A multi-level hierarchical aggregation control method and system for adjustable resources

Abstract

This invention belongs to the field of load regulation technology and provides a multi-level hierarchical aggregated regulation method and system for adjustable resources. The method includes acquiring historical electricity consumption data and clustering buildings based on electricity consumption; performing controllability analysis on the electrical equipment in each building group to determine the load response level of the equipment; classifying building energy consumption into different energy consumption patterns; inputting historical electricity consumption data into a prediction model to obtain electricity load prediction results for different response levels under different energy consumption patterns; and regulating the electricity load of the equipment based on the prediction results. This invention performs zoned aggregated regulation according to the priority of energy consumption pattern, controllability, and response speed.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of load regulation, and particularly relates to a multi-level hierarchical aggregation regulation method and system for adjustable resources. BACKGROUND

[0002] The statements in this section merely provide background information related to the present disclosure and do not necessarily constitute the prior art.

[0003] With the gradual improvement of power facilities, the situation of continuous power shortage is alleviated, but the power grid still faces great peak shaving pressure during the load peak. In some areas, the phenomenon of power shortage still exists. The imbalance between power demand and resource distribution and the seasonal imbalance between supply and demand lead to the fact that the power situation is still severe. At the same time, with the reform of the power market and the development of smart grid technology, power users can actively change their power consumption mode through the demand response mechanism to realize load reduction and transfer, so as to realize the balance between supply and demand of the power system through the control of the load side.

[0004] The existing aggregation regulation scheme only considers the resources that can be adjusted, and does not consider the response speed of the equipment and the different energy consumption modes due to different seasons, so it cannot realize fine aggregation regulation, and the stability of the system is insufficient. SUMMARY

[0005] In view of the above defects, the present disclosure provides a multi-level hierarchical aggregation regulation method and system for adjustable resources, which performs partition aggregation regulation according to the priority of three levels of energy consumption mode, adjustability and response speed, can more finely implement aggregation regulation on the load, improves the adaptability and response speed of the system, and realizes comprehensive, collaborative and efficient management of the multi-energy system.

[0006] In order to achieve the above purpose, the present disclosure adopts the following technical scheme:

[0007] The first aspect of the present disclosure provides a multi-level hierarchical aggregation regulation method for adjustable resources, comprising the following steps:

[0008] Obtain historical power consumption data, and cluster the buildings according to the power consumption;

[0009] Perform adjustability analysis on the power consumption equipment of each building group to determine the load response level of the power consumption equipment;

[0010] Divide the building energy consumption into different energy consumption modes;

[0011] Input the historical power consumption data into a prediction model to obtain power consumption load prediction results of different response levels under different energy consumption modes;

[0012] According to the prediction result, the power consumption load of the power consumption equipment is regulated.

[0013] As a further implementation manner, the controllability analysis is the response and regulation capability of the equipment to the power system.

[0014] As a further implementation manner, the load response level includes instantaneous response, short-time response and medium-long time response.

[0015] As a further implementation manner, the energy consumption mode includes cold, heat and electricity.

[0016] As a further implementation manner, the three energy consumption modes have different power consumption peak periods and power consumption characteristics.

[0017] As a further implementation manner, according to the prediction result, the power consumption load of the power consumption equipment is regulated, specifically:

[0018] The trend and change rule of the predicted load are analyzed, and when the actual load is higher or lower than the change range of the predicted load by more than a set threshold, the controllable equipment is adjusted to reduce or increase the load.

[0019] As a further implementation manner, the historical power consumption data includes the power consumption of each building, the energy consumption time distribution and the load curve.

[0020] The second aspect of the present disclosure provides a multi-level hierarchical aggregated regulation system for adjustable resources, comprising:

[0021] A clustering module is configured to obtain historical power consumption data, and cluster the buildings according to the power consumption;

[0022] A load response level determination module is configured to analyze the controllability of the power consumption equipment of each building group, and determine the load response level of the power consumption equipment;

[0023] An energy consumption mode division module is configured to divide the building energy consumption into different energy consumption modes;

[0024] A model prediction module is configured to input the historical power consumption data into a prediction model, and obtain the power consumption load prediction result of different response levels under different energy consumption modes;

[0025] A regulation module is configured to regulate the power consumption load of the power consumption equipment according to the prediction result.

[0026] The third aspect of the present disclosure provides a medium having a program stored thereon, wherein the program is executed by a processor to implement the steps of the multi-level hierarchical aggregated regulation method for adjustable resources according to the first aspect of the present disclosure.

