Industrial park distributed energy management method, device, system, and storage medium

Abstract

The present application provides a method, device, system, and storage medium for distributed energy management in an industrial park, which belongs to the field of energy management technology. The method includes: predicting the first photovoltaic power generation power of the industrial park in the prediction period based on a physical model; determining a first eigenvector based on fault prediction data of photovoltaic power generation equipment and meteorological data of the industrial park; performing feature matching on the first eigenvector with multiple second eigenvectors in the target strategy library to obtain multiple third eigenvectors; determining a second correction coefficient based on the first correction coefficients corresponding to the multiple third eigenvectors, so as to obtain a second photovoltaic power generation power based on the second correction coefficient; and scheduling and managing the distributed energy of the industrial park based on the load demand data of the industrial park and the second photovoltaic power generation power. The method, device, system, and storage medium for distributed energy management in an industrial park provided by the present application can improve the utilization efficiency of distributed energy in the industrial park.

Description

Technical Field

[0001] The present application belongs to the field of energy management technology, and more specifically, relates to a distributed energy management method, device, system, and storage medium for an industrial park. Background Art

[0002] With the development of distributed energy technologies, renewable energy sources such as photovoltaic power generation are increasingly being used in industrial parks. However, the intermittent and fluctuating power generation of photovoltaic power generation conflicts with the stable power supply characteristics of traditional power grids. This complicates the coordinated scheduling of multiple energy systems (electricity, heat, cooling, and gas), making efficient energy utilization difficult.

[0003] Therefore, it is urgent to propose a new distributed energy management method for industrial parks to improve energy utilization efficiency. Summary of the Invention

[0004] The purpose of this application is to provide an industrial park distributed energy management method and device, system, and storage medium to improve the utilization efficiency of distributed energy in industrial parks.

[0005] A first aspect of an embodiment of the present application provides a method for managing distributed energy in an industrial park, including:

[0006] Based on the physical parameters of the photovoltaic power generation equipment and the solar irradiance during the forecast period, a physical model is used to predict the first photovoltaic power generation power of the industrial park during the forecast period, wherein the solar irradiance during the forecast period is obtained based on the meteorological data during the forecast period;

[0007] Determining a first eigenvector based on fault prediction data of photovoltaic power generation equipment during a prediction period and meteorological data of the industrial park during the prediction period;

[0008] Performing feature matching on the first feature vector and a plurality of second feature vectors in a target strategy library to obtain a plurality of third feature vectors; wherein each second feature vector in the target strategy library has a corresponding first correction coefficient, and the third feature vector is a vector among the plurality of second feature vectors that matches the first feature vector;

[0009] determining a second correction coefficient corresponding to the first eigenvector based on first correction coefficients corresponding to a plurality of third eigenvectors, so as to correct the first photovoltaic power generation power based on the second correction coefficient to obtain a second photovoltaic power generation power;

[0010] Distributed energy in the industrial park is dispatched and managed based on the load demand data corresponding to the industrial park in the forecast period and the second photovoltaic power generation power.

[0011] A second aspect of an embodiment of the present application provides an industrial park distributed energy management device, including:

[0012] a photovoltaic prediction module for predicting the first photovoltaic power generation power of the industrial park during the prediction period using a physical model based on the physical parameters of the photovoltaic power generation equipment and the solar irradiance during the prediction period, wherein the solar irradiance during the prediction period is obtained based on the meteorological data during the prediction period;

[0013] a feature extraction module, configured to determine a first feature vector based on fault prediction data of the photovoltaic power generation equipment during a prediction period and meteorological data of the industrial park during the prediction period;

[0014] a feature matching module, configured to perform feature matching on the first feature vector with a plurality of second feature vectors in a target strategy library to obtain a plurality of third feature vectors; wherein each second feature vector in the target strategy library has a corresponding first correction coefficient, and the third feature vector is a vector among the plurality of second feature vectors that matches the first feature vector;

[0015] a parameter determination module, configured to determine a second correction coefficient corresponding to the first eigenvector based on first correction coefficients corresponding to the plurality of third eigenvectors, so as to correct the first photovoltaic power generation power based on the second correction coefficient to obtain a second photovoltaic power generation power;

[0016] The scheduling management module is used to schedule and manage the distributed energy of the industrial park based on the load demand data corresponding to the industrial park in the forecast period and the second photovoltaic power generation power.

[0017] The third aspect of the embodiments of the present application provides an industrial park distributed energy management system, including a controller, the controller including a memory, a processor, and a computer program stored in the memory and running on the processor, and the processor implements the steps of the above-mentioned industrial park distributed energy management method when executing the computer program.

