A load step-based industrial regulation potential evaluation method, device, equipment and medium

Abstract

This invention discloses a method, apparatus, equipment, and medium for assessing industrial load regulation potential based on load steps, addressing the technical problems of existing industrial load regulation potential assessment methods being highly dependent on process parameters, complex in modeling, and difficult to promote on a large scale to users. The invention includes: collecting historical active power data of the target industrial user; preprocessing the historical active power data to obtain preprocessed data; extracting multiple daily load sequences from the preprocessed data and extracting daily load behavior features from the daily load sequences; constructing a daily load behavior feature vector based on the daily load behavior features; clustering the daily load behavior feature vector using a clustering algorithm to determine the typical electricity consumption patterns of the target industrial user; performing load stepping processing on the typical daily load sequences corresponding to the typical electricity consumption patterns to obtain load steps; and calculating the industrial regulation potential corresponding to the target industrial user based on the load steps corresponding to all typical electricity consumption patterns.

Description

Technical Field

[0001] This invention relates to the field of industrial regulation potential assessment technology, and in particular to a method, apparatus, equipment and medium for assessing industrial regulation potential based on load steps. Background Technology

[0002] With the large-scale integration of renewable energy and the increasing demand for power system flexibility, the power grid is paying increasing attention to the development and utilization of flexible resources on the load side. Industrial user loads, characterized by large capacity, long operating times, and high levels of automation, can generate considerable load reduction potential over a certain time scale by adjusting production plans and starting / stopping some equipment within the limits of technological capabilities. This makes them an important regulating resource in the new power system. Accurately assessing the adjustable potential of industrial loads has become an urgent problem to be solved in achieving planned demand response.

[0003] Existing methods for assessing industrial load regulation potential largely rely on production mechanisms and equipment parameters. These methods involve establishing detailed equipment-level models, incorporating substantial internal enterprise data such as output, process constraints, and inventory constraints, and then optimizing these models to obtain the adjustable capacity and duration. Current technologies propose the following schemes for assessing industrial load regulation potential: A method for evaluating the regulation capability of electrolytic aluminum is proposed, which calculates the regulation capability of electrolytic aluminum based on parameters such as the upper and lower limits of power and temperature of the electrolytic cell, electrolyte mass, specific heat capacity of electrolyte, and heat absorption and dissipation.

[0004] A method for estimating the adjustable power range of an industrial park is proposed, which calculates the adjustable power range of the entire industrial park based on model parameters of production equipment such as electrolytic aluminum, electric arc furnace, and polysilicon.

[0005] A method for assessing the adjustability potential of industrial loads considering safety constraints is proposed. This method establishes load curves using industrial production load models of steel, cement, electrolytic aluminum, and other industries, and then assesses the adjustability potential of industrial loads.

[0006] However, while methods based on detailed equipment models are highly accurate, the information generated, such as process parameters, equipment configurations, and production plans, often involves core business secrets of enterprises and is difficult to obtain accurately. Furthermore, there are significant differences between different enterprises and different processes, resulting in poor universality and scalability of such methods. Summary of the Invention

[0007] This invention provides a method, apparatus, equipment, and medium for assessing industrial load regulation potential based on load steps, which addresses the technical problems of existing industrial load regulation potential assessment methods being highly dependent on process parameters, having complex modeling, and being difficult to promote on a large scale to users.

[0008] This invention provides a method for assessing industrial regulation potential based on load steps, comprising: Collect historical active power data of the target industrial user, and preprocess the historical active power data to obtain preprocessed data; Multiple daily load sequences are extracted from the preprocessed data, and daily load behavior features are extracted from the daily load sequences; Construct a daily load behavior feature vector based on the described daily load behavior characteristics; Clustering algorithms are used to cluster the daily load behavior feature vectors to determine the typical electricity consumption patterns of the target industrial users; The typical daily load sequence corresponding to the typical electricity consumption pattern is subjected to load stepping processing to obtain load steps, and the industrial regulation potential corresponding to the target industrial user is calculated based on the load steps corresponding to all typical electricity consumption patterns.

[0009] Optionally, the step of collecting historical active power data of the target industrial user and preprocessing the historical active power data to obtain preprocessed data includes: Collect historical active power data of the target industrial users and remove outliers from the historical active power data to obtain preliminary preprocessed data; Missing values ​​are filled into the preliminary preprocessed data to obtain the preprocessed data.

[0010] Optionally, the step of extracting daily load sequences from the preprocessed data and extracting daily load behavior features from the daily load sequences includes: Extract daily load sequences from the preprocessed data; Obtain the active power values ​​at each sampling time of the daily load sequence; The daily average power characteristics of the daily load sequence are calculated based on the active power values ​​at each sampling time. Extract the maximum and minimum values ​​of the active power value; Calculate the daily peak-valley power difference characteristics based on the maximum and minimum values; Calculate the power difference between adjacent sampling times based on the active power value; Based on the preset steady-state indication function and the power difference, a steady-state determination is performed on the time interval between adjacent sampling moments to obtain the determination result; The cumulative duration of steady-state power is calculated based on the determination results and interval duration at each time interval; The daily load behavior characteristics of the daily load sequence are generated by using the daily average power characteristics, the daily peak-valley power difference characteristics, and the cumulative duration of steady-state power.

[0011] Optionally, the step of clustering the daily load behavior feature vector using a clustering algorithm to determine the typical electricity consumption pattern of the target industrial user includes: Several daily load behavior feature vector samples are selected from the multiple daily load behavior feature vectors as initial cluster centers; Calculate the Euclidean distance between each of the daily load behavior feature vectors and the initial cluster center; The daily load behavior feature vector is clustered based on the Euclidean distance to obtain several clusters; Calculate the new cluster center of the cluster, replace the initial cluster center with the new cluster center, and return to the step of calculating the Euclidean distance between each of the daily load behavior feature vectors and the initial cluster center, until the preset iteration termination condition is met, and take each cluster as a typical power consumption pattern of the target industrial user.