[0027] The fourth aspect of the present disclosure provides an electronic device, comprising a memory, a processor, and a program stored in the memory and executable on the processor, wherein the processor implements the steps of the multi-level hierarchical aggregation regulation method for adjustable resources according to the first aspect of the present disclosure when executing the program.

[0028] Compared with the prior art, the present disclosure has the following beneficial effects:

[0029] The multi-level hierarchical aggregation regulation method and system for adjustable resources according to the present disclosure can implement more refined aggregation regulation of the load according to the priorities of the energy consumption mode, controllability, and response speed, making the energy system more flexible and capable of adapting to changes in different times, seasons, and business demands, thereby improving the adaptability and response speed of the system.

[0030] The multi-level hierarchical aggregation regulation method and system for adjustable resources according to the present disclosure propose a TAM-CNN-LSTM multi-task learning method for extracting seasonal features, and add SARIMA and VAE models according to the different characteristics of instantaneous short-time and medium-long-time responses, which can improve the representation and generalization capabilities of the model for load data, thereby improving the accuracy and stability of load prediction.

[0031] The advantages of the additional aspects of the present disclosure will be partially given in the following description, partially become obvious from the following description, or be understood through the practice of the present disclosure. BRIEF DESCRIPTION OF DRAWINGS

[0032] The accompanying drawings, which form a part of the present disclosure, are used to provide further understanding of the present disclosure, and the illustrative embodiments of the present disclosure and their description serve to explain the present disclosure, and do not constitute improper limitations on the present disclosure.

[0033] Figure 1 is a flowchart of a multi-level hierarchical aggregation regulation method for adjustable resources according to the present disclosure. DETAILED DESCRIPTION

[0034] The present disclosure will be further described below in conjunction with the accompanying drawings and embodiments.

[0035] It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the present disclosure. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present disclosure belongs.

[0036] The embodiments in the present disclosure and the features in the embodiments can be combined with each other as long as there is no conflict.

[0037] Embodiment one

[0038] As Figure 1 shown, the embodiment of the present disclosure provides a multi-level hierarchical aggregation regulation method for adjustable resources, including the following steps:

[0039] S1 Obtain historical power consumption data, and cluster buildings according to power consumption;

[0040] S2 Analyze the controllability of power consumption equipment of each building group, and determine the load response level of the power consumption equipment;

[0041] S3 Divide building energy consumption into different energy consumption modes;

[0042] S4 Input historical power consumption data into a prediction model to obtain power consumption load prediction results of different response levels under different energy consumption modes;

[0043] S5 According to the prediction results, the power consumption load of the power consumption equipment is regulated.

[0044] According to the priority of the three levels of energy consumption mode, controllability and response speed, the partition aggregation regulation is carried out.

[0045] The energy consumption mode is taken as the first priority factor, because the energy consumption mode can directly reflect the importance of actual demand: the energy consumption mode determines whether the power demand of a region or system has regularity and predictability. The priority of those regions with long-term stable and high load demand (such as hospitals and communication base stations) is naturally higher, because they are more dependent on energy supply and power outage will have a serious impact. Therefore, the regulation must first ensure the power supply of these regions.

[0046] Optimize resource allocation: By prioritizing different types of energy consumption modes, energy regulation resources can be more reasonably allocated. In the case of limited regulation capacity, high-priority power demand scenarios are prioritized to maximize energy use efficiency.

[0047] Controllability is taken as the second priority factor, because the flexibility of controllability affects the scheduling strategy: controllability determines whether a load can be flexibly involved in energy scheduling. In energy regulation, the more easily a load can be regulated, the greater its contribution to grid stability. Therefore, loads with high controllability can better cope with power supply fluctuations, peak shaving and other scheduling needs. Although the energy consumption mode determines the rigidity of demand, if the load controllability is strong, it can be adjusted effectively while meeting stable demand, reducing the pressure on the power grid.

[0048] Reduce energy waste: Loads with high controllability can adjust power according to real-time grid conditions, thereby avoiding unnecessary energy waste. Under the dual carbon target, the scheduling system must prioritize controllable loads to maximize energy efficiency.

[0049] Response speed is prioritized as the third factor because it impacts the system's emergency response capability: response speed measures how quickly the load can adjust after receiving a dispatch instruction. While response speed is crucial for emergency dispatching or peak shaving, it builds upon the first two factors. Even if a region or load possesses rapid response capabilities, its dispatch value diminishes if its energy consumption patterns are unimportant or uncontrollable. Therefore, while response speed is important, it only becomes a key indicator for optimizing control after a reasonable selection of energy consumption patterns and controllable loads.

