An industrial adjustable load potential analysis method and system based on a clustering algorithm

Abstract

The present application relates to the technical field of power grid load scheduling, and proposes an industrial adjustable load potential analysis method and system based on a clustering algorithm, which comprises obtaining historical data of a plurality of adjustable load devices as a sample set, including device name, industry to which the device belongs, rated power, operating capacity, peak shaving regulation capacity and valley filling regulation capacity; the sample set is subjected to clustering processing and divided into K categories; the average peak shaving regulation capacity and average valley filling regulation capacity of all adjustable load devices in each category are calculated, and a feature library of adjustable load resource pool is constructed. Through the above technical solution, the adjustable load potential of the device can be quickly and accurately predicted, and reliable data basis is provided for load scheduling.

Description

Technical Field

[0001] This invention relates to the field of power grid load dispatching technology, specifically to a method and system for analyzing the potential of industrial adjustable loads based on clustering algorithms. Background Technology

[0002] With the gradual adjustment of the economic structure, the total electricity consumption of the whole society continues to rise. Although power construction has developed rapidly, local and seasonal power shortages still exist. At the same time, in order to promote the transformation of the energy structure and achieve energy conservation and emission reduction, the proportion of new energy power generation such as wind power and photovoltaic power generation has risen rapidly, resulting in a significant shortage of peak-shaving capacity of power generation resources, which seriously affects the safe and stable operation of the power system.

[0003] To address the increasingly severe supply-demand imbalance, demand-side resources are receiving more and more attention. Load dispatch, as an important component of demand-side resources, is one of the key means to alleviate power shortages. Adjustable load potential is an important reference for realizing load dispatch; however, there is a lack of research on the prediction of adjustable load potential in existing technologies. Summary of the Invention

[0004] This invention proposes a method and system for analyzing the industrial adjustable load potential based on a clustering algorithm. By clustering historical data of adjustable load equipment, the method enables the prediction of adjustable load potential and solves the aforementioned technical problems.

[0005] The technical solution of the present invention is as follows:

[0006] Firstly, a method for analyzing the industrial adjustable load potential based on clustering algorithms includes the following steps:

[0007] Enter the adjustable load equipment parameters for the target user, including equipment name, industry of the equipment, rated power, and operating capacity;

[0008] Based on the input adjustable load equipment parameters, the target category is searched in the adjustable load resource pool feature library, and the average peak shaving adjustment capacity and average valley filling adjustment capacity corresponding to the target category are obtained as the prediction results. The target category is specifically: equipment name and industry.

[0009] The construction process of the adjustable load resource pool feature library includes:

[0010] Historical data from multiple adjustable load devices were collected as a sample set, including device name, industry, rated power, operating capacity, peak shaving capacity, and valley filling capacity.

[0011] The sample set is clustered into K categories;

[0012] Calculate the average peak shaving capacity and average valley filling capacity of all adjustable load devices in each category, and construct a feature library of adjustable load resource pool.

[0013] Secondly, an industrial adjustable load potential analysis system based on clustering algorithms includes,

[0014] The first acquisition module is used to acquire historical data of multiple adjustable load devices as a sample set, including device name, industry of the device, rated power, operating capacity, peak shaving capacity and valley filling capacity;

[0015] The first processing module is used to cluster the sample set into K categories.

[0016] The first calculation module is used to calculate the average peak shaving capacity and average valley filling capacity of all adjustable load devices in each category, and to build a feature library of the adjustable load resource pool.

[0017] The first input module is used to input the adjustable load equipment parameters of the target user, including equipment name, industry of the equipment, rated power, and operating capacity;

[0018] The second calculation module is used to search for target categories in the adjustable load resource pool feature library based on the input adjustable load equipment parameters, and obtain the average peak shaving adjustment capacity and average valley filling adjustment capacity corresponding to the target category as prediction results.

[0019] Thirdly, a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of a method for analyzing the potential of industrial adjustable loads based on a clustering algorithm.

[0020] The working principle and beneficial effects of this invention are as follows:

[0021] This invention utilizes the K-means clustering algorithm to classify historical data of adjustable load equipment, constructing a feature library for the adjustable load resource pool. Based on the parameters of the adjustable load equipment input by the target user, it automatically finds similar categories and matches the peak-shaving and valley-filling response capacities within each category. This achieves the effect of automatically predicting the adjustable load potential based on equipment characteristic parameters. Through this invention, the adjustable load potential of equipment can be predicted quickly and accurately, providing a reliable data foundation for load scheduling. Attached Figure Description

[0022] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0023] Figure 1 This is a flowchart illustrating the design of the adjustable load feature library of the present invention.