[0012] Optionally, the step of performing load stepping processing on the typical daily load sequence corresponding to the typical electricity consumption pattern to obtain load steps, and calculating the industrial regulation potential corresponding to the target industrial user based on the load steps corresponding to all typical electricity consumption patterns, includes: The typical daily load sequence corresponding to the typical electricity consumption pattern is subjected to load stepping processing to obtain load steps, and the load step information matrix of the load steps is constructed. A reference load step and a target load step are selected from the load step information matrix, and the industrial regulation potential corresponding to the target industrial user is calculated based on the reference load step and the target load step corresponding to all typical power consumption patterns.

[0013] Optionally, the step of performing load stepping processing on the typical daily load sequence corresponding to the typical electricity consumption pattern to obtain load steps, and constructing the load step information matrix of the load steps, includes: A candidate step array is selected from the typical daily load sequence corresponding to the typical electricity consumption pattern; The typical daily load sequence that is closest to the candidate step array is added to the candidate step array to obtain the updated step array; Calculate the local rate of change of the updated step array within the local time window; When the local rate of change is less than a preset threshold, the candidate step array is updated using the updated step array, and the step array is updated by adding the typical daily load sequence that is closest to the candidate step array into the candidate step array. When the local rate of change is not less than a preset threshold, the updated step array is used as the target step array, and the step of selecting a candidate step array in the typical daily load sequence corresponding to the typical electricity consumption pattern is returned. The sampling duration of each target step array is sequentially determined to be not less than the preset minimum duration criterion; If so, the target step array is used as the load step; When all sampling times in the typical daily load sequence have been traversed, output all load steps; A load step information matrix for the typical power consumption pattern is generated using all the aforementioned load steps.

[0014] Optionally, the step of selecting a reference load step and a target load step in the load step information matrix, and calculating the industrial regulation potential corresponding to the target industrial user based on the reference load step and the target load step corresponding to all typical power consumption patterns, includes: Select the baseline load step and the target load step from the load step information matrix; Calculate the power difference between the baseline load step and the target load step to obtain the adjustable capacity corresponding to the typical power consumption mode; Calculate the percentage of adjustable capacity based on the adjustable capacity; Obtain the start and end times of the target load step; Calculate the adjustment time based on the start time and the end time; The industrial regulation potential corresponding to the target industrial user is generated based on the regulation capacity percentage and regulation time corresponding to all typical power consumption patterns.

[0015] The present invention also provides an industrial regulation potential assessment device based on load steps, comprising: The data acquisition and preprocessing module is used to acquire historical active power data of the target industrial user, and preprocess the historical active power data to obtain preprocessed data. The daily load behavior feature extraction module is used to extract multiple daily load sequences from the preprocessed data and extract daily load behavior features from the daily load sequences. The daily load behavior feature vector construction module is used to construct a daily load behavior feature vector based on the daily load behavior features. The typical electricity consumption pattern determination module is used to cluster the daily load behavior feature vector using a clustering algorithm to determine the typical electricity consumption pattern of the target industrial user. The industrial regulation potential calculation module is used to perform load stepping processing on the typical daily load sequence corresponding to the typical power consumption pattern to obtain the load step, and calculate the industrial regulation potential corresponding to the target industrial user based on the load step corresponding to all typical power consumption patterns.

[0016] The present invention also provides an electronic device, the device comprising a processor and a memory: The memory is used to store program code and transmit the program code to the processor; The processor is configured to execute, according to instructions in the program code, the industrial regulation potential assessment method based on load steps as described above.

[0017] The present invention also provides a computer-readable storage medium for storing program code for performing the industrial regulation potential assessment method based on load steps as described in any of the preceding claims.

[0018] As can be seen from the above technical solutions, the present invention has the following advantages: The embodiments of the present invention provide a method for assessing industrial regulation potential based on load steps, specifically disclosing: collecting historical active power data of target industrial users; preprocessing the historical active power data to obtain preprocessed data; extracting multiple daily load sequences from the preprocessed data and extracting daily load behavior features from the daily load sequences; constructing a daily load behavior feature vector based on the daily load behavior features; clustering the daily load behavior feature vector using a clustering algorithm to determine the typical electricity consumption patterns of the target industrial users; performing load stepping processing on the typical daily load sequences corresponding to the typical electricity consumption patterns to obtain load steps; and calculating the industrial regulation potential corresponding to the target industrial users based on the load steps corresponding to all typical electricity consumption patterns. The present invention can quantitatively assess the regulation capacity and duration based solely on the industrial user's electricity consumption data, without requiring additional information such as production processes and equipment parameters, and has strong scalability. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 A flowchart illustrating the steps of an industrial regulation potential assessment method based on load steps, provided for embodiments of the present invention; Figure 2 The power consumption versus load step curve diagram under conventional non-regulating power consumption mode provided in the embodiments of the present invention; Figure 3 This invention provides a power consumption versus load step curve diagram under a long-term low-power regulation power consumption mode, as shown in the embodiments of the present invention. Figure 4The power consumption versus load step curve diagram under the midday power consumption mode provided in this embodiment of the invention; Figure 5 The power consumption versus load step curve diagram under the nighttime medium-duration power consumption mode provided in this embodiment of the invention; Figure 6 The power consumption versus load step curve diagram under the afternoon power consumption mode with adjustable power consumption is provided in the embodiment of the present invention; Figure 7 This is a power consumption versus load step curve diagram under a short-time high-power regulation power consumption mode provided in an embodiment of the present invention; Figure 8 This is a schematic diagram illustrating the range of adjustment power ratio and adjustment time provided in an embodiment of the present invention; Figure 9 This is a structural block diagram of an industrial regulation potential assessment device based on load steps, provided as an embodiment of the present invention. Detailed Implementation

[0021] This invention provides a method, apparatus, equipment, and medium for assessing industrial load regulation potential based on load steps, which addresses the technical problems of existing industrial load regulation potential assessment methods being highly dependent on process parameters, having complex modeling, and being difficult to promote on a large scale to users.