[0018] In a fourth aspect of an embodiment of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned industrial park distributed energy management method are implemented.

[0019] The beneficial effects of the industrial park distributed energy management method, device, system, and storage medium provided by the embodiments of the present application are:

[0020] The embodiment of the present application first predicts the first photovoltaic power generation power of the industrial park in the prediction period based on the physical model. Then, taking into account the impact of the fault status of the photovoltaic power generation equipment and meteorological data on the output power of the photovoltaic power generation equipment, a first eigenvector is determined based on the fault prediction data of the photovoltaic power generation equipment in the prediction period and the meteorological data of the industrial park in the prediction period. Through similarity search of the eigenvectors, the most relevant second eigenvector is automatically selected from the target strategy library. The second correction coefficient corresponding to the first eigenvector is determined based on the first correction coefficient corresponding to the second eigenvector. Based on the second correction coefficient, the first power generation power output by the physical model is accurately corrected in real time to obtain the second photovoltaic power generation power. Distributed energy management of the industrial park based on the second photovoltaic power generation power is conducive to improving energy utilization efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the embodiments or descriptions of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0022] Figure 1 A flow chart of a distributed energy management method for an industrial park provided in one embodiment of the present application;

[0023] Figure 2 A structural block diagram of a distributed energy management device for an industrial park provided in one embodiment of the present application;

[0024] Figure 3 A schematic block diagram of a controller provided in accordance with an embodiment of the present application. DETAILED DESCRIPTION

[0025] In the following description, specific details such as specific system structures and techniques are provided for purposes of illustration rather than limitation to facilitate a thorough understanding of the embodiments of the present application. However, it will be apparent to those skilled in the art that the present application may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obscuring the description of the present application with unnecessary detail.

[0026] In order to make the purpose, technical solutions and advantages of this application clearer, specific embodiments will be described below with reference to the accompanying drawings.

[0027] Please refer to Figure 1 , Figure 1A flowchart of a distributed energy management method for an industrial park provided in an embodiment of the present application may be executed by a controller in a distributed energy management system for an industrial park. The method may include:

[0028] S101: Based on the physical parameters of the photovoltaic power generation equipment and the solar irradiance during the forecast period, a physical model is used to predict the first photovoltaic power generation power of the industrial park during the forecast period. The solar irradiance during the forecast period is obtained based on the meteorological data during the forecast period.

[0029] In this embodiment, the physical parameters of the photovoltaic power generation equipment include parameters such as the number of photovoltaic modules, the area of ​​a single module, and the efficiency of the inverter. The solar irradiance refers to the solar radiation power received per unit area. The solar irradiance is related to meteorological data such as temperature, cloud cover, and precipitation probability. The mapping relationship between meteorological data and solar irradiance can be pre-constructed based on historical solar irradiance and historical meteorological data (for example, a regression model can be constructed). On this basis, a data interface can be established with the meteorological department to obtain meteorological data for the forecast period. The meteorological data for the forecast period can be input into the regression model to obtain the solar irradiance for the forecast period.

[0030] Furthermore, by inputting the physical parameters of the photovoltaic power generation equipment and the solar irradiance during the forecast period into the physical model, the first photovoltaic power generation power of the industrial park during the forecast period can be predicted.

[0031] The prediction period can be several hours to seven days in the future, and the physical model can adopt the existing single diode model, whose calculation formula is:

[0032] ;

[0033] in, represents the first photovoltaic power generation power, Indicates the number of photovoltaic modules, Represents the area of ​​a single component, represents the solar irradiance during the forecast period, represents the component efficiency, Indicates the power temperature coefficient, usually 0.4% / ℃, represents the temperature of the photovoltaic cell, Indicates the inverter efficiency.

[0034] S102: Determine a first eigenvector based on fault prediction data of the photovoltaic power generation equipment during the prediction period and meteorological data of the industrial park during the prediction period.

[0035] In this embodiment, a fault prediction model can be pre-built based on a random forest model or a deep learning model (such as a convolutional neural network), and real-time operating data such as current, voltage, and temperature of the photovoltaic power generation equipment can be input into the fault prediction model to obtain fault prediction data of the photovoltaic power generation equipment in the prediction period (including fault type and corresponding fault level).

[0036] On this basis, considering that the fault state of photovoltaic power generation equipment will affect its power generation power, at the same time, changes in temperature, precipitation, atmospheric transparency, wind speed, etc. will also affect the output power of photovoltaic power generation equipment, therefore, the first photovoltaic power generation power output by the physical model can be corrected based on the fault prediction data of photovoltaic power generation equipment in the prediction period and the meteorological data of the industrial park in the prediction period to improve the accuracy of photovoltaic power prediction.