[0024] Figure 2 This is a flowchart illustrating the user-adjustable load potential prediction process of the present invention. Detailed Implementation

[0025] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0026] Example 1

[0027] like Figures 1-2 As shown, this embodiment proposes a method for analyzing the industrial adjustable load potential based on a clustering algorithm, including:

[0028] (1) Obtain historical data of multiple adjustable load devices that have participated in the power consumption response as a sample set, including device name, industry of the device, rated power, operating capacity, peak shaving regulation capacity and valley filling regulation capacity.

[0029] Historical data of adjustable load equipment is preprocessed. Before clustering, the adjustable load data is standardized and outlier data is filtered. The processed data includes equipment name, industry (the same type of equipment may have different adjustable loads in different industries), rated power, operating capacity, peak shaving capacity, and valley filling capacity.

[0030] (2) Cluster the sample set into K categories, select a reasonable value for K, use Gap Statistic to automatically find the best value for K, and randomly select K center points;

[0031] The compact support measure of clustering is represented by the Euclidean distance between sample points within a cluster, denoted as D. k D k Smaller clusters have better compact support.

[0032] After standardization,

[0033] Reference measure values ​​are obtained using Monte Carlo sampling.

[0034] Gap n (k)=E n *logW k -logW k ,

[0035] in, B represents the number of samples;

[0036] To correct for errors introduced by Monte Carlo sampling, we calculate the standard deviation. in,

[0037] Select that meets the gap. k ≥Gap k+1 -s k+1 The minimum K value is taken as the optimal K value.

[0038] Select K center points, denoted as μ1, μ2, ..., μ k .

[0039] (3) Define the loss function

[0040] Where x i Let c represent the i-th sample. i It is x i The cluster to which it belongs This represents the cluster center point, and M is the total number of samples.

[0041] Let t = 0, 1, 2, ... be the iteration steps. Repeat the convergence steps until the loss function J monotonically decreases to the minimum value. This will allow you to adjust the clustering results of the load equipment to converge.

[0042] The convergence steps are as follows:

[0043] For each sample x i Assign it to the nearest center

[0044] For each category, the center of that category is recalculated.

[0045] (5) Based on the clustering results, calculate the average peak shaving capacity and average valley filling capacity of all adjustable load equipment in each category, and construct an adjustable load resource pool feature library;

[0046] (6) Input the adjustable load equipment parameters of users who are not participating in the power response, including equipment name, industry, rated power, and operating capacity; based on the input adjustable load equipment parameters, search for the target category in the above-mentioned adjustable load resource pool feature library, and obtain the average peak shaving regulation capacity and average valley filling regulation capacity corresponding to the target category as the prediction result. If the target user has multiple adjustable load equipment, sum the prediction results of all adjustable load equipment to obtain the adjustable potential parameter of the target user.

[0047] Example 2

[0048] Based on the same concept as Embodiment 1 above, this embodiment also proposes an industrial adjustable load potential analysis system based on clustering algorithm, including:

[0049] The first acquisition module is used to acquire historical data of multiple adjustable load devices that have participated in the power consumption response as a sample set, including device name, industry of the device, rated power, operating capacity, peak shaving regulation capacity and valley filling regulation capacity;

[0050] The first processing module is used to cluster the sample set into K categories.

[0051] The first calculation module is used to calculate the average peak shaving capacity and average valley filling capacity of all adjustable load devices in each category, and to build a feature library of the adjustable load resource pool.

[0052] The first input module is used to input the parameters of adjustable load equipment of users who are not participating in the power response, including equipment name, industry of equipment, rated power and operating capacity;

[0053] The second calculation module is used to find the target category in the above-mentioned adjustable load resource pool feature library based on the input adjustable load equipment parameters, and obtain the average peak shaving adjustment capacity and average valley filling adjustment capacity corresponding to the target category as the prediction result.

[0054] Furthermore, the aforementioned first processing module includes,

[0055] The second processing module is used to automatically determine the optimal number of categories K using Gap Statistic;

[0056] The first marking module is used to select K center points, denoted as μ1, μ2, ..., μ3. k ;

[0057] The second labeling module is used to define the loss function. Where x i Let c represent the i-th sample. i It is x i The cluster to which it belongs This represents the cluster center point, and M is the total number of samples.

[0058] The third processing module is used to repeatedly execute the convergence step until the loss function J monotonically decreases to its minimum value, which can adjust the convergence of the clustering results of the load equipment.

[0059] Furthermore, the aforementioned second processing module includes,

[0060] The third calculation module is used to calculate the Euclidean distance between sample points within each class.