[0022] To make the objectives, features, and advantages of this invention more apparent and understandable, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described below are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0023] Please see Figure 1 , Figure 1 A flowchart illustrating the steps of an industrial regulation potential assessment method based on load steps, provided in this embodiment of the invention.

[0024] This invention provides a method for assessing industrial regulation potential based on load steps, which may specifically include the following steps: Step 101: Collect historical active power data of the target industrial user, preprocess the historical active power data to obtain preprocessed data; In this embodiment of the invention, historical active power data of the target industrial user can be collected to assess the industrial regulation potential.

[0025] First, industrial users within the target area are numbered and documented. Historical active power data for these users over one year is collected using smart meters, distribution automation systems, or park energy consumption monitoring platforms. The sampling period can be a fixed 15-minute interval, dividing the day into 96 sampling times. Active power data from 365 consecutive days is selected as the sample for modeling and evaluation, forming a daily load power sequence arranged continuously by date. Let the industrial user be on the... The active power at each sampling time is ,in, , For discrete sampling time indexes, This represents the number of sampling points.

[0026] Next, to avoid the impact of outliers or missing values ​​on the evaluation results, the historical active power data of the target industrial users can be preprocessed to obtain preprocessed data.

[0027] In one example, step 101 may include the following sub-steps: S11: Collect historical active power data of the target industrial users and remove outliers from the historical active power data to obtain preliminary preprocessed data. During the historical power data acquisition process for industrial users, invalid power data may be generated due to factors such as metering equipment failure, communication disturbances, or instantaneous sampling anomalies. In order to reduce the impact of such abnormal sampling points on subsequent analysis results, this example first performs outlier detection and removal processing on the original load power sequence.

[0028] Specifically, a load power differential sequence is constructed based on the variation characteristics of load power between adjacent sampling times. The difference value corresponding to each sampling time is defined as: ; in, Indicates the first The change in active power at each sampling time point relative to the previous sampling time point. Further, statistical difference sequences... The mean and standard deviation are denoted as . and .use The criterion performs anomaly detection on the difference sequence. If the first... The difference values ​​corresponding to each sampling time point satisfy: ; If an anomaly is detected, it will be marked as an anomaly and set to null. All detected anomaly sampling points will be uniformly marked as null values ​​and processed in subsequent missing data filling steps.

[0029] S12, fill in missing values ​​in the preliminary preprocessed data to obtain the preprocessed data.

[0030] After outlier removal, the original power sequence may contain missing data due to communication interruptions, data loss, or outlier removal. To ensure the continuity and integrity of the load time series and to avoid data gaps affecting subsequent feature calculations and pattern recognition results, this embodiment further performs data filling processing on the missing data.

[0031] Specifically, for data loss caused by communication interruptions, the average of adjacent time points is used to fill in the gaps. The filling formula is as follows: ; in, , These represent the normal load power values ​​adjacent to the missing time t.

[0032] In one embodiment, when the missing sampling point is located at the beginning or end of the sequence, or when there is invalid data in the adjacent sampling values ​​before and after the missing sampling point, the two valid sampling values ​​closest to the missing sampling point are selected for mean filling.

[0033] Step 102: Extract multiple daily load sequences from the preprocessed data, and extract daily load behavior features from the daily load sequences; In this embodiment of the invention, after completing the acquisition and preprocessing of historical active power data of the target industrial user to obtain a continuous, complete load power time series with outliers removed as preprocessed data, in order to further characterize the intraday load operation characteristics of the target industrial user and provide input for subsequent typical electricity consumption pattern identification and adjustability potential assessment, this embodiment of the invention extracts multiple daily load sequences from the preprocessed data, and extracts daily load behavior features that can reflect the user's load level, fluctuation degree, and stable operation characteristics from the daily load sequence curves.

[0034] In one example, this embodiment of the invention selects "daily average power characteristics," "daily peak-valley power difference characteristics," and "steady-state power duration characteristics" as dimensions for representing the load behavior of industrial users, in order to construct a daily load behavior feature vector for industrial users. Step 102 may include the following sub-steps: S21, Extract the daily load sequence from the preprocessed data; S22, obtain the active power value at each sampling time of the daily load sequence; S23, calculate the daily average power characteristics of the daily load sequence based on the active power values ​​at each sampling time; In the specific implementation, based on the preprocessed active power sequence, the first... The daily load sequence for each day is represented as follows: ; in, Indicates that industrial users are in the first Heavenly The active power value at each sampling time; Indicates the sample day index; Indicates the index of discrete sampling times within a day; This indicates the total number of sampling points in a single day.

[0035] Extracting daily average power characteristics: Daily average power characteristics are used to reflect the power consumption of industrial users in the first... The average load level within a day can characterize the overall energy consumption characteristics of an industrial user during the day's production and operation, and provide a basis for distinguishing between operating states such as full-load production, low-load maintenance, and shutdown for maintenance.

[0036] The daily average power characteristic is defined as: ; in, For the first The daily average power characteristic value.

[0037] S24, extract the maximum and minimum values ​​of active power; S25, calculate the daily peak-valley power difference characteristics based on the maximum and minimum values; The daily peak-valley power difference characteristic is used to characterize the degree of load fluctuation of industrial users within a single day, and can reflect the power adjustment range and operational flexibility during the production process.

[0038] Specifically, the daily peak-valley power difference characteristic is defined as: ; ; in, The first The maximum and minimum active power of the day. This refers to the difference between peak and off-peak power.