[0037] Specifically, the fault prediction data for photovoltaic power generation equipment during the forecast period and the weather data for the industrial park during the forecast period can be concatenated into a vector to obtain a first eigenvector, which provides a basis for subsequent correction of the first power generation capacity. For example, the first eigenvector can be [fault type, fault level, weather type].

[0038] It should be noted that, in this embodiment, step S102 may be executed after step S101 or before step S101. Therefore, the specific numbers of step S101 and step S102 do not limit the execution order of the two steps.

[0039] S103: Feature matching is performed on the first eigenvector with multiple second eigenvectors in the target strategy library to obtain multiple third eigenvectors; wherein each second eigenvector in the target strategy library has a corresponding first correction coefficient, and the third eigenvector is a vector among the multiple second eigenvectors that matches the first eigenvector.

[0040] In this embodiment, a target strategy library can be pre-built based on multiple sets of historical deviation data. Specifically, each set of historical deviation data includes historical fault data, historical meteorological data, and a historical correction coefficient. For any set of historical deviation data, a second eigenvector in the target strategy library can be built based on the historical fault data and historical meteorological data in the set of historical deviation data. At the same time, the historical correction coefficient in the set of historical deviation data is used as the first correction coefficient corresponding to the second eigenvector. The historical correction coefficient in each set of historical deviation data can be obtained by calculating the ratio between the historically predicted first photovoltaic power generation power and the actual historical photovoltaic power generation power under the operating conditions corresponding to the historical deviation data, where the operating conditions corresponding to the historical deviation data are determined by the historical fault data and historical meteorological data in the historical deviation data.

[0041] On this basis, the first feature vector can be feature matched with multiple second feature vectors in the target strategy library to select multiple third feature vectors that match the first feature vector. The multiple third feature vectors have a high degree of similarity with the first feature vector. Therefore, the second correction coefficient corresponding to the first feature vector can be determined based on the first correction coefficients corresponding to the multiple third feature vectors.

[0042] Specifically, the Euclidean distances between the first eigenvector and multiple second eigenvectors in the target strategy library can be calculated respectively. The closer the Euclidean distances are, the higher the similarity between the two vectors is. Therefore, the multiple second eigenvectors with the closest distances can be selected as the third eigenvector.

[0043] S104: Determine a second correction coefficient corresponding to the first eigenvector based on the first correction coefficients corresponding to the plurality of third eigenvectors, so as to correct the first photovoltaic power generation power based on the second correction coefficient to obtain a second photovoltaic power generation power.

[0044] In this embodiment, the average value of the first correction coefficients corresponding to multiple third eigenvectors can be determined as the second correction coefficient corresponding to the first eigenvector, or other methods can be used to calculate the second correction coefficient corresponding to the first eigenvector. See the following embodiments for details.

[0045] After the second correction coefficient is obtained, the second correction coefficient may be multiplied by the first photovoltaic power generation power to obtain the second photovoltaic power generation power.

[0046] S105: Dispatching and managing the distributed energy of the industrial park based on the load demand data corresponding to the industrial park in the forecast period and the second photovoltaic power generation power.

[0047] In this embodiment, the historical load data of the industrial park can be analyzed based on the Autoregressive Integrated Moving Average Model (ARIMA model) to capture its periodicity and trend, and then predict the load demand data corresponding to the industrial park in the forecast period. Then, with satisfying the load demand data as a constraint and minimizing the electricity purchase cost and the curtailment rate as the scheduling goal, the photovoltaic power generation equipment, energy storage system and power grid of the industrial park are dispatched and managed.

[0048] As can be seen from the above, this embodiment first predicts the first generated power of the industrial park during the prediction period based on the physical model. Then, a first eigenvector is determined based on the fault prediction data of the photovoltaic power generation equipment during the prediction period and the meteorological data of the industrial park during the prediction period. Through a similarity search of the eigenvectors, the most relevant second eigenvector is automatically selected from the target strategy library. The second correction coefficient corresponding to the first eigenvector is determined based on the first correction coefficient corresponding to the second eigenvector. Based on the second correction coefficient, the first generated power output by the physical model is accurately corrected in real time to obtain the second photovoltaic generated power. Distributed energy management of the industrial park based on the second photovoltaic generated power is conducive to improving energy utilization efficiency.

