and using the stored energy during the day when the electricity price is high can effectively reduce costs. The device status information involves the status monitoring of all devices related to energy consumption, such as industrial equipment, air conditioning systems, lighting facilities, etc. Sensors are installed on these devices to monitor their operating status in real time, including power consumption, operating efficiency, fault warnings, etc. By analyzing this data, inefficient or faulty devices can be identified and corresponding adjustments or repairs can be made to avoid unnecessary energy waste. The user behavior patterns include information such as working hours, rest hours, peak usage periods, etc. With the help of smart meters and other smart home devices, the system can track the user's behavior patterns, such as when electrical appliances are turned on or off.

[0024] Step S20, extract features based on the energy-related data to obtain energy feature information.

[0025] It should be noted that, based on the energy-related data, preprocessing is carried out, and then the feature information useful for the prediction model is extracted. Then, further refinement is selected for the feature information. Through statistical tests, model-based methods, or recursive feature elimination techniques, the features most influential for the prediction task can be selected while removing redundant or irrelevant features. For example, using the Pearson correlation coefficient to screen out the features highly correlated with the target variable (i.e., energy consumption), and only retaining those features with a correlation coefficient greater than the set threshold (such as ±0.5).

[0026] Furthermore, step S20 includes: performing data cleaning on the energy-related data to obtain the cleaned data, then performing outlier processing and replacement on the cleaned energy-related data to obtain the data after outlier processing, and then performing standard normalization based on the data after outlier processing to obtain the normalized data. Specifically, the purpose of data cleaning is to remove noise and irrelevant information. This includes deleting duplicate records, correcting format errors, and filling in missing values. For example, for continuous numerical features such as energy consumption data, the forward filling strategy is used to fill in the missing values; for categorical features, the mode is used for filling. Then, outliers are identified by calculating the mean and standard deviation, and any data points exceeding the mean ± 3 times the standard deviation will be regarded as outliers and replaced with the median of the feature. After completing data cleaning, enter the outlier processing stage. The system will replace or adjust the detected outliers according to preset rules to ensure the consistency and reliability of the dataset. For example, if the real-time energy price at a certain time period significantly deviates from the normal range, the system will reasonably replace the outlier based on historical data or data at adjacent time points. The last step is standard normalization, that is, converting data of different magnitudes into a unified scale. Common methods include min-max normalization and Z-score standardization. For example, using the min-max normalization method, all features are scaled between 0 and 1. The above processing is to eliminate the influence caused by unit differences, so that data from different sources can be compared and analyzed on the same scale, thereby improving the performance and accuracy of the prediction model.

[0027] Furthermore, perform a Fourier transform on the normalized data to obtain feature information, and screen the feature information through statistical tests to obtain energy feature information. Specifically, the Fourier transform is a method that converts a time-domain signal into a frequency-domain representation, which can help us identify periodic patterns in the data. For the normalized energy consumption data, by applying the Fast Fourier Transform (FFT), we can convert the original time-series data into its frequency components. This method is particularly suitable for detecting and analyzing periodic trends hidden behind complex data, such as daily or seasonal energy consumption fluctuations. For example, when processing the energy usage data of a building, the Fourier transform can reveal the energy consumption difference patterns between weekdays and weekends, or between different seasons. After obtaining the feature information in the frequency domain, the next step is the feature screening stage. In this process, statistical tests are usually used to evaluate the importance of each feature and select the features that are most helpful for improving the prediction accuracy of the model. A common method is to calculate the Pearson correlation coefficient between the feature and the target variable (such as energy consumption). By setting a threshold (e.g., ±0.5), only those features with a high positive or negative correlation are retained. In addition, model-based methods such as Recursive Feature Elimination (RFE) can be used. By repeatedly building models and removing the features with the lowest contribution until the best feature subset is found. In addition to the above methods, other advanced techniques can also be considered for feature selection, such as L1 regularization (Lasso regression), which can force the weights of some unimportant features to become zero by increasing the penalty term, thus achieving an automated feature selection process. The energy feature information screened out through these steps not only contains the most critical periodic and aperiodic changes in the data, but also removes noise and redundant information, making the final dataset used to train the prediction model more refined and efficient.

[0028] Step S30: Input the energy feature information into a preset energy consumption prediction model to obtain an energy consumption prediction result.

[0029] It should be noted that in this embodiment, the preset energy consumption model is a long short-term memory network model because it is good at capturing long-term dependencies and other complex patterns in time-series data.