[0050] It can improve system flexibility: fast-response loads can react quickly in emergencies, helping the power grid maintain stability during periods of severe load fluctuations. However, such loads are usually designed to handle extreme situations or short-term fluctuations and are not a primary consideration in daily operation.

[0051] First, S1 acquires historical electricity consumption data, including the electricity consumption, energy consumption time distribution, and load curves of each building. The k-means clustering algorithm is used to cluster buildings based on their electricity consumption. The basic idea of ​​the k-means clustering algorithm is to iteratively find a partitioning scheme with k clusters. A primary prerequisite for k-means is determining the value of k; this paper uses the silhouette coefficient method to ultimately determine the value of k. This minimizes the loss function corresponding to the clustering results, as shown in equation (1).

[0052]

[0053] Where x i Represents the i-th sample, c i It is x i The cluster to which it belongs The silhouette coefficient represents the centroid of the cluster, and M is the total number of samples. It is a metric used to evaluate the quality of clustering, measuring whether each data point is appropriately assigned to a cluster and the degree of separation between clusters. The formula for the silhouette coefficient is shown in equation (2):

[0054]

[0055] Where a(i) represents the cohesion of the sample points, and the calculation method is shown in equation (3):

[0056]

[0057] Here, j represents other sample points within the same class as sample i, and distance represents the distance between i and j. Therefore, the smaller a(i) is, the closer the classes are. b(i) is calculated in a similar way to a(i).

[0058] By considering factors such as geographical location and building attributes in each clustered building group, the similarity of individual buildings in adjacent areas but not in the same cluster can be measured, and individual buildings far from the building group can be assigned to adjacent large building groups.

[0059] Then, S2 performs a controllability analysis on the electrical equipment in each building complex (the user or equipment's response and control capabilities to the power system) to determine the load response level of the electrical equipment. Users or equipment with higher controllability can participate in control activities more flexibly, improving the system's stability and responsiveness.

[0060] The load response levels of equipment include instantaneous response, short-time response, and medium-to-long-time response. The specific classification criteria are as follows:

[0061] (1) Instantaneous response

[0062] Required time: Response time ranges from a few seconds to one minute.

[0063] Applicable scenarios: Suitable for loads that are sensitive to power changes and have the ability to adjust quickly, such as lighting and instantaneous adjustments of certain production equipment.

[0064] (2) Short-time response

[0065] Required time: Response time is from a few minutes to half an hour.

[0066] Applicable scenarios: Suitable for loads that require relatively quick adjustments but do not have extremely high requirements for response speed, such as the start-up and shutdown of some production lines, and the adjustment of air conditioning temperature.

[0067] (3) Medium and long time response

[0068] Required time: Response time ranges from several hours to one day.

[0069] Applicable scenarios: Suitable for loads with relatively slow power changes and relatively long production processes, such as the switching of some industrial production lines and the overall adjustment of building energy consumption.

[0070] S3 categorizes building energy consumption into different energy consumption patterns: cooling, heating, and electricity. Each of these patterns has different peak electricity consumption periods and characteristics. Dividing building energy consumption into these three aspects offers multiple benefits. First, this detailed division facilitates refined management, making energy consumption data clearer and providing a deeper understanding of the energy consumption of each part of the building system. Second, by separately monitoring and analyzing cooling, heating, and electricity consumption, energy bottlenecks in the system can be identified more accurately, allowing for targeted energy-saving measures to be proposed and implemented, minimizing energy waste. Furthermore, this division facilitates comprehensive system optimization, improving the overall energy efficiency of the building system. Through scientific decision-making, building managers can better understand the usage of each energy sector and formulate more scientific and rational decisions and plans.

[0071] S4 inputs historical electricity consumption data into the prediction model to obtain electricity load prediction results for different response levels under different energy consumption patterns.

[0072] Since cold, heat, and electricity loads have different peak electricity consumption periods and characteristics, they are subject to zoned aggregated regulation based on the different energy consumption habits of cold, heat, and electricity loads and according to the response levels of each load (instantaneous response, short-term response, and medium-to-long-term response).

[0073] First, historical electricity consumption data for each zone was analyzed to assess the characteristics of cooling, heating, and electricity loads. Historical load data was collected and analyzed, including load curves, peak loads, and load fluctuations for different time periods. Load curves for different load zones were plotted, showing daily, weekly, and monthly load changes. Fourier analysis was then used to decompose and analyze the complex load curves, accurately identifying periodic changes and performing quantitative analysis to gain a deeper understanding of the daily variation patterns of cooling, heating, and electricity loads.