[0049] In one embodiment of the present application, the process of constructing the target policy library includes:

[0050] Acquire multiple sets of historical deviation data; each set of historical deviation data includes historical fault data, historical meteorological data, and historical correction coefficients;

[0051] Perform cluster analysis on multiple groups of historical deviation data to obtain cluster centers corresponding to multiple cluster areas;

[0052] Based on the size of the historical correction coefficient corresponding to each cluster center, multiple feature data points are selected from each cluster area;

[0053] Determine the cluster center and multiple feature data points of each cluster area as multiple target data points of the corresponding cluster area;

[0054] For each cluster region, determining the second eigenvector and the first correction coefficient corresponding to each target data point in the cluster region, thereby obtaining multiple second eigenvectors and multiple first correction coefficients for the cluster region; wherein any second eigenvector is determined based on historical fault data and historical meteorological data in the corresponding target data point, and any first correction coefficient is a historical correction coefficient in the corresponding target data point;

[0055] A target strategy library is constructed based on the multiple second eigenvectors and the multiple first correction coefficients of each cluster area.

[0056] In this embodiment, a K-means or DBSCAN clustering method can be used to perform cluster analysis on multiple sets of historical deviation data to obtain multiple cluster regions, each of which represents an operating condition of the photovoltaic power generation equipment. Each cluster region has different data distribution characteristics, and each cluster region can be distinguished based on the size of the historical correction coefficient corresponding to the cluster center of each cluster region. For example, if the historical correction coefficient corresponding to the cluster center of a cluster region is large, it indicates that the fault status of the photovoltaic power generation equipment and meteorological data have little impact on the output power of the photovoltaic power generation equipment (the impact of fault status and meteorological data on photovoltaic power generation equipment usually causes the actual photovoltaic power generation value to be less than the predicted value). The predicted and actual photovoltaic power generation values ​​corresponding to this cluster region are close. If the historical correction coefficient corresponding to the cluster center of a cluster region is small, it indicates that the predicted and actual photovoltaic power generation values ​​corresponding to this cluster region are significantly different.

[0057] Based on this, different methods can be used to select multiple feature data points from each cluster region. These multiple feature data points and the cluster center of each cluster region are used as representative target data points within that cluster region. For each target data point, a second feature vector is constructed based on the historical fault data and historical meteorological data contained in that target data point, and the historical correction coefficient of that target data point is used as the first correction coefficient corresponding to the second feature vector. Using this method, multiple second feature vectors and multiple first correction coefficients can be obtained, and then a target strategy library can be constructed based on these multiple second feature vectors and multiple first correction coefficients for each cluster region.

[0058] From the above, it can be concluded that this embodiment compresses a large amount of historical deviation data into a small number of representative target data points based on cluster analysis, and constructs a target decision library based on the target data points. This allows the target decision library to cover various working conditions of photovoltaic power generation equipment without causing an excessive amount of data.

[0059] In one embodiment of the present application, feature matching is performed on the first feature vector with multiple second feature vectors in the target strategy library to obtain multiple third feature vectors, including:

[0060] For each cluster region, respectively calculate the distance between the first eigenvector and each second eigenvector in the cluster region to obtain multiple first distances, and calculate the average of the multiple first distances to obtain the second distance corresponding to the cluster region;

[0061] Selecting a target cluster area from the plurality of cluster areas based on the second distances corresponding to the respective cluster areas;

[0062] The plurality of second eigenvectors of the target cluster region are determined as a plurality of third eigenvectors.

[0063] In this embodiment, each cluster region represents an operating condition of the photovoltaic power generation equipment. Therefore, the similarity between the first eigenvector and each cluster region can be determined first. When determining the similarity between the first eigenvector and any cluster region, the distances (such as Euclidean distances and cosine distances) between the first eigenvector and all second eigenvectors within the cluster region can be calculated to obtain multiple first distances. The average value (or weighted average value) of the multiple first distances is determined as the second distance, and the second distance is determined as the similarity between the first eigenvector and the cluster region. Using the above method, the similarity between the first eigenvector and each cluster region can be obtained, the cluster region with the greatest similarity is selected as the target cluster region, and the multiple second eigenvectors within the target cluster region are determined as multiple third eigenvectors.

[0064] From the above, it can be concluded that this embodiment first comprehensively considers the first distance corresponding to the second eigenvector of each clustering area to obtain the second distance, determines the target clustering area based on the second distance, and then determines multiple second eigenvectors based on the target clustering area, which can avoid strategy misselection caused by single vector deviation.