[0030] Specifically, a three-layer LSTM structure is adopted, with each layer containing 50 neurons, which can provide sufficient capacity to learn complex energy consumption patterns. During the training process, the Adam optimizer is used, with a learning rate set to 0.001, a batch size of 32, and the mean squared error (MSE) is selected as the loss function. Hyperparameters are adjusted through grid search or random search to ensure that the model can achieve optimal performance. Then, the energy feature information is input into the trained model. These feature information include but are not limited to timestamps, seasonal factors, weather conditions, user behavior patterns, etc. When new data enters the system, it will be used together with historical data for real-time prediction of energy demand in the next few hours or even days. For example, if historical data shows that the energy consumption surges due to the use of office equipment between 9 am and 5 pm on weekdays, then the model can identify this pattern and make accurate predictions under similar conditions. Finally, the prediction results are generated and applied to actual operations. Using the trained LSTM model combined with the latest input data for prediction, the energy demand estimate for a specific future time period can be obtained. The prediction results not only provide a decision-making basis for intelligent scheduling but also help formulate more flexible cost-saving strategies. For example, when the prediction shows that the energy consumption will increase in the next few hours and the market electricity price is low, it will decide to purchase additional electricity to fill the energy storage device; on the contrary, if the electricity price is high, non-critical industrial processes will be postponed to reduce costs.

[0031] Step S40, according to the energy usage strategy, perform energy allocation on the energy consumption prediction result to obtain the energy allocation result.

[0032] It should be noted that according to the predicted energy demand, combined with the real-time energy price, the operating status of the equipment, and the charge and discharge status of the energy storage device, the energy allocation plan is dynamically adjusted. For example, during peak energy demand periods, the energy storage device will be preferentially scheduled to release energy, or low-cost electricity will be purchased from the smart grid; while during low-demand periods, the energy storage device will be arranged to charge, or unnecessary energy procurement will be reduced.

[0033] Furthermore, perform specific energy distribution operations according to the generated strategy above to obtain the energy distribution result. It automatically adjusts the energy supply and distribution plan through interfaces with the power grid and other distributed energy resources. For example, if the prediction indicates an upcoming high-energy consumption period with expensive electricity prices, it schedules the battery energy storage system to store the electricity when the price is low to meet the demand. At the same time, it can also communicate with the smart grid, request more electricity from the grid, or sell the excess electricity to the grid when the demand is low, so as to achieve the balance between supply and demand while maximizing economic benefits. In addition, to ensure user participation and support, an intuitive operation interface is provided, allowing users to view the prediction data, understand the current energy usage situation and optimization suggestions, and make corresponding adjustments according to personal preferences. For example, during peak power demand periods, users can choose to turn off some non-essential electrical appliances, or heat water when the electricity is sufficient and the cost is low to respond to the system's energy-saving suggestions.

[0034] In this way, the system can ensure the efficient matching of energy supply and demand, while optimizing the energy usage cost and environmental impact. The final energy distribution result not only meets the users' energy needs, but also realizes the intelligence and sustainability of energy management.

[0035] Step S50: Optimize the energy distribution result according to the evaluation and optimization strategy to achieve energy management based on the closed-loop feedback mechanism.

[0036] It should be noted that the evaluation and optimization strategy usually relies on multi-objective optimization algorithms, aiming to balance cost, efficiency, and environmental impact. For example, if it is predicted that the energy consumption will increase in the next few hours and the market electricity price is low, it will recommend purchasing additional electricity to fill the energy storage device; on the contrary, when the electricity price is high, non-critical industrial processes will be postponed to reduce costs. Then, the closed-loop feedback mechanism enables the system to continuously learn and improve in actual operation. This includes real-time monitoring of the energy consumption situation and changes in external variables (such as real-time energy prices, weather changes, etc.), and feeding back this information. For example, suppose it was originally planned to charge the battery during the low electricity price period, but suddenly the weather forecast shows that the upcoming storm may cause grid instability. At this time, it should be able to automatically adjust its strategy, complete the charging in advance or find alternative solutions to ensure the safety and stability of energy supply.

[0037] Further, step S50 further includes: performing an environmental impact assessment on the energy distribution result to obtain an environmental impact index value; performing a cost and efficiency assessment on the energy distribution result to obtain an energy usage cost value and an efficiency index value. Combining the environmental impact index value, the energy usage cost value, and the efficiency index value and performing energy distribution according to the target optimization function to obtain an updated energy distribution result. Before combining the environmental impact index value, the energy usage cost value, and the efficiency index value and performing energy distribution according to the target optimization function to obtain an updated energy distribution result, it is necessary to construct a cost function, an efficiency function, and an environmental impact function. The above cost function is specifically expressed as: Wherein, represents the energy consumption in the time period, represents the real-time energy price in the time period, represents the time period. In this embodiment, it is assumed that the energy price in the time period is 0.15 per kilowatt-hour, and the electricity consumption in the same time period is 1000 kilowatt-hours. Sometimes, in order to encourage users to reduce electricity consumption during high-demand periods, the power grid company will provide a demand response program. If users participate and successfully reduce the expected peak load, they will receive a reward. If they receive a demand response reward of 50, then the cost in the

[0038] time period is The above efficiency function is specifically expressed as: represents the actual output in the time period. If a factory's output in the time period is 500 units of products and it consumes 2000 kilowatt-hours, then