[0039] S26, Calculate the power difference between adjacent sampling times based on the active power value; S27, perform steady-state determination on the time interval between adjacent sampling times based on the preset steady-state indication function and power difference, and obtain the determination result; S28, calculate the cumulative duration of steady-state power based on the judgment results and interval duration at each time interval; The cumulative duration of steady-state power characterizes the cumulative duration of the stable operating segment in the load curve of industrial users, reflecting the ability of the load to remain near a relatively constant level. This feature can be used to identify industrial load operation modes with long-term stable operation characteristics.

[0040] Specifically, the power difference between adjacent time points is defined as follows: ; in, Indicates the first Heavenly The change in power at each sampling time relative to the previous sampling time.

[0041] Furthermore, the steady-state determination threshold is set as follows: In one embodiment, It can be determined based on the rated capacity of industrial users, historical power fluctuation distribution, or preset experience ratio, and is used to determine whether the power change between adjacent sampling times is within an acceptable small fluctuation range.

[0042] Furthermore, the steady-state indication function is constructed as follows: ; Among them, when When, it indicates the first Heavenly The load in each sampling interval is in steady-state operation; when When the time is specified, it indicates that the load within the sampling interval has fluctuated significantly.

[0043] Then the first The characteristic of the cumulative duration of steady-state power over a day can be expressed as: ; in, This represents the cumulative duration of steady-state power on day d when the load is in a steady-state operating state. This represents the time interval between adjacent sampling times.

[0044] S29 uses daily average power characteristics, daily peak-valley power difference characteristics, and steady-state power cumulative duration to generate daily load behavior characteristics of daily load sequences.

[0045] After calculating the daily average power characteristics, daily peak-valley power difference characteristics, and steady-state power cumulative duration, the daily load behavior characteristics of the daily load sequence can be generated.

[0046] Step 103: Construct a daily load behavior feature vector based on the daily load behavior characteristics; After obtaining the daily load behavior characteristics of the daily load sequence, a daily load behavior feature vector can be constructed.

[0047] In one example, the three-dimensional vector of the industrial daily load behavior characteristics on day d is as follows: ; It should be noted that, due to the differences in numerical ranges among the daily average power characteristics, daily peak-valley power difference characteristics, and steady-state power cumulative duration characteristics, directly performing cluster analysis based on the daily load behavior feature vector may lead to one dimension dominating distance calculations, affecting the accuracy of typical electricity consumption pattern identification. Therefore, the constructed daily load behavior feature vector is normalized to unify the numerical scale of each dimension.

[0048] Let the first Heavenly dimensional eigenvalues , for the After performing Min-Max normalization on the dimension, we get: ; in, For the first The maximum and minimum values ​​of the dimensional feature across all sample days; After normalization, the first The daily load behavior feature vector is represented as follows: ; in, This is the normalized daily load behavior feature vector for day d. This represents the normalized daily average power characteristic on day d. The normalized daily peak-valley power difference characteristics on day d. The cumulative duration of steady-state power on day d after normalization.

[0049] Step 104: Cluster the daily load behavior feature vectors using a clustering algorithm to determine the typical electricity consumption patterns of the target industrial users; In real-world scenarios, industrial users' electricity demand is significantly heterogeneous over time due to factors such as production plans, order fluctuations, and equipment maintenance. To accurately classify the multi-condition electricity consumption patterns of industrial users, unsupervised clustering analysis can be performed on the daily load behavior feature vectors based on the K-means clustering algorithm. This uncovers typical electricity consumption models inherent in the data, laying the foundation for subsequent load tiering processing and adjustment potential assessment based on condition classification.

[0050] In one example, step 104 may include the following sub-steps: S41, Select several daily load behavior feature vector samples from multiple daily load behavior feature vectors as initial cluster centers; In practical implementation, the selection of initial cluster centers affects the convergence speed and stability of the K-means clustering algorithm. To effectively segment typical electricity consumption patterns of industrial users, the number of cluster centers K needs to be determined in advance. Based on the given number of clusters K, a random sampling method is used to select several daily load behavior feature vector samples from multiple daily load behavior feature vectors as initial cluster centers.

[0051] Let the set of normalized daily load behavior characteristic vectors be: ; in, This represents the total number of sample days. Indicates the first The normalized daily load behavior feature vector corresponding to each day.

[0052] From the set Random selection The feature vectors of the daily samples are used as the initial cluster centers, denoted as follows: ; in, Indicates the first 1 initial cluster center vector, .

[0053] Let the first During the nth iteration The cluster center vectors are: ; in, This refers to the feature dimension number. In this embodiment of the invention, These correspond to three characteristic dimensions: average daily load, daily peak-valley power difference, and steady-state power duration, respectively. Indicates the first During the nth iteration The cluster centers at the in Components on the dimensional features, .

[0054] S42, calculate the Euclidean distance between the daily load behavior feature vector and the initial cluster center; In this embodiment of the invention, in order to quantify the feature similarity between the samples to be clustered and each initial cluster center, Euclidean distance is used as a similarity metric to achieve preliminary category assignment of the samples.

[0055] Let the normalized daily load behavior feature vector of each d-th sample be: ; Then the first During the nth iteration, the 1st The sample and the first The Euclidean distance between cluster centers can be expressed as: ; in, For the first During the nth iteration Sample and the first The distance between cluster centers For the first The first day of the sample The eigenvalues ​​of the normalized daily load behavior feature vector; For the first During the nth iteration The first cluster center feature vector Dimensional components.

[0056] S43, cluster the daily load behavior feature vectors according to Euclidean distance to obtain several clusters; According to the minimum distance principle, each sample is assigned to the category corresponding to the cluster center with the smallest Euclidean distance, thus completing the initial category assignment of the samples. Let the th... During the nth iteration The clusters are Then we have: ; in, Indicates the first In the next iteration, it was divided into the... A sample set of classes.