[0065] In one embodiment of the present application, for each cluster region, based on the size of the historical correction coefficient corresponding to each cluster center, multiple feature data points are selected from the cluster region, including:

[0066] If the historical correction coefficient corresponding to the cluster center of the cluster area is greater than a first threshold, multiple feature data points are selected from the cluster area based on the standard deviation of each data point in the cluster area;

[0067] If the historical correction coefficient corresponding to the cluster center of the cluster area is between the first threshold and the second threshold, all data points in the cluster area are secondary clustered, and the cluster centers determined by the secondary clustering are determined as multiple feature data points; the first threshold is greater than the second threshold;

[0068] If the historical correction coefficient corresponding to the cluster center of the cluster area is less than the second threshold, multiple edge points of the cluster area are selected based on the convex hull algorithm, and the multiple edge points are determined as multiple feature data points.

[0069] In this embodiment, for cluster regions where the historical correction coefficient corresponding to the cluster center is greater than the first threshold, the corresponding photovoltaic power generation equipment is operating normally. Under this operating condition, the fault level of the photovoltaic power generation equipment is 0 (indicating no faults), the weather type is sunny, and the historical deviation data is concentrated in the center of the cluster region, decreasing symmetrically toward both sides, with no obvious outliers (normal distribution). Based on the data distribution characteristics of this cluster region, multiple characteristic data points can be selected from the cluster region based on the standard deviation of each data point within the cluster region. Specifically, data points at a distance of one standard deviation from the cluster center are selected to represent the normal fluctuation range.

[0070] For cluster regions where the historical correction coefficient corresponding to the cluster center lies between the first and second thresholds, the corresponding photovoltaic power generation equipment is in an abnormal operating state. Under these operating conditions, the fault level of the photovoltaic power generation equipment is 1 (indicating a primary fault), or the weather type is cloudy or overcast. Under different fault types, the historical deviation data will form multiple clusters (with a multimodal distribution). Based on the data distribution characteristics of each cluster region, all data points in the cluster region can be secondary clustered. The cluster center determined by the secondary clustering is then used as multiple characteristic data points to obtain the first correction coefficient for each fault type.

[0071] For cluster regions where the historical correction coefficient corresponding to the cluster center is less than the second threshold, the corresponding photovoltaic power generation equipment is operating in an extreme state. Under these conditions, the fault level of the photovoltaic power generation equipment is 2 (indicating a secondary fault), or the weather type is extreme weather (including strong sandstorms, extremely high or low temperatures, etc.). The meteorological data under secondary fault conditions mostly shows normal weather conditions, with a few extreme weather conditions. Alternatively, the equipment failure data under extreme weather conditions mostly shows normal conditions, with a few secondary faults. Therefore, the historical deviation data may contain a small number of outliers. Based on the data distribution characteristics of this cluster region, multiple edge points of the cluster region are selected using the convex hull algorithm and identified as multiple feature data points as correction references for extreme conditions.

[0072] It should be noted that if the convex hull algorithm yields a large number of edge points, the edge of the cluster region can be segmented, and the point with the maximum curvature within each segment can be selected as the edge point of that segment. The first threshold and the second threshold are both preset constants, and those skilled in the art can flexibly design their specific values ​​based on actual needs.

[0073] From the above, it can be concluded that this embodiment uses different standards to determine multiple feature data points based on the data distribution characteristics of different clustering areas, which can avoid omission of key data points and thus more accurately characterize the data characteristics within the clustering area.

[0074] In one embodiment of the present application, before performing cluster analysis on multiple groups of historical deviation data, the method further includes:

[0075] Determine a detection threshold corresponding to each set of historical deviation data based on the historical fault data and the historical meteorological data in each set of historical deviation data;

[0076] Selecting target historical deviation data from the multiple sets of historical deviation data based on a detection threshold corresponding to each set of historical deviation data; wherein the historical correction coefficient included in the target historical deviation data is greater than or equal to the detection threshold;

[0077] Among them, cluster analysis is performed on multiple groups of historical deviation data, including:

[0078] Perform cluster analysis on target historical deviation data.

[0079] In this embodiment, when the power generated by photovoltaic power generation equipment does not match the power demand, it is necessary to forcibly limit the power output of the photovoltaic power generation equipment to maintain grid stability, which is the phenomenon of "power curtailment." In this case, the historical correction coefficients in the historical deviation data cannot truly reflect the errors predicted by the physical model. Therefore, before constructing the target strategy library based on multiple sets of historical deviation data, this embodiment first performs anomaly detection on the data in the multiple sets of historical deviation data and selects the target historical deviation data from them for use in constructing the target strategy library.