[0039] The specific expression of the above environmental impact function is: Wherein, represents the emission factor in the time period. If is 0.7 kilograms of carbon dioxide per kilowatt-hour and kg. According to the above cost function, efficiency function and environmental impact function, the corresponding energy usage cost value, energy usage cost value and environmental impact index value are obtained. The above energy usage cost value, the above energy usage cost value and the above environmental impact index value are all constant values. Then, based on the energy usage cost value, energy usage cost value and environmental impact index value, an objective optimization function is comprehensively constructed, which is specifically expressed as: where, represents the energy usage cost weight, energy usage efficiency weight and environmental impact index weight. The above energy usage cost weight is dynamically adjusted according to the user's budget and market energy price fluctuations. Specifically, by real-time monitoring of the energy market price and combining with the user's budget limit, the cost weight is dynamically calculated. For example, when the energy price rises significantly, the system will increase the cost weight to prompt the optimization algorithm to preferentially select energy usage strategies with lower costs; conversely, when the energy price is low, the cost weight can be appropriately reduced to balance other optimization goals. The above energy usage efficiency weight is set based on the equipment operation efficiency. Specifically, by analyzing the historical operation data and performance indicators of the equipment, the efficiency of the equipment in different operation states is evaluated. For equipment with high efficiency operation, the system will assign a higher efficiency weight to encourage the optimization algorithm to preferentially schedule these equipment to improve the overall energy utilization efficiency. In addition, the system can further adjust the efficiency weight according to the user's production goals (such as maximizing production or minimizing production time) to meet specific production requirements. The above environmental impact index weight is determined based on the carbon emission target and environmental protection requirements. Specifically, according to the user's carbon emission target (such as reducing carbon emissions) and relevant environmental protection policy requirements, the environmental impact weight is set. For example, if the user sets a strict carbon emission target, the environmental impact weight will be increased to prompt the optimization algorithm to preferentially select low-carbon or carbon-free energy usage strategies. At the same time, the regulations on carbon emission limits will also be considered to ensure that the optimization results meet the environmental protection requirements.

[0040] Furthermore, the environmental impact index value, energy usage cost value, and efficiency index value are combined to perform energy allocation according to the objective optimization function, obtaining an updated energy allocation result, including: combining the environmental impact index value, energy usage cost value, and efficiency index value to generate candidate solutions through genetic algorithm operations according to the objective optimization function; locally optimizing the candidate solutions through particle swarm optimization operations to obtain optimized solutions; globally optimizing the optimized solutions through simulated annealing operations to obtain globally optimal solutions; and using the globally optimal solutions as the updated energy allocation result. Specifically, based on the above objective optimization function, a genetic algorithm (GA) is used to generate initial candidate solutions. In the genetic algorithm, each chromosome represents a possible energy usage strategy, and the gene loci represent the on / off states of devices at different time periods. Through roulette wheel selection, single-point crossover (crossover probability 0.7), and random mutation (mutation probability 0.01), the algorithm can quickly explore the solution space and generate diverse candidate solutions. Subsequently, these candidate solutions are input into a particle swarm optimization (PSO) algorithm for local optimization. In PSO, each particle represents a potential energy management strategy, and the velocity and position of the particle are updated through the individual best position and the swarm best position, with the learning factors both set to 2.0 and the maximum number of iterations set to 100 times. PSO can quickly converge to a local optimal solution, further improving the quality of the candidate solutions. The obtained locally optimized solutions are then globally optimized through a simulated annealing (SA) algorithm. The initial temperature of the SA algorithm is set to 10,000, and the cooling rate is 0.95. The algorithm stops when the temperature drops to 0.01 or the number of iterations reaches 1,000 times. Through the random search and temperature annealing mechanism of simulated annealing, the algorithm can explore the optimal solution globally and effectively avoid local optimal traps. Finally, the globally optimal solution optimized by simulated annealing is used as the updated energy allocation result, ensuring that the energy allocation strategy can adapt to changing market conditions and user needs while meeting economic efficiency and environmental friendliness, realizing the intelligent and sustainable development of the energy management system.

[0041] Evaluate according to the updated energy distribution results until the preset energy distribution requirements are met to form a closed-loop feedback mechanism. Specifically, after obtaining the global optimal solution through genetic algorithms, particle swarm optimization, and simulated annealing algorithms, use this solution as the updated energy distribution result and apply it in practice. Subsequently, conduct real-time evaluation on the updated energy distribution results. The evaluation indicators include energy usage cost, efficiency indicators, and environmental impact indicators, etc., to ensure that the energy distribution strategy achieves the expected optimization goals during actual operation. If the evaluation results show that the energy distribution results do not fully meet the preset energy distribution requirements, such as the energy cost exceeding the budget, the equipment operation efficiency not meeting the standard, or the environmental impact exceeding the allowable range, obtain these evaluation results. Readjust the optimization strategy according to the feedback information, generate a new energy distribution strategy again, and apply it. This process loops continuously until the energy distribution results meet all preset requirements. Through this closed-loop feedback mechanism, it can dynamically adapt to the fluctuations of the energy market, the changes in equipment status, and the adjustment of user needs, ensuring that the energy distribution strategy always remains in an optimal state.