[0074] Throughout the year, electricity and cooling loads show a trend of high in the middle and low at both ends, with strong local randomness and volatility, and peaks mainly concentrated in summer. Compared with electricity and cooling loads, heating load demand is low throughout the year and fluctuates greatly, with peaks mainly concentrated in winter. Clearly, climate change has a direct impact on user load demand, and the changes in each load throughout the year show a significant temporal correlation. Multi-task learning is used to predict these loads simultaneously. Electricity load features are extracted normally, while for cooling and heating loads, features are extracted based on the importance of cooling and heating usage in different seasons for building clusters. For example, in residential areas during summer, only electricity and cooling loads undergo further feature extraction. A seasonal multi-step feature extraction method is proposed for prediction, and the prediction model uses the TAM-CNN-LSTM model.

[0075] TAM (Time-Based Aspect Ratio) allows models to focus on the most relevant time points or regions in a time series, thereby improving their ability to perceive important features. By learning the weights of each time step in the time series, TAM enables the model to dynamically focus on information at different time points, thus more effectively capturing trends and correlations in the time series. CNN (Convolutional Networks) can effectively capture local features in time series data, such as sudden changes or periodic patterns. Through convolutional operations, CNNs can extract features at different scales and combine these features to better represent the structural information of the time series. LSTM (Laser-Based Memory) networks can capture long-term dependencies in time series data and have a better ability to handle the memory within the sequence.

[0076] The TAM-CNN-LSTM model first receives time-series data, typically a three-dimensional tensor including the number of samples, time step, and number of features. Then, based on different seasons, further feature extraction is performed on high-energy-consumption data (e.g., electricity and cooling loads in summer). The data is first processed through a Temporal Attention (TAM) layer, which establishes temporal dependencies in the time-series data, helping the model better focus on important time steps. The TAM layer usually consists of multiple attention mechanisms used to focus on and adjust different aspects of the time-series data. Next, feature extraction is performed through convolutional layers (CNNs), with the kernel size set to (k, F), where k is the kernel size and F is the number of features. Convolutional operations capture local relationships of features in the temporal dimension. Finally, the heating, cooling, and electricity data are processed through LSTM layers. LSTM units can memorize and update information in the sequence, thus achieving long-term dependency modeling of the sequence data. Finally, fully connected layers integrate the features extracted by the LSTM layers and output the final prediction result. In fully connected layers, dropout layers can be added to prevent overfitting. Overall, the TAM-CNN-LSTM model dynamically and selectively focuses on important information at different time points through a temporal attention mechanism. It extracts local features and improves the representation of time series data through CNNs, and captures long-term dependencies and processes memory through LSTMs, thus effectively extracting important features from time series data. This model can capture the temporal correlation, periodicity, and trend features in load data, thereby achieving accurate load forecasting. The TAM-CNN-LSTM model can effectively process time series data, extract important features, and achieve accurate predictions.

[0077] I. Power load forecasting at instantaneous and short-term response levels.

[0078] Since instantaneous response requires the prediction of instantaneous load and more accurate prediction is needed, a seasonal autoregressive integrated moving average (SARIMA) model is added to the aggregated control method at the instantaneous response level.

[0079] First, the SARIMA model is used to predict the electrical load data. The SARIMA model's prediction results are then used as additional features input into the TAM-CNN-LSTM model. The input data for the TAM-CNN-LSTM model includes: the original electrical load data, i.e., the historical load data sequence; and the SARIMA model's prediction results, i.e., the predicted load data for future time points. These two sets of data will serve as input features for the TAM-CNN-LSTM model to perform load prediction.

[0080] The output of the TAM-CNN-LSTM model is a prediction of electrical load data at future time points. By learning from historical load data and its correlation with the SARIMA model's predictions, the TAM-CNN-LSTM model can generate more accurate load forecasts. This fully utilizes the predictive power of time series models, providing richer and more accurate input features.

[0081] The prediction results of the SARIMA model are fused with the features extracted by the TAM-CNN-LSTM model. A feature concatenation method, such as weighted averaging, is used.