[0080] Specifically, considering that historical correction coefficients vary depending on historical fault data and historical meteorological data, for example, on sunny days with no faults, the historical correction coefficient is relatively large, close to 1. Alternatively, when a photovoltaic power generation system experiences a secondary fault and the weather is cloudy, the corresponding historical correction coefficient is relatively small, such as 0.6. Therefore, this embodiment determines the detection threshold corresponding to each set of historical deviation data based on the historical fault data and historical meteorological data in each set of historical deviation data, compares the historical correction coefficient in each set of historical deviation data with the corresponding detection threshold, and selects as target deviation data the sets of historical deviation data whose corresponding historical correction coefficients are less than the detection threshold.

[0081] From the above, it can be concluded that this embodiment determines the detection threshold corresponding to each group of historical deviation data based on the historical fault data and historical meteorological data in each group of historical deviation data, performs anomaly detection on the data in multiple groups of historical deviation data based on the detection threshold, and selects the target historical deviation data for the construction of the target strategy library. The obtained target strategy library can more accurately characterize the prediction error of the physical model.

[0082] In one embodiment of the present application, for any set of historical deviation data, determining a detection threshold corresponding to the set of historical deviation data based on historical fault data and historical meteorological data in the set of historical deviation data includes:

[0083] Obtaining a reference value for the detection threshold;

[0084] determining a first adjustment coefficient based on historical failure data in the set of historical deviation data;

[0085] determining a second adjustment coefficient based on the historical meteorological data in the set of historical deviation data;

[0086] The reference value of the detection threshold is adjusted based on the first adjustment coefficient and the second adjustment coefficient to obtain the detection threshold corresponding to the set of historical deviation data.

[0087] In this embodiment, the reference value of the detection threshold can be determined by the ratio between the historically predicted first photovoltaic power generation power and the actual historical photovoltaic power generation power under sunny and fault-free conditions. On this basis, for any set of historical deviation data, a first adjustment coefficient is determined based on the historical fault data in the set of historical deviation data. Specifically, a first mapping relationship between the fault level in the historical fault data and the first adjustment coefficient can be pre-established, where the higher the fault level, the smaller the first adjustment coefficient. At the same time, a second adjustment coefficient is determined based on the historical meteorological data in the set of historical deviation data. Specifically, a second mapping relationship between the weather type in the historical meteorological data and the second adjustment coefficient can be pre-established, where the second adjustment coefficient corresponding to a sunny day is greater than the second adjustment coefficient corresponding to a cloudy or overcast day, and the second adjustment coefficient corresponding to a cloudy or overcast day is greater than the second adjustment coefficient corresponding to extreme weather.

[0088] For any set of historical deviation data, after obtaining the first and second adjustment coefficients, the reference value of the detection threshold, the first and second adjustment coefficients can be multiplied together to obtain the detection threshold corresponding to that set of historical deviation data. Using the same method, the detection threshold corresponding to each set of historical deviation data can be obtained.

[0089] Specifically, the detection threshold can be calculated using the following first formula:

[0090] ;

[0091] in, represents the detection threshold, represents the reference value of the detection threshold, represents the first adjustment coefficient, represents the second adjustment coefficient. The first adjustment coefficient can be obtained by searching the first mapping relationship based on the fault type, and the second adjustment coefficient can be obtained by searching the second mapping relationship based on the weather type.

[0092] In one embodiment of the present application, determining the second correction coefficient corresponding to the first eigenvector based on the first correction coefficients corresponding to the plurality of third eigenvectors includes:

[0093] Based on the first distance corresponding to each third eigenvector, the first correction coefficients corresponding to multiple third eigenvectors are weighted and summed to obtain the second correction coefficient corresponding to the first eigenvector; wherein the weight of each first correction coefficient is negatively correlated with the corresponding first distance.

[0094] In this embodiment, the smaller the first distance corresponding to the third eigenvector, the higher the similarity between the third eigenvector and the first eigenvector. Therefore, based on the first distances corresponding to each third eigenvector, the first correction coefficients corresponding to multiple third eigenvectors are weighted and summed. This can make the first correction coefficients with high corresponding similarity have higher weights, which is conducive to obtaining a more accurate second correction coefficient.

[0095] Corresponding to the industrial park distributed energy management method of the above embodiment, Figure 2 This is a structural block diagram of an industrial park distributed energy management device provided by an embodiment of the present application. For ease of explanation, only the parts related to the embodiment of the present application are shown. Figure 2 The industrial park distributed energy management device 20 includes: a photovoltaic prediction module 21, a feature extraction module 22, a feature matching module 23, a parameter determination module 24 and a scheduling management module 25.

[0096] The photovoltaic prediction module 21 is used to predict the first photovoltaic power generation power of the industrial park in the prediction period using a physical model based on the physical parameters of the photovoltaic power generation equipment and the solar irradiance in the prediction period. The solar irradiance in the prediction period is obtained based on the meteorological data in the prediction period.