[0042] This embodiment provides an energy management method based on a closed-loop feedback mechanism. By obtaining energy-related data, extracting features according to the energy-related data to obtain energy feature information, inputting the energy feature information into a preset energy consumption prediction model to obtain an energy consumption prediction result, and performing energy distribution on the energy consumption prediction result according to an energy usage strategy to obtain an energy distribution result, and optimizing the energy distribution result according to an evaluation and optimization strategy to achieve energy management based on a closed-loop feedback mechanism. By collecting data such as real-time energy prices, equipment status, and user behavior patterns, inputting them into a preset energy consumption prediction model after feature extraction to obtain a prediction result, formulating an energy usage strategy and performing distribution based on this result, and then optimizing the distribution result through a closed-loop feedback mechanism to achieve intelligent energy management, improve the efficiency and accuracy of energy usage, and ensure the automatic adaptability of the system through continuous optimization to meet the requirements of modern energy management.

[0043] Based on the first embodiment of the present application, in the second embodiment of the present application, the content that is the same as or similar to the above-mentioned embodiment one can be referred to the above introduction and will not be repeated hereinafter. On this basis, please refer to Figure 2 , step S40 of the energy management method based on a closed-loop feedback mechanism further includes steps S201 to S203: Step S201, make a judgment according to the energy consumption prediction result to obtain the energy demand status.

[0044] It should be noted that the energy demand status refers to the amount of energy demanded within a specific time period (such as per hour, per day, or per week). Based on the results of energy consumption prediction, analysis and judgment are carried out to understand the energy demand trend in the future. For example, if the energy consumption prediction results show that the electricity demand surges due to the use of office equipment between 9 am and 5 pm on a working day, while the demand drops significantly at night, this period can be defined as the "high demand state". On the contrary, on weekends or holidays when the overall demand level is low, it can be regarded as the "low demand state".

[0045] Step S202: Analyze the energy demand status according to the energy usage strategy to obtain the corresponding energy supply methods for different time periods.

[0046] It should be noted that by obtaining the energy demand status in real time, including the current energy consumption, equipment operation status, user behavior patterns, and external environmental factors (such as meteorological conditions), etc. These data are transmitted to the data processing unit in real time for analysis and processing to ensure the accuracy and availability of the data. Subsequently, according to the energy usage strategy, the energy demand status is processed. The energy usage strategy is generated through an adaptive optimization algorithm, which comprehensively considers multiple factors such as cost, efficiency, and environmental impact. The strategy contains information such as the on / off status of equipment, energy procurement plans, and charge / discharge arrangements of energy storage devices in different time periods. According to these strategies, the energy distribution plan is dynamically adjusted to calculate the corresponding energy supply methods for different time periods. For example, during the high demand period from 9 am to 5 pm on a working day, if real-time monitoring shows that the grid load is approaching the upper limit, energy storage devices will be dispatched to release energy to relieve the grid pressure; while during the low demand period at night, energy storage devices will be arranged to charge, or unnecessary energy procurement will be reduced to lower the operating cost. In addition, when the real-time energy price is detected to be lower than expected, the procurement volume may be temporarily increased to prepare for future demand peaks.

[0047] Step S203: Based on the energy supply method, adjust the energy distribution in real time to obtain the energy distribution result.

[0048] It should be noted that based on the energy supply method, the energy distribution plan will be dynamically adjusted, including adjusting the charge / discharge strategy of energy storage devices, rearranging the work schedules of equipment, or interacting with the smart grid to buy or sell excess electricity. For example, during the low electricity consumption period (such as late at night), low-cost electricity can be used to charge energy storage devices for use during peak periods.

[0049] In addition, fine-tuning will be carried out in combination with external environmental factors and user behavior patterns. For example, if the weather forecast shows that an upcoming storm may cause a decrease in photovoltaic power generation, the proportion of electricity purchased from the grid will be increased accordingly to ensure the continuity and stability of energy supply.

[0050] In this embodiment, the demand status is determined through energy consumption prediction, the supply quantity is processed in combination with strategies, and the allocation is adjusted in real time to achieve precise energy management and improve energy use efficiency.

[0051] Based on the first embodiment of this application, in the third embodiment of this application, the same or similar content as that in the above-mentioned first embodiment can be referred to the above introduction and will not be repeated hereinafter. On this basis, please refer to Figure 3 , before step S30 of the energy management method based on the closed-loop feedback mechanism, steps S301 to S303 are further included: Step S301: Divide the energy-related data into a training set, a test set, and a validation set.

[0052] It should be noted that, in order to improve the accuracy and reliability of the energy consumption prediction model, the collected energy-related data is divided into a training set, a test set, and a validation set. Historical energy consumption data, real-time energy prices, meteorological conditions, equipment status, and user behavior patterns are obtained from multiple data sources (such as smart meters, weather stations, user equipment, etc.). After being cleaned, standardized, and feature-extracted by the data processing unit, these data are divided into three parts: the training set accounts for 70% of the total data and is used to train the prediction model. Through a large amount of historical data, the model can learn the patterns and trends of energy consumption, thereby improving the accuracy of prediction. The validation set accounts for 15% of the total data and is used to adjust the hyperparameters of the model and evaluate the performance of the model. Through the validation set, the structure and parameters of the model can be optimized to ensure that the model has good generalization ability on unseen data. The test set accounts for 15% of the total data and is used to finally evaluate the performance of the model. The data in the test set is not used during the model training and validation process, so the actual prediction ability of the model can be evaluated. Through this division method, the system can make full use of data resources and improve the accuracy and reliability of the model.