[0082] The SARIMA model and the TAM-CNN-LSTM model are sequentially integrated into a single model. First, the SARIMA model is used to make predictions on the data, yielding preliminary results. Then, these predictions are used as input to the TAM-CNN-LSTM model for further predictions, resulting in the final predictions from the integrated model. This sequential integration method fully leverages the advantages of both models, improving prediction accuracy and stability.

[0083] When training the ensemble model, techniques such as cross-validation are used to optimize the model's parameters and hyperparameters to improve the model's generalization ability and prediction performance.

[0084] After model training is complete, the integrated model is evaluated and optimized. The model is evaluated on a validation set, and the model parameters and feature fusion strategy are adjusted based on the evaluation results to further improve the model's prediction accuracy and stability.

[0085] Through the above steps, the SARIMA model is organically combined with the TAM-CNN-LSTM model, fully leveraging their respective advantages to improve the accuracy and reliability of load forecasting. The SARIMA model is a classic time series model suitable for modeling and forecasting data with seasonal, trend, and periodic variations. Instantaneous forecasting typically involves predicting data over a relatively short future timeframe, and the SARIMA model effectively captures the seasonal and periodic variations of data, demonstrating good predictive ability for data with significant seasonal characteristics. Furthermore, the SARIMA model's modeling and forecasting process is relatively simple and computationally fast, making it suitable for instantaneous forecasting tasks requiring rapid response and real-time prediction.

[0086] II. Electricity load forecasting for medium- and long-term response levels.

[0087] A variational autoencoder (VAE) is incorporated into a medium- to long-term load prediction model. First, historical load data is prepared as input, along with corresponding labels (i.e., load data for future timeframes) as the target for supervised learning. Simultaneously, the historical load data is normalized to ensure the stability and convergence of model training. Then, a variational autoencoder (VAE) model is constructed, comprising an encoder and a decoder. The encoder maps the input historical load data to distribution parameters in the latent space, modeled using a multi-layer neural network structure. The decoder decodes the representation in the latent space into reconstructed input data, also modeled using a multi-layer neural network structure. Finally, the VAE model is used to encode the historical load data at the input of the TAM-CNN-LSTM model, obtaining a latent representation, which is then input into the TAM-CNN-LSTM model along with the original historical load data. The entire model is trained end-to-end, adjusting model parameters by minimizing the loss function between predicted values ​​and true labels. Simultaneously, the reconstruction error and KL divergence in the VAE model are minimized to ensure that the VAE model learns appropriate latent representations. The ensemble model is evaluated using methods such as validation sets or cross-validation to assess its performance. Based on the evaluation results, the model is fine-tuned by adjusting hyperparameters and network structure to improve its prediction accuracy and generalization ability. By integrating VAE and TAM-CNN-LSTM models, the latent representations of data learned by the VAE model can be fully utilized, improving the TAM-CNN-LSTM model's ability to extract features and model load data, thereby enhancing the accuracy and stability of load forecasting.

[0088] S5 regulates the power load of electrical equipment based on the prediction results.

[0089] Specifically:

[0090] For instantaneous and short-term response level electricity load forecasting, the trends and patterns of forecast load data are analyzed. When the actual load is more than 10% higher or lower than the forecast load, controllable equipment will be adjusted to reduce or increase the load. Based on the established control strategy, corresponding control measures are implemented to adjust the actual load level to approximate the forecast load.

[0091] Energy purchase costs are high during peak periods. Instantaneous and short-term control measures can reduce these costs, improve energy utilization efficiency and system operating efficiency, and avoid energy waste and unnecessary costs. Timely response to load fluctuations and emergencies can reduce system operational risks and instability, ensuring the safe and stable operation of the system.

[0092] For medium- and long-term response level power load forecasting, since the forecasting time is longer, compared with instantaneous and short-term forecasting, control measures are implemented in advance before the load changes are predicted. The load of controllable equipment is adjusted to adapt to the expected changes in future load. Control measures are implemented in a timely manner through intelligent control system or operation and maintenance personnel, and the actual load changes are monitored.

[0093] By predicting and planning load trends and changes over a longer period, investment and operational strategies for equipment can be optimized, avoiding over-investment and resource waste, thereby reducing the overall system cost. Rational allocation of resources and scheduling of equipment operation maximizes the use of existing resources, improving overall system efficiency and productivity. Proactive identification and response to potential risks and problems reduce operational risks and uncertainties, enhancing the system's resilience and reliability.

[0094] Example 2

[0095] Embodiment 2 of this disclosure provides a multi-level hierarchical aggregation control system for adjustable resources, including:

[0096] The clustering module is configured to: acquire historical electricity consumption data and cluster buildings based on electricity consumption;

[0097] The load response level determination module is configured to perform controllability analysis on the electrical equipment in each building complex and determine the load response level of the electrical equipment.