[0097] A feature extraction module 22 is configured to determine a first feature vector based on the fault prediction data of the photovoltaic power generation equipment during the prediction period and the meteorological data of the industrial park during the prediction period;

[0098] A feature matching module 23 is configured to perform feature matching on the first feature vector with a plurality of second feature vectors in the target strategy library to obtain a plurality of third feature vectors; wherein each second feature vector in the target strategy library has a corresponding first correction coefficient, and the third feature vector is a vector among the plurality of second feature vectors that matches the first feature vector;

[0099] a parameter determination module 24 for determining a second correction coefficient corresponding to the first eigenvector based on the first correction coefficients corresponding to the plurality of third eigenvectors, so as to correct the first photovoltaic power generation power based on the second correction coefficient to obtain a second photovoltaic power generation power;

[0100] The scheduling management module 25 is used to schedule and manage the distributed energy of the industrial park based on the load demand data corresponding to the industrial park in the forecast period and the second photovoltaic power generation power.

[0101] In one embodiment of the present application, the feature matching module 23 is specifically configured to:

[0102] Acquire multiple sets of historical deviation data; each set of historical deviation data includes historical fault data, historical meteorological data, and historical correction coefficients;

[0103] Perform cluster analysis on multiple groups of historical deviation data to obtain cluster centers corresponding to multiple cluster areas;

[0104] Based on the size of the historical correction coefficient corresponding to each cluster center, multiple feature data points are selected from each cluster area;

[0105] Determine the cluster center and multiple feature data points of each cluster area as multiple target data points of the corresponding cluster area;

[0106] For each cluster region, determining the second eigenvector and the first correction coefficient corresponding to each target data point in the cluster region, thereby obtaining multiple second eigenvectors and multiple first correction coefficients for the cluster region; wherein any second eigenvector is determined based on historical fault data and historical meteorological data in the corresponding target data point, and any first correction coefficient is a historical correction coefficient in the corresponding target data point;

[0107] A target strategy library is constructed based on the multiple second eigenvectors and the multiple first correction coefficients of each cluster area.

[0108] In one embodiment of the present application, the feature matching module 23 is further configured to:

[0109] For each cluster region, respectively calculate the distance between the first eigenvector and each second eigenvector in the cluster region to obtain multiple first distances, and calculate the average of the multiple first distances to obtain the second distance corresponding to the cluster region;

[0110] Selecting a target cluster area from the plurality of cluster areas based on the second distances corresponding to the respective cluster areas;

[0111] The plurality of second eigenvectors of the target cluster region are determined as a plurality of third eigenvectors.

[0112] In one embodiment of the present application, for each cluster region, the feature matching module 23 is further configured to:

[0113] If the historical correction coefficient corresponding to the cluster center of the cluster area is greater than a first threshold, multiple feature data points are selected from the cluster area based on the standard deviation of each data point in the cluster area;

[0114] If the historical correction coefficient corresponding to the cluster center of the cluster area is between the first threshold and the second threshold, all data points in the cluster area are secondary clustered, and the cluster centers determined by the secondary clustering are determined as multiple feature data points; the first threshold is greater than the second threshold;

[0115] If the historical correction coefficient corresponding to the cluster center of the cluster area is less than the second threshold, multiple edge points of the cluster area are selected based on the convex hull algorithm, and the multiple edge points are determined as multiple feature data points.

[0116] In one embodiment of the present application, before performing cluster analysis on multiple groups of historical deviation data, the feature matching module 23 is further configured to:

[0117] Determine a detection threshold corresponding to each set of historical deviation data based on the historical fault data and the historical meteorological data in each set of historical deviation data;

[0118] Selecting target historical deviation data from the multiple sets of historical deviation data based on a detection threshold corresponding to each set of historical deviation data; wherein the historical correction coefficient included in the target historical deviation data is greater than or equal to the detection threshold;

[0119] Among them, when performing cluster analysis on multiple groups of historical deviation data, the feature matching module 23 is specifically used to:

[0120] Perform cluster analysis on target historical deviation data.

[0121] In one embodiment of the present application, for any set of historical deviation data, the feature matching module 23 is further configured to:

[0122] Obtaining a reference value for the detection threshold;

[0123] determining a first adjustment coefficient based on historical failure data in the set of historical deviation data;

[0124] determining a second adjustment coefficient based on the historical meteorological data in the set of historical deviation data;

[0125] The reference value of the detection threshold is adjusted based on the first adjustment coefficient and the second adjustment coefficient to obtain the detection threshold corresponding to the set of historical deviation data.