[0057] S44, calculate the new cluster center of the cluster, replace the initial cluster center with the new cluster center, and return to the step of calculating the Euclidean distance between the daily load behavior feature vector and the initial cluster center, until the preset iteration termination condition is met, and each cluster is taken as a typical power consumption pattern of the target industrial user.

[0058] To ensure that the cluster centers gradually approximate the true centroids of the sample clusters, the cluster centers are dynamically updated based on the category assignment results, thereby optimizing the clustering results. For each cluster... Calculate the mean of each dimension of the feature vector of all samples within the cluster to obtain the cluster centers for the next iteration: ; in, For clusters The number of samples included.

[0059] Through multiple iterations, a two-way optimization of sample categories and cluster centers is achieved. Convergence conditions are set to determine the final clustering result, ensuring the scientific rigor and effectiveness of the clustering analysis. Steps S42-S43 are repeated until the algorithm converges or reaches the maximum number of iterations. After all iterations are complete, the final result is obtained. Each cluster consists of a category and its corresponding cluster center. Each cluster corresponds to a typical electricity consumption pattern, reflecting the representative load behavior characteristics of the target industrial user under specific operating conditions.

[0060] Step 105: Perform load stepping processing on the typical daily load sequence corresponding to the typical power consumption pattern to obtain the load step, and calculate the industrial regulation potential corresponding to the target industrial user based on the load steps corresponding to all typical power consumption patterns.

[0061] In this embodiment of the invention, after identifying typical power consumption patterns, the load curves of typical daily load sequences corresponding to the clustering results can be processed by load stepping, dividing the continuously changing load curves into several stable load steps, and calculating the industrial regulation potential corresponding to typical power consumption patterns based on the load steps.

[0062] In one example, step 105 may include the following sub-steps: S51, load stepping is performed on the typical daily load sequence corresponding to the typical electricity consumption pattern to obtain the load step, and the load step information matrix of the load step is constructed. In this embodiment of the invention, after identifying typical electricity consumption patterns, the load curves of typical daily load sequences corresponding to the clustering results can be subjected to load stepping processing. The continuously changing load curves are divided into several stable load steps, and the average load power, start time, and end time corresponding to each step are recorded. Based on this, a load step information matrix is ​​constructed, providing a basis for subsequent adjustment capability assessment. Specifically, this may include the following sub-steps: S511, select a candidate step array from the typical daily load sequence corresponding to the typical electricity consumption pattern; In practical implementation, to achieve standardized analysis of load data for each typical electricity consumption pattern, the load power time series of typical days for each cluster is extracted to form an analysis sample with a unified time scale. The load power of typical days is read at a uniform 15-minute interval to obtain the typical day load sequence for that typical day. ; in, This represents the current electricity consumption scenario in the [number]th [year]. Load power at each sampling time; This represents the total number of sampling points per day. In this embodiment of the invention, the sampling interval is 15 minutes.

[0063] To identify stable load steps in a load sequence, in the load sequence Define a candidate step array This is used to temporarily store the currently identified local load data during the time series analysis. In the initial stage, the first four sampling points of the load sequence are loaded into a candidate step array, denoted as: .

[0064] S512, add the typical daily load sequence that is closest to the candidate step array into the candidate step array to obtain the updated step array; S513, calculate the local rate of change of the updated step array within the local time window; S514, When the local rate of change is less than the preset threshold, the candidate step array is updated by updating the step array, and the typical daily load sequence that is closest to the candidate step array is added to the candidate step array to obtain the step array update step. S515, when the local rate of change is not less than the preset threshold, the updated step array is used as the target step array, and the step of selecting the candidate step array in the typical daily load sequence corresponding to the typical power consumption mode is returned. To determine whether the local load segment corresponding to the candidate step array meets the load step condition, a local rate of change is introduced. As an indicator for determining load steps, the local rate of change is used. for: ; ; in, The maximum load difference within the local time frame covered by the candidate step array. and These represent the maximum and minimum load values ​​within the local time window covered by the candidate step array, respectively. and These represent the mean and standard deviation of the corresponding load sequence difference sequence, respectively.

[0065] In this embodiment, the comprehensive criterion for the establishment of an effective load step is set as follows: the local rate of change threshold is set as follows: .

[0066] Starting from the 5th sampling point, the load sequence... Perform time series traversal. During the traversal, attempt to incorporate the current sampling point into the candidate step array. and dynamically update the local rate of change. Based on this, it can be determined whether the current load segment belongs to the same load step.

[0067] When traversing to the th When sampling points, the current point Add the candidate steps to the array and recalculate the local rate of change. The traversal method is as follows: ; That is: when joining Still satisfied At that time, Append to the current candidate array It was assumed that the load was still within the same load step; however, after the addition of [the load], [the following occurred]. At that time, the four most recent sampling points Reconstruct the candidate array to indicate the end of the original steps and the start of detecting new steps.

[0068] S516, sequentially determine whether the sampling duration of each target step array is not less than the preset minimum duration criterion; S517, if so, use the target step array as the load step; S518: When all sampling times in a typical daily load sequence have been traversed, output all load steps; S519 uses all load steps to generate a load step information matrix for typical power consumption patterns.

[0069] In the specific implementation, during the traversal process, the time index is adjusted based on whether the current sampling point belongs to the step region. Update: when When determined to be within the load step, let: ; when Not on the load step, but at its previous sampling point When on a load step, skip a few transient points, let: ; when and If none of the steps are on the load step, continue traversing backwards with a step size of 1. make: ; When candidate step data The local rate of change criterion is always satisfied during the continuous expansion process. And the duration was kept secret. If the time period is valid, it is recorded as an effective load step; otherwise, it is considered a transient process and is not recorded as an effective step.

[0070] Through the above traversal and judgment process, several effective load steps that meet the stability and duration conditions can be identified from the load curves corresponding to each power consumption scenario.