[0098] The energy consumption mode classification module is configured to classify building energy consumption into different energy consumption modes.

[0099] The model prediction module is configured to input historical electricity consumption data into the prediction model to obtain electricity load prediction results with different response levels under different energy consumption patterns.

[0100] The control module is configured to control the electrical load of electrical equipment based on the prediction results.

[0101] The more detailed steps are the same as in Example 1, and will not be repeated here.

[0102] Example 3

[0103] This disclosure provides a medium on which a program is stored, which, when executed by a processor, implements the steps of a multi-level hierarchical aggregation control method for adjustable resources as described in this disclosure, embodiment one.

[0104] The more detailed steps are the same as in Example 1, and will not be repeated here.

[0105] Example 4

[0106] This disclosure provides an electronic device, including a memory, a processor, and a program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps in a multi-level hierarchical aggregation control method for adjustable resources as described in this disclosure.

[0107] The more detailed steps are the same as in Example 1, and will not be repeated here.

[0108] The above description is merely a preferred embodiment of this disclosure and is not intended to limit this disclosure. Various modifications and variations can be made to this disclosure by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A multi-level hierarchical aggregated regulation method for adjustable resources, characterized in that, The method comprises the following steps: obtaining historical power consumption data, and clustering buildings according to power consumption; analyzing the controllability of power consumption equipment of each building group to determine the load response level of the power consumption equipment; dividing building energy consumption into different energy consumption modes, including cold, heat and electricity; inputting the historical power consumption data into a prediction model to obtain power load prediction results of different response levels under different energy consumption modes, the prediction model using a TAM-CNN-LSTM model; controlling the power load of the power consumption equipment according to the prediction results; The controllability is the response and control ability of the equipment to the power system, and the load response level includes instantaneous response, short-time response and medium-long-time response. According to the priority of the three levels of energy consumption mode, controllability and response level, the zone is aggregated and controlled, wherein the energy consumption mode is taken as the first priority factor, the controllability is taken as the second priority factor, and the response speed is taken as the third priority factor. For the power load prediction of users with instantaneous and short-time response levels, first, use the SARIMA model to predict the data to obtain the preliminary prediction results; then, use the prediction results as the input of the TAM-CNN-LSTM model to predict again to obtain the final prediction results of the overall model. For the power load prediction of users with medium-long-time response levels, first, use the VAE model to encode the historical load data at the input end of the TAM-CNN-LSTM model to obtain the latent representation, and then input it together with the original historical load data into the TAM-CNN-LSTM model.

2. The multi-level hierarchical aggregation regulation method for adjustable resources according to claim 1, characterized in that, The three energy consumption modes have different power consumption peak periods and power consumption characteristics.

3. The multi-level hierarchical aggregation regulation method for adjustable resources as claimed in claim 1, characterized in that, According to the prediction results, the power load of the power consumption equipment is controlled, specifically: analyze the trend and change rule of the predicted load, and when the actual load is higher or lower than the change range of the predicted load by more than a set threshold, adjust the controllable equipment to reduce or increase the load.

4. The adjustable resource oriented multi-level hierarchical aggregation regulation method as claimed in claim 1, wherein, The historical power consumption data includes the power consumption, energy consumption time distribution and load curve of each building.

5. A multi-level hierarchical aggregated regulation system for adjustable resources, using a multi-level hierarchical aggregated regulation method for adjustable resources as claimed in any one of claims 1-4, characterized in that, It comprises: a clustering module configured to obtain historical power consumption data and cluster buildings according to power consumption; a load response level determination module configured to analyze the controllability of power consumption equipment of each building group to determine the load response level of the power consumption equipment; an energy consumption mode division module configured to divide building energy consumption into different energy consumption modes; a model prediction module configured to input the historical power consumption data into a prediction model to obtain power load prediction results of different response levels under different energy consumption modes; a control module configured to control the power load of the power consumption equipment according to the prediction results.

6. A medium having stored thereon a program, characterized by The program is executed by the processor to realize the steps in the multi-level hierarchical aggregation control method for adjustable resources according to any one of claims 1-4.

7. An electronic device comprising a memory, a processor, and a program stored in the memory and executable on the processor, characterized by The processor executes the program to realize the steps in the multi-level hierarchical aggregation control method for adjustable resources according to any one of claims 1-4.

Images