[0061] The fourth processing module is used to obtain the results after standardization.

[0062] The fifth processing module is used to obtain the reference measure value Gap using the Monte Carlo sampling method. n (k)=E n *logW k -logW k ,in B represents the number of samples;

[0063] The fourth calculation module is used to calculate the standard deviation. in,

[0064] The first selection module is used to select the option that meets the gap. k ≥Gap k+1 -s k+1 The minimum K value is taken as the optimal K value.

[0065] The various variations and specific examples of the industrial adjustable load potential analysis method based on clustering algorithm in the foregoing embodiments are also applicable to the industrial adjustable load potential analysis system based on clustering algorithm in this embodiment. Through the detailed description of the industrial adjustable load potential analysis method based on clustering algorithm described above, those skilled in the art can clearly understand the implementation method of the industrial adjustable load potential analysis system based on clustering algorithm in this embodiment. Therefore, for the sake of brevity, it will not be described in detail here.

[0066] Example 3

[0067] Based on the same concept as Embodiment 1 above, this embodiment also proposes a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of an industrial adjustable load potential analysis method based on a clustering algorithm.

[0068] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for analyzing the potential of adjustable industrial loads based on clustering algorithms, characterized in that, Includes the following steps: Enter the adjustable load equipment parameters for the target user, including equipment name, industry, rated power, and operating capacity; Based on the input adjustable load equipment parameters, the target category is searched in the adjustable load resource pool feature library, and the average peak shaving adjustment capacity and average valley filling adjustment capacity corresponding to the target category are obtained as the prediction results. The target category is specifically: equipment name and industry. The construction process of the adjustable load resource pool feature library includes: Historical data from multiple adjustable load devices were collected as a sample set, including device name, industry, rated power, operating capacity, peak shaving capacity, and valley filling capacity. Use Gap Statistic to automatically determine the optimal number of categories K; Calculate the Euclidean distance between sample points within each class. ; After standardization, ; Reference measure values ​​are obtained using Monte Carlo sampling. ,in B represents the number of samples; Calculate the standard deviation ,in, , ; Select the one that satisfies The minimum K value is taken as the optimal K value; Select K center points, denoted as , , ..., ; Define loss function ,in Represents the i-th sample. yes The cluster to which it belongs This represents the cluster center point, and M is the total number of samples. Repeat the convergence step until the loss function J monotonically decreases to its minimum value, thus completing the clustering process; Calculate the average peak shaving capacity and average valley filling capacity of all adjustable load devices in each category, and construct a feature library of adjustable load resource pool.

2. The method for analyzing the industrial adjustable load potential based on clustering algorithm according to claim 1, characterized in that, The convergence step includes: For each sample Assign it to the nearest center , For each category, the center of that category is recalculated. , where t=0,1,2,...,t is the iteration step.

3. An industrial adjustable load potential analysis system based on clustering algorithm, characterized in that, include, The first acquisition module is used to acquire historical data of multiple adjustable load devices as a sample set, including device name, industry of the device, rated power, operating capacity, peak shaving capacity and valley filling capacity; The first processing module is used to cluster the sample set into K categories. The first processing module includes, The second processing module is used to automatically determine the optimal number of categories K using Gap Statistic; The second processing module includes, The third calculation module is used to calculate the Euclidean distance between sample points within each class. ; The fourth processing module is used to obtain the results after standardization. ; The fifth processing module is used to obtain reference measure values ​​using the Monte Carlo sampling method. ,in B represents the number of samples; The fourth calculation module is used to calculate the standard deviation. ,in, , ; The first selection module is used to select those that meet the requirements. The minimum K value is taken as the optimal K value; The first marking module is used to select K center points, denoted as... , , ..., ; The second labeling module is used to define the loss function. ,in Represents the i-th sample. yes The cluster to which it belongs This represents the cluster center point, and M is the total number of samples. The third processing module is used to repeatedly execute the convergence step until the loss function J monotonically decreases to the minimum value, which can adjust the convergence of the clustering results of the load equipment. The first calculation module is used to calculate the average peak shaving capacity and average valley filling capacity of all adjustable load devices in each category, and to build a feature library of the adjustable load resource pool. The first input module is used to input the adjustable load equipment parameters of the target user, including equipment name, industry of the equipment, rated power, and operating capacity; The second calculation module is used to search for target categories in the adjustable load resource pool feature library based on the input adjustable load equipment parameters, and obtain the average peak shaving adjustment capacity and average valley filling adjustment capacity corresponding to the target category as prediction results.

4. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the industrial adjustable load potential analysis method based on a clustering algorithm as described in any one of claims 1 or 2.

Images