[0071] For each identified effective load step, its corresponding average load power and start and end times are extracted and recorded, and a load step feature vector is constructed accordingly. Let the first... i The first typical electricity consumption day identified j The load step is: ; in This indicates that the step is at the 1st The average load power of each load step within the corresponding time period; Indicates the start time of the step; Indicates the end time of the step; This indicates the number of steps identified in the entire daily load curve.

[0072] Based on this, the first All load step feature vectors identified under a specific electricity consumption scenario are arranged in chronological order to construct the load step information matrix corresponding to that electricity consumption scenario: ; For all electricity consumption scenarios, the load step information matrix corresponding to each scenario can be further summarized to obtain the overall load step information set: ; in, This represents the number of typical electricity consumption patterns.

[0073] S52, select the benchmark load step and the target load step in the load step information matrix, and calculate the industrial regulation potential corresponding to the target industrial user based on the benchmark load step and the target load step corresponding to all typical power consumption modes.

[0074] After completing the load stepping process for each typical power consumption mode and forming a load step information matrix, the adjustment capability of the target industrial user under different power consumption scenarios is evaluated based on the stable load steps identified by each typical power consumption mode.

[0075] In one example, S52 may include the following sub-steps: S521, Select the baseline load step and the target load step in the load step information matrix; In practical implementation, for typical power consumption patterns, a stable load step representing normal production conditions is selected from its stepped load curve and denoted as the "baseline load step," with its step power denoted as... This step corresponds to the normal operating state when users do not participate in demand response. The load fluctuates slightly around this power level and can be considered as continuous and stable.

[0076] In the same typical power consumption pattern, considering constraints such as equipment operating status and process continuity, a lower load step that can be stably maintained after adjustment is selected from the stepped load curve and denoted as the "target load step," with its step power denoted as... This step represents the reduced load state in which the user can operate stably for a long time under the premise of ensuring production safety and process continuity, and is a target reference for adjustment potential.

[0077] S522, calculate the power difference between the baseline load step and the target load step to obtain the adjustable capacity corresponding to the typical power consumption mode; S523, calculate the percentage of adjustable capacity based on adjustable capacity; Selected reference load step Target load step After that, the The regulating capacity in a given power consumption scenario is defined as the power difference between the two: ; in, This represents the adjustable capacity in the m-th power consumption scenario.

[0078] The corresponding regulation capacity percentage is defined as the proportion of regulation capacity relative to the baseline load: ; in, Indicates the first Adjustable capacity for various power consumption scenarios.

[0079] S524, obtain the start and end times of the target load step; S525, calculate the adjustment time based on the start and end times; In this embodiment of the invention, the start and end times of the target load step can be used as a reference. , Calculate the adjustment time during which the load can be stably maintained after adjustment in this power consumption scenario: .

[0080] Therefore, the regulation potential point corresponding to the m-th typical electricity consumption pattern can be expressed as: ; Where, the x-axis Indicates adjustment time, y-axis This indicates the percentage of adjusted capacity.

[0081] For all power consumption scenarios, the set of adjustment potential points for the target industrial user can be obtained: ; Where M represents the number of electricity consumption scenarios formed by clustering.

[0082] S526 generates the industrial regulation potential for the target industrial user based on the regulation capacity percentage and regulation time corresponding to all typical power consumption modes.

[0083] In this embodiment of the invention, the percentage of adjustment capacity corresponds to all power consumption scenarios. and adjustment time To summarize, the maximum percentage limit of the maximum regulating capacity and the maximum regulating time for this industrial user are defined as follows: ; This indicates the maximum continuous adjustment time that the target industrial user can achieve across all power consumption scenarios.

[0084] Given adjustment time requirements The maximum adjustable capacity percentage is: ; The user's adjustment potential range in the power-time plane is: ; in, This represents the set of adjustment potential regions for users on the power-time plane; the horizontal axis represents the adjustment time T, and the vertical axis represents the percentage of adjustment capacity R.

[0085] This invention can quantitatively evaluate the regulation capacity and duration based solely on the power consumption data of industrial users, without requiring additional information such as production processes and equipment parameters, and has strong scalability.

[0086] For ease of understanding, the embodiments of the present invention will be described below through specific examples: The actual power consumption data of a paper mill was selected as input to simulate, evaluate, and verify the aforementioned method. The paper mill recorded active power data at 15-minute intervals throughout the year. The power consumption and load steps for each power consumption mode are shown below. Figures 2-7 As shown, the user's power adjustment ratio and adjustment time range are as follows: Figure 8 As shown.

[0087] Please see Figure 9 , Figure 9 This is a structural block diagram of an industrial regulation potential assessment device based on load steps, provided as an embodiment of the present invention.

[0088] This invention provides an industrial regulation potential assessment device based on load steps, comprising: The data acquisition and preprocessing module 901 is used to acquire historical active power data of the target industrial user, preprocess the historical active power data, and obtain preprocessed data. The daily load behavior feature extraction module 902 is used to extract multiple daily load sequences from preprocessed data and extract daily load behavior features from the daily load sequences. The daily load behavior feature vector construction module 903 is used to construct a daily load behavior feature vector based on the daily load behavior characteristics. The typical electricity consumption pattern determination module 904 is used to cluster the daily load behavior feature vector using a clustering algorithm to determine the typical electricity consumption pattern of the target industrial user. The industrial regulation potential calculation module 905 is used to perform load stepping processing on the typical daily load sequence corresponding to the typical power consumption pattern to obtain the load step, and calculate the industrial regulation potential corresponding to the target industrial user based on the load step corresponding to all typical power consumption patterns.

[0089] In this embodiment of the invention, the data acquisition and preprocessing module 901 includes: The abnormal data removal submodule is used to collect historical active power data of target industrial users and remove outliers from the historical active power data to obtain preliminary preprocessed data. The missing value imputation submodule is used to impute missing values ​​in the preliminary preprocessed data to obtain the preprocessed data.