[0053] Step S302: Obtain the energy data of the training set and establish an initial energy consumption prediction model.

[0054] It should be noted that, in this embodiment, a long short-term memory network (LSTM) is used to establish an initial energy consumption prediction model. The LSTM model can effectively process time series data and capture the long-term dependence relationship and dynamic change trend of energy consumption. During the model construction process, by defining the network architecture, setting hyperparameters (such as learning rate, batch size, number of layers, etc.), and selecting an appropriate loss function (such as mean square error, MSE), it is ensured that the model has good learning ability and generalization performance. Specifically, in order to optimize the model parameters, the Adam optimizer can be used, and an appropriate initial learning rate (such as 0.001) can be set, and at the same time, the mean square error (MSE) is selected as the loss function to measure the gap between the predicted value and the actual value.

[0055] Step S303: Optimize the initial energy consumption prediction model according to the energy data in the training set to obtain a preset energy consumption prediction model.

[0056] It should be noted that for the training of the initial energy consumption prediction model, first, obtain the actual energy consumption value. Specifically, it is necessary to collect the actual energy consumption values in the historical data as a reference standard. These data usually come from devices such as smart meters and sensors, which record the real energy consumption at different time points. Then initialize the model parameters of the initial energy consumption prediction model, and perform normalization processing on the energy data in the training set to obtain the processed data. Specifically, initialize the LSTM weights and bias terms, and then perform normalization processing on the input features. Common methods include min-max normalization and Z-score standardization. For example, scale all energy consumption data to between 0 and 1. Then input the processed data into the initial energy consumption prediction model to obtain the predicted energy consumption value. Calculate according to the loss function to obtain the error value between the predicted energy consumption value and the actual energy consumption value. It should be understood that the commonly used loss function is the mean square error (MSE), which measures the gap between the model prediction value and the actual value. A smaller MSE means more accurate model prediction. Obtain the learning rate, and iteratively update the model parameters through the optimization algorithm according to the learning rate until the maximum number of iterations is reached or the error value calculated by the loss function converges to a preset threshold to obtain a preset energy consumption prediction model. Specifically, choosing an appropriate learning rate (such as 0.001) is crucial because it determines the amplitude of parameter update in each iteration. Too high a learning rate may cause the model not to converge, while too low a learning rate will make the training process extremely slow. Use the Adam optimizer, combined with the backpropagation algorithm, to gradually adjust the model parameters according to the calculated error until the predetermined maximum number of iterations is reached or the error value drops below the preset threshold. Once the model training is completed, that is, the optimal state is reached, a preset energy consumption prediction model is obtained, providing a reliable basis for subsequent energy management decisions.

[0057] After obtaining the preset energy consumption prediction model, it further includes: evaluating the preset energy consumption prediction model to obtain an evaluation result; if the evaluation result does not meet the detection requirements, retrain the preset energy consumption prediction model until the detection requirements are met.

[0058] This embodiment realizes precise energy management, improves prediction accuracy, reduces costs, enhances energy use efficiency and system flexibility by dividing the data set, establishing and optimizing the prediction model.

[0059] This application also provides an energy management device based on a closed-loop feedback mechanism. Please refer to Figure 4 The device includes: An acquisition module 10 for acquiring energy-related data, where the energy-related data includes real-time energy prices, device status information, and user behavior patterns.

[0060] A feature extraction module 20 for extracting features based on the energy-related data to obtain energy feature information.

[0061] A calculation module 30 for inputting the energy feature information into a preset energy consumption prediction model to obtain an energy consumption prediction result, where the preset energy consumption model is a long short-term memory network model.

[0062] An allocation module 40 for allocating energy based on the energy consumption prediction result according to an energy usage strategy to obtain an energy allocation result.

[0063] An optimization module 50 for optimizing the energy allocation result according to an evaluation and optimization strategy to achieve energy management based on a closed-loop feedback mechanism; also for performing an environmental impact assessment on the energy allocation result to obtain an environmental impact index value; performing a cost and efficiency assessment on the energy allocation result to obtain an energy usage cost value and an efficiency index value; combining the environmental impact index value, the energy usage cost value, and the efficiency index value and performing energy allocation according to an objective optimization function to obtain an updated energy allocation result; and performing an evaluation based on the updated energy allocation result until a preset energy allocation requirement is met to form a closed-loop feedback mechanism.

[0064] The energy management device based on a closed-loop feedback mechanism provided by the present application adopts the energy management method based on a closed-loop feedback mechanism in the above embodiment, and can solve the technical problem of how to improve the automation and efficiency of energy management. Compared with the prior art, the beneficial effects of the energy management device based on a closed-loop feedback mechanism provided by the present application are the same as those of the energy management method based on a closed-loop feedback mechanism provided by the above embodiment, and other technical features in the energy management device based on a closed-loop feedback mechanism are the same as the features disclosed in the method of the above embodiment, and will not be elaborated here.