[0126] In one embodiment of the present application, the parameter determination module 24 is specifically configured to:

[0127] Based on the first distance corresponding to each third eigenvector, the first correction coefficients corresponding to multiple third eigenvectors are weighted and summed to obtain the second correction coefficient corresponding to the first eigenvector; wherein the weight of each first correction coefficient is negatively correlated with the corresponding first distance.

[0128] The distributed energy management system for an industrial park provided in one embodiment of the present application includes a controller 300, photovoltaic power generation equipment and an energy storage system. Figure 3 , Figure 3 This is a schematic block diagram of a controller 300 provided in one embodiment of the present application. Figure 3 The industrial park controller 300 in the embodiment shown may include: one or more processors 301, one or more input devices 302, one or more output devices 303 and one or more memories 304. The processors 301, input devices 302, output devices 303 and memories 304 communicate with each other via a communication bus 305. The memory 304 is used to store computer programs, which include program instructions. The processor 301 is used to execute the program instructions stored in the memory 304. The processor 301 is configured to call the program instructions to execute the functions of each module / unit in the above-mentioned device embodiments, such as Figure 2 The functions of the photovoltaic prediction module 21, the feature extraction module 22 and the feature matching module 23, the parameter determination module 24 and the scheduling management module 25 are shown.

[0129] It should be understood that in the embodiment of the present application, the processor 301 may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc.

[0130] The input device 302 may include a touchpad, a fingerprint collection sensor (for collecting user fingerprint information and fingerprint direction information), a microphone, etc. The output device 303 may include a display (LCD, etc.), a speaker, etc.

[0131] The memory 304 may include a read-only memory and a random access memory, and provides instructions and data to the processor 301. A portion of the memory 304 may also include a non-volatile random access memory. For example, the memory 304 may also store preset constants such as a first threshold, a second threshold, and a third threshold.

[0132] In a specific implementation, the processor 301, input device 302, and output device 303 described in the embodiment of the present application can execute the implementation method described in the industrial park distributed energy management method provided in the embodiment of the present application, and can also execute the implementation method of the controller 300 described in the embodiment of the present application, which will not be repeated here.

[0133] In another embodiment of the present application, a computer-readable storage medium is provided. The computer-readable storage medium stores a computer program. The computer program includes program instructions. When the program instructions are executed by a processor, all or part of the process of the method in the above embodiment is implemented. The computer program can also be used to instruct related hardware to complete the process. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of each of the above method embodiments are implemented. The computer program includes computer program code, which can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium can include: any entity or device capable of carrying computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal and software distribution medium.

[0134] The computer-readable storage medium can be the internal storage unit of the industrial park distributed energy management system in any of the aforementioned embodiments, such as the hard disk or memory of the industrial park distributed energy management system. The computer-readable storage medium can also be an external storage device of the industrial park distributed energy management system, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a flash memory card, etc. Furthermore, the computer-readable storage medium can include both the internal storage unit of the industrial park distributed energy management system and an external storage device. The computer-readable storage medium is used to store computer programs and other programs and data required by the industrial park distributed energy management system. The computer-readable storage medium can also be used to temporarily store data that has been output or is about to be output.

[0135] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0136] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the industrial park distributed energy management system and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0137] In the several embodiments provided in this application, it should be understood that the disclosed industrial park distributed energy management system and method can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of units is only a logical function division. There may be other division methods in actual implementation, such as modules can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces or units, or it can be an electrical, mechanical or other form of connection.

[0138] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiments of the present application.

[0139] In addition, the functional modules in the various embodiments of the present application may be integrated into a processing module, or each module may exist physically separately, or two or more modules may be integrated into a single module. The above-mentioned integrated modules may be implemented in the form of hardware or software functional modules.

[0140] The above are only specific embodiments of the present application, but the scope of protection of the present application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and such modifications or substitutions should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A distributed energy management method for an industrial park, characterized in that: include: Based on the physical parameters of the photovoltaic power generation equipment and the solar irradiance during the forecast period, a physical model is used to predict the first photovoltaic power generation power of the industrial park during the forecast period, wherein the solar irradiance during the forecast period is obtained based on the meteorological data during the forecast period; Determining a first eigenvector based on fault prediction data of photovoltaic power generation equipment during a prediction period and meteorological data of the industrial park during the prediction period; Performing feature matching on the first feature vector and a plurality of second feature vectors in a target strategy library to obtain a plurality of third feature vectors; wherein each second feature vector in the target strategy library has a corresponding first correction coefficient, and the third feature vector is a vector among the plurality of second feature vectors that matches