[0090] In this embodiment of the invention, the daily load behavior feature extraction module 902 includes: The daily load sequence extraction submodule is used to extract the daily load sequence from the preprocessed data. The active power value acquisition submodule is used to acquire the active power value at each sampling time of the daily load sequence; The daily average power characteristic calculation submodule is used to calculate the daily average power characteristic of the daily load sequence based on the active power value at each sampling time. The maximum and minimum value extraction submodule is used to extract the maximum and minimum values ​​of active power. The daily peak-valley power difference characteristic calculation submodule is used to calculate the daily peak-valley power difference characteristic based on the maximum and minimum values. The power difference calculation submodule is used to calculate the power difference between adjacent sampling times based on the active power value; The determination submodule is used to perform steady-state determination on the time interval between adjacent sampling times based on a preset steady-state indication function and power difference, and obtain the determination result. The steady-state power cumulative duration calculation submodule is used to calculate the steady-state power cumulative duration based on the judgment results and interval duration at each time interval; The daily load behavior feature generation submodule is used to generate daily load behavior features of the daily load sequence by using daily average power features, daily peak-valley power difference features, and steady-state power cumulative duration.

[0091] In this embodiment of the invention, the typical power consumption pattern determination module 904 includes: The initial cluster center selection submodule is used to select several daily load behavior feature vector samples as initial cluster centers from multiple daily load behavior feature vectors; The Euclidean distance calculation submodule is used to calculate the Euclidean distance between the daily load behavior feature vector and the initial cluster center; The clustering submodule is used to cluster the daily load behavior feature vectors based on Euclidean distance to obtain several clusters; The typical power consumption pattern determination submodule is used to calculate the new cluster center of the cluster, replace the initial cluster center with the new cluster center, and return to the step of calculating the Euclidean distance between the daily load behavior feature vector and the initial cluster center, until the preset iteration termination condition is met, and each cluster is regarded as a typical power consumption pattern of the target industrial user.

[0092] In this embodiment of the invention, the industrial regulation potential calculation module 905 includes: The load step information matrix construction submodule is used to perform load stepping processing on the typical daily load sequence corresponding to the typical electricity consumption pattern, obtain the load steps, and construct the load step information matrix of the load steps. The industrial regulation potential calculation submodule is used to select the benchmark load step and the target load step in the load step information matrix, and calculate the industrial regulation potential corresponding to the target industrial user based on the benchmark load step and the target load step corresponding to all typical power consumption patterns.

[0093] In this embodiment of the invention, the load step information matrix construction submodule includes: The candidate step array selection unit is used to select a candidate step array from the typical daily load sequence corresponding to a typical power consumption pattern; The updated step array generation unit is used to add the typical daily load sequence that is closest to the candidate step array into the candidate step array to obtain the updated step array; The local rate of change calculation unit is used to calculate the local rate of change of the updated step array within the local time window; The first return unit is used to update the candidate step array by updating the step array when the local rate of change is less than a preset threshold, and to return the step array by adding the typical daily load sequence that is closest to the candidate step array into the candidate step array. The second return unit is used to update the step array as the target step array when the local rate of change is not less than a preset threshold, and return the steps of selecting the candidate step array in the typical daily load sequence corresponding to the typical power consumption mode. The sampling duration judgment unit is used to sequentially determine whether the sampling duration of each target step array is not less than the preset minimum duration criterion; The load step determination unit is used to determine the target step array as the load step if the condition is met. The load step output unit is used to output all load steps when all sampling times in a typical daily load sequence have been traversed. The load step information matrix generation unit is used to generate a load step information matrix for typical power consumption patterns using all load steps.

[0094] In this embodiment of the invention, the industrial regulation potential calculation submodule includes: The reference load step and target load step selection unit is used to select the reference load step and target load step from the load step information matrix; The adjustable capacity acquisition unit is used to calculate the power difference between the baseline load step and the target load step to obtain the adjustable capacity corresponding to a typical power consumption mode. Adjustable capacity percentage calculation unit, used to calculate the adjustable capacity percentage based on the adjustable capacity; The start and end time extraction unit is used to obtain the start and end times of the target load step. The adjustment time calculation unit is used to calculate the adjustment time based on the start time and the end time. The industrial regulation potential generation unit is used to generate the industrial regulation potential corresponding to the target industrial user based on the regulation capacity percentage and regulation time corresponding to all typical power consumption modes.

[0095] This invention also provides an electronic device, which includes a processor and a memory: The memory is used to store program code and transfer the program code to the processor; The processor is used to execute the load step-based industrial regulation potential assessment method of this invention according to instructions in the program code.

[0096] This invention also provides a computer-readable storage medium for storing program code for executing the load step-based industrial regulation potential assessment method of this invention.

[0097] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0098] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0099] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, apparatus, or computer program products. Therefore, embodiments of the present invention can take the form of entirely hardware embodiments, entirely software embodiments, or embodiments combining software and hardware aspects. Furthermore, embodiments of the present invention can take the form of computer program products implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0100] Embodiments of the present invention are described with reference to flowchart illustrations and / or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0101] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing terminal device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0102] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal equipment, causing a series of operational steps to be performed on the computer or other programmable terminal equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable terminal equipment for implementing the process. Figure 1One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0103] Although preferred embodiments of the present invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the embodiments of the present invention.

[0104] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0105] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal device that includes said element.

[0106] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for assessing industrial regulation potential based on load steps, characterized in that, include: Collect historical active power data of the target industrial user, and preprocess the historical active power data to obtain preprocessed data; Multiple daily load sequences are extracted from the preprocessed data, and daily load behavior features are extracted from the daily load sequences; Construct a daily load behavior feature vector based on the described daily load behavior characteristics; Clustering algorithms are used to cluster the daily load behavior feature vectors to determine the typical electricity consumption patterns of the target industrial users; The typical daily load sequence corresponding to the typical electricity consumption pattern is subjected to load stepping processing to obtain load steps, and the industrial regulation potential corresponding to the target industrial user is calculated based on the load steps corresponding to all typical electricity consumption patterns.