[0065] In one embodiment, the allocation module 40 is further configured to make a judgment based on the energy consumption prediction result to obtain an energy demand status; analyze the energy demand status according to the energy usage strategy to obtain corresponding energy supply methods for different time periods; and adjust the energy allocation in real time based on the energy supply methods to obtain an energy allocation result.

[0066] In one embodiment, the optimization module 50 is further configured to construct a cost function, an efficiency function, and an environmental impact function, and obtain corresponding energy usage cost values, energy usage cost values, and environmental impact index values according to the cost function, the efficiency function, and the environmental impact function. The energy usage cost values, the energy usage cost values, and the environmental impact index values are all constant values. A target optimization function is comprehensively constructed based on the energy usage cost values, the energy usage cost values, and the environmental impact index values.

[0067] In one embodiment, the optimization module 50 is further configured to combine the environmental impact index value, the energy usage cost value, and the efficiency index value, generate candidate solutions through genetic algorithm operations according to the target optimization function, perform local optimization on the candidate solutions through particle swarm optimization operations to obtain optimized solutions, perform global optimization on the optimized solutions through simulated annealing operations to obtain global optimal solutions, and use the global optimal solutions as the updated energy allocation results.

[0068] In one embodiment, the calculation module 30 is further configured to divide the energy-related data into a training set, a test set, and a validation set, obtain the energy data of the training set, and establish an initial energy consumption prediction model. The initial energy consumption prediction model is optimized according to the energy data of the training set to obtain a preset energy consumption prediction model.

[0069] In one embodiment, the calculation module 30 is further configured to obtain the actual energy consumption value, initialize the model parameters of the initial energy consumption prediction model, perform normalization processing on the energy data of the training set to obtain processed data, input the processed data into the initial energy consumption prediction model to obtain a predicted energy consumption value, calculate the error value between the predicted energy consumption value and the actual energy consumption value according to the loss function, obtain the learning rate, and iteratively update the model parameters through an optimization algorithm according to the learning rate until the maximum number of iterations is reached or the error value calculated by the loss function converges to a preset threshold to obtain a preset energy consumption prediction model.

[0070] In one embodiment, the feature extraction module 20 is configured to extract energy feature information according to the energy-related data, and is further configured to perform data cleaning on the energy-related data to obtain cleaned data, perform outlier processing and replacement on the cleaned energy-related data to obtain data after outlier processing, perform standard normalization on the data after outlier processing to obtain normalized data, perform Fourier transform on the normalized data to obtain feature information, and screen the feature information through statistical tests to obtain energy feature information.

[0071] The present application provides an energy management device based on a closed-loop feedback mechanism. The energy management device based on the closed-loop feedback mechanism includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the energy management method based on the closed-loop feedback mechanism in Embodiment 1 above.

[0072] Reference is made below Figure 5 , which shows a schematic structural diagram of an energy management device based on a closed-loop feedback mechanism suitable for implementing the embodiments of the present application. The energy management device based on the closed-loop feedback mechanism in the embodiments of the present application may include, but is not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions: tablet computers), PMPs (Portable Media Players), vehicle terminals (such as vehicle navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 5 The energy management device based on the closed-loop feedback mechanism shown is only an example and should not impose any limitations on the functions and usage scope of the embodiments of the present application.

[0073] As Figure 5As described above, the energy management device based on the closed-loop feedback mechanism may include a processing device 1001 (such as a central processing unit, a graphics processing unit, etc.), which may perform various appropriate actions and processes according to a program stored in a ROM (Read Only Memory) 1002 or a program loaded from a storage device 1003 into a RAM (Random Access Memory) 1004. In the RAM 1004, various programs and data required for the operation of the energy management device based on the closed-loop feedback mechanism are also stored. The processing device 1001, the ROM 1002, and the RAM 1004 are connected to each other via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems may be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touchpad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 may allow the energy management device based on the closed-loop feedback mechanism to communicate with other devices wirelessly or wiredly to exchange data. Although the figure shows an energy management device based on the closed-loop feedback mechanism having various systems, it should be understood that it is not required to implement or have all the shown systems. More or fewer systems may be implemented or had alternatively.

[0074] In particular, according to the embodiments disclosed in the present application, the processes described above with reference to the flowcharts may be implemented as computer software programs. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable storage medium, and the computer program includes program codes for performing the methods described in the flowcharts. In such an embodiment, the computer program may be downloaded and installed from a network through the communication device, or installed from the storage device 1003, or installed from the ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the methods of the embodiments disclosed in the present application are executed.

[0075] The energy management device based on the closed-loop feedback mechanism provided in the present application adopts the energy management method based on the closed-loop feedback mechanism in the above-mentioned embodiment, and can solve the technical problem of how to improve the automation and efficiency of energy management. Compared with the prior art, the beneficial effects of the energy management device based on the closed-loop feedback mechanism provided in the present application are the same as those of the energy management method based on the closed-loop feedback mechanism provided in the above-mentioned embodiment, and other technical features in the energy management device based on the closed-loop feedback mechanism are the same as the features disclosed in the method of the previous embodiment, and will not be elaborated here.

[0076] It should be understood that each part disclosed in this application can be implemented by hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in a suitable manner in any one or more embodiments or examples.

[0077] As described above, it is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art within the technical scope disclosed in this application can easily think of changes or substitutions, which should all be covered within the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

[0078] This application provides a computer-readable storage medium, having computer-readable program instructions (i.e., computer programs) stored thereon for performing calculations to obtain machine-readable program instructions for executing the energy management method based on a closed-loop feedback mechanism in the above embodiments.

[0079] The computer-readable storage medium provided by this application can be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination of the above. More specific examples of computer-readable storage media can include, but are not limited to: electrical connections with one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM) or flash memory, optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the above. In this embodiment, for performing calculations, the machine-readable storage medium can be any tangible storage medium that contains or stores a program, and this program can be used by or combined with an instruction execution system, device, or device. The program code contained on the computer-readable storage medium can be transmitted by any appropriate storage medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination of the above.

[0080] The above computer-readable storage medium can be included in the energy management device based on a closed-loop feedback mechanism; it can also exist alone without being assembled into the energy management device based on a closed-loop feedback mechanism.

[0081] The above computer-readable storage medium carries one or more programs. When the one or more programs are executed by an energy management device based on a closed-loop feedback mechanism, the energy management device based on the closed-loop feedback mechanism can write computer program code for performing the operations of this application in one or more programming languages or combinations thereof. The programming languages include object-oriented programming languages such as Java, Smalltalk, C++, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, executed as an independent software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (for example, by using an Internet service provider to connect through the Internet).

[0082] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code that contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than that marked in the accompanying drawings. For example, two consecutively represented blocks may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and the combination of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system for performing the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0083] The modules involved in the embodiments described in this application can be implemented in software or in hardware. Among them, the name of the module does not constitute a limitation to the unit itself in some cases.

[0084] The readable storage medium provided by this application is a computer-readable storage medium. The computer-readable storage medium stores computer-readable program instructions (i.e., computer programs) for executing the above energy management method based on a closed-loop feedback mechanism, and can solve the technical problem of how to improve the automation and efficiency of energy management. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided by this application are the same as those of the energy management method based on a closed-loop feedback mechanism provided by the above embodiments, and will not be elaborated here.

[0085] This application also provides a computer program product, including a computer program. When the computer program is executed by a processor, it implements the steps of the energy management method based on a closed-loop feedback mechanism as described above.

[0086] The computer program product provided by this application can solve the technical problem of how to improve the automation and efficiency of energy management. Compared with the prior art, the beneficial effects of the computer program product provided by this application are the same as those of the energy management method based on a closed-loop feedback mechanism provided by the above embodiments, and will not be elaborated here.

[0087] The above are only some embodiments of this application, and do not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application by using the content of the specification and drawings of this application, or directly / indirectly applied in other related technical fields, are included in the patent protection scope of this application.

Claims

1. An energy management method based on a closed-loop feedback mechanism, characterized in that The method includes: Obtaining energy-related data, where the energy-related data includes real-time energy prices, equipment status information, and user behavior patterns; Performing feature extraction based on the energy-related data to obtain energy feature information; Inputting the energy feature information into a preset energy consumption prediction model to obtain an energy consumption prediction result, where the preset energy consumption model is a long short-term memory network model; Performing energy allocation on the energy consumption prediction result according to an energy usage strategy to obtain an energy allocation result; Optimizing the energy allocation result according to an evaluation and optimization strategy to achieve energy management based on a closed-loop feedback mechanism; The step of optimizing the energy allocation result according to an evaluation and optimization strategy to achieve energy management based on a closed-loop feedback mechanism includes: Performing an environmental impact assessment on the energy allocation result to obtain an environmental impact index value; Performing a cost and efficiency assessment on the energy allocation result to obtain an energy usage cost value and an efficiency index value; Combining the environmental impact index value, the energy usage cost value, and the efficiency index value and performing energy allocation according to the target optimization function to obtain an updated energy allocation result; Evaluating according to the updated energy allocation result until a preset energy allocation requirement is met to form a closed-loop feedback mechanism.

2. The method according to claim 1, wherein The step of performing energy allocation on the energy consumption prediction result according to an energy usage strategy to obtain an energy allocation result includes: Making a judgment based on the energy consumption prediction result to obtain an energy demand status; Analyzing the energy demand status according to an energy usage strategy to obtain corresponding energy supply methods for different time periods; Adjusting the energy allocation in real time based on the energy supply methods to obtain an energy allocation result.