2. The method according to claim 1, characterized in that, The steps of collecting historical active power data from target industrial users and preprocessing the historical active power data to obtain preprocessed data include: Collect historical active power data of the target industrial users and remove outliers from the historical active power data to obtain preliminary preprocessed data; Missing values ​​are filled into the preliminary preprocessed data to obtain the preprocessed data.

3. The method according to claim 1, characterized in that, The steps of extracting daily load sequences from the preprocessed data and extracting daily load behavior features from the daily load sequences include: Extract daily load sequences from the preprocessed data; Obtain the active power values ​​at each sampling time of the daily load sequence; The daily average power characteristics of the daily load sequence are calculated based on the active power values ​​at each sampling time. Extract the maximum and minimum values ​​of the active power value; Calculate the daily peak-valley power difference characteristics based on the maximum and minimum values; Calculate the power difference between adjacent sampling times based on the active power value; Based on the preset steady-state indication function and the power difference, a steady-state determination is performed on the time interval between adjacent sampling moments to obtain the determination result; The cumulative duration of steady-state power is calculated based on the determination results and interval duration at each time interval; The daily load behavior characteristics of the daily load sequence are generated by using the daily average power characteristics, the daily peak-valley power difference characteristics, and the cumulative duration of steady-state power.

4. The method according to claim 1, characterized in that, The step of clustering the daily load behavior feature vector using a clustering algorithm to determine the typical electricity consumption pattern of the target industrial user includes: Several daily load behavior feature vector samples are selected from the multiple daily load behavior feature vectors as initial cluster centers; Calculate the Euclidean distance between each of the daily load behavior feature vectors and the initial cluster center; The daily load behavior feature vector is clustered based on the Euclidean distance to obtain several clusters; Calculate the new cluster center of the cluster, replace the initial cluster center with the new cluster center, and return to the step of calculating the Euclidean distance between each of the daily load behavior feature vectors and the initial cluster center, until the preset iteration termination condition is met, and take each cluster as a typical power consumption pattern of the target industrial user.

5. The method according to claim 1, characterized in that, The step of performing load stepping processing on the typical daily load sequence corresponding to the typical electricity consumption pattern to obtain load steps, and calculating the industrial regulation potential corresponding to the target industrial user based on the load steps corresponding to all typical electricity consumption patterns, includes: The typical daily load sequence corresponding to the typical electricity consumption pattern is subjected to load stepping processing to obtain load steps, and the load step information matrix of the load steps is constructed. A reference load step and a target load step are selected from the load step information matrix, and the industrial regulation potential corresponding to the target industrial user is calculated based on the reference load step and the target load step corresponding to all typical power consumption patterns.

6. The method according to claim 5, characterized in that, The step of performing load stepping processing on the typical daily load sequence corresponding to the typical electricity consumption pattern to obtain load steps, and constructing the load step information matrix of the load steps, includes: A candidate step array is selected from the typical daily load sequence corresponding to the typical electricity consumption pattern; The typical daily load sequence that is closest to the candidate step array is added to the candidate step array to obtain the updated step array; Calculate the local rate of change of the updated step array within the local time window; When the local rate of change is less than a preset threshold, the candidate step array is updated using the updated step array, and the step array is updated by adding the typical daily load sequence that is closest to the candidate step array into the candidate step array. When the local rate of change is not less than a preset threshold, the updated step array is used as the target step array, and the step of selecting a candidate step array in the typical daily load sequence corresponding to the typical electricity consumption pattern is returned. The sampling duration of each target step array is sequentially determined to be not less than the preset minimum duration criterion; If so, the target step array is used as the load step; When all sampling times in the typical daily load sequence have been traversed, output all load steps; A load step information matrix for the typical power consumption pattern is generated using all the aforementioned load steps.

7. The method according to claim 5, characterized in that, The step of selecting a reference load step and a target load step in the load step information matrix, and calculating the industrial regulation potential corresponding to the target industrial user based on the reference load step and the target load step corresponding to all typical power consumption patterns, includes: Select the baseline load step and the target load step from the load step information matrix; Calculate the power difference between the baseline load step and the target load step to obtain the adjustable capacity corresponding to the typical power consumption mode; Calculate the percentage of adjustable capacity based on the adjustable capacity; Obtain the start and end times of the target load step; Calculate the adjustment time based on the start time and the end time; The industrial regulation potential corresponding to the target industrial user is generated based on the regulation capacity percentage and regulation time corresponding to all typical power consumption patterns.

8. An industrial regulation potential assessment device based on load steps, characterized in that, include: The data acquisition and preprocessing module is used to acquire historical active power data of the target industrial user, and preprocess the historical active power data to obtain preprocessed data. The daily load behavior feature extraction module is used to extract multiple daily load sequences from the preprocessed data and extract daily load behavior features from the daily load sequences. The daily load behavior feature vector construction module is used to construct a daily load behavior feature vector based on the daily load behavior features. The typical electricity consumption pattern determination module is used to cluster the daily load behavior feature vector using a clustering algorithm to determine the typical electricity consumption pattern of the target industrial user. The industrial regulation potential calculation module is used to perform load stepping processing on the typical daily load sequence corresponding to the typical power consumption pattern to obtain the load step, and calculate the industrial regulation potential corresponding to the target industrial user based on the load step corresponding to all typical power consumption patterns.

9. An electronic device, characterized in that, The device includes a processor and a memory: The memory is used to store program code and transmit the program code to the processor; The processor is configured to execute the industrial regulation potential assessment method based on load steps as described in any one of the claims 1-7 according to the instructions in the program code.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium is used to store program code for executing the industrial regulation potential assessment method based on load steps as described in any one of claims 1-7.

Images