3. The method according to claim 1, wherein Before the step of combining the environmental impact index value, the energy usage cost value, and the efficiency index value and performing energy allocation according to the target optimization function to obtain an updated energy allocation result, it further includes: Constructing a cost function, an efficiency function, and an environmental impact function, The specific expression of the cost function is: Among them, represents the energy consumption during the time period, represents the real-time energy price during the time period, represents the reward or penalty for demand response during the time period, represents the time period, The specific expression of the efficiency function is: Among them, represents the actual output during the time period The specific expression of the environmental impact function is: Among them, represents the emission factor for the time period; According to the cost function, the efficiency function, and the environmental impact function, obtaining corresponding energy usage cost values, energy usage efficiency values, and environmental impact index values, where the energy usage cost values, the energy usage cost values, and the environmental impact index values are all constant values; Comprehensively constructing a target optimization function according to the energy usage cost values, energy usage efficiency values, and environmental impact index values, and the specific expression is: Among them, represents the weight of energy usage cost, the weight of energy usage efficiency, and the weight of environmental impact indicators. The weight of energy usage cost is dynamically adjusted according to the user's budget and market energy price fluctuations. The weight of energy usage efficiency is set based on the equipment operation efficiency. The weight of environmental impact indicators is determined according to the carbon emission target and environmental protection requirements.

4. The method according to claim 1, wherein The step of combining the environmental impact index value, the energy usage cost value, and the efficiency index value and performing energy allocation according to the target optimization function to obtain an updated energy allocation result includes: Combining the environmental impact index value, the energy usage cost value, and the efficiency index value and generating candidate solutions through genetic algorithm operations according to the target optimization function; Performing local optimization on the candidate solutions through particle swarm optimization operations to obtain optimized solutions; Performing global optimization on the optimized solutions through simulated annealing operations to obtain global optimal solutions; Take the global optimal solution as the updated energy allocation result.

5. The method according to claim 1, wherein Before the step of inputting the energy characteristic information into a preset energy consumption prediction model to obtain an energy consumption prediction result, it includes: Divide the energy-related data into a training set, a test set, and a validation set; Obtain the energy data of the training set and establish an initial energy consumption prediction model; Optimize the initial energy consumption prediction model according to the energy data of the training set to obtain a preset energy consumption prediction model.

6. The method according to claim 5, characterized in that, The step of optimizing the initial energy consumption prediction model according to the energy data of the training set to obtain a preset energy consumption prediction model includes: Obtain the actual energy consumption value; Initialize the model parameters of the initial energy consumption prediction model; Perform normalization processing on the energy data of the training set to obtain processed data; Input the processed data into the initial energy consumption prediction model to obtain a predicted energy consumption value; Calculate according to the loss function to obtain the error value between the predicted energy consumption value and the actual energy consumption value; Obtain the learning rate; Iteratively update the model parameters according to the learning rate through an optimization algorithm until the maximum number of iterations is reached or the error value calculated by the loss function converges to a preset threshold to obtain a preset energy consumption prediction model.

7. The method according to claim 1, wherein The step of extracting features from the energy-related data to obtain energy characteristic information includes: Perform data cleaning on the energy-related data to obtain cleaned data; Perform outlier processing and replacement on the cleaned energy-related data to obtain data after outlier processing; Perform standard normalization based on the data after outlier processing to obtain normalized data; Perform Fourier transform according to the normalized data to obtain characteristic information; Screen the characteristic information through statistical tests to obtain energy characteristic information.

8. An energy management device based on a closed-loop feedback mechanism, characterized in that, The device includes: An acquisition module for acquiring energy-related data, where the energy-related data includes real-time energy prices, device status information, and user behavior patterns; A feature extraction module for extracting features from the energy-related data to obtain energy characteristic information; A calculation module for inputting the energy characteristic information into a preset energy consumption prediction model to obtain an energy consumption prediction result, where the preset energy consumption model is a long short-term memory network model; An allocation module for performing energy allocation on the energy consumption prediction result according to an energy usage strategy to obtain an energy allocation result; An optimization module for optimizing the energy allocation result according to an evaluation and optimization strategy to achieve energy management based on a closed-loop feedback mechanism; and also for performing an environmental impact assessment on the energy allocation result to obtain an environmental impact index value; performing a cost and efficiency assessment on the energy allocation result to obtain an energy usage cost value and an efficiency index value; combining the environmental impact index value, the energy usage cost value, and the efficiency index value and performing energy allocation according to the target optimization function to obtain an updated energy allocation result; and performing an evaluation on the updated energy allocation result until the preset energy allocation requirements are met to form a closed-loop feedback mechanism.

9. An energy management device based on a closed-loop feedback mechanism, characterized in that, The device includes: a memory, a processor, and an energy management program based on a closed-loop feedback mechanism stored on the memory and running on the processor, and the energy management program based on the closed-loop feedback mechanism is configured to implement the steps of the energy management method based on the closed-loop feedback mechanism according to any one of claims 1-7.

10. A storage medium, characterized in that, An energy management program based on a closed-loop feedback mechanism is stored on the storage medium, and when the energy management program based on the closed-loop feedback mechanism is executed by a processor, the steps of the energy management method based on the closed-loop feedback mechanism according to any one of claims 1-7 are implemented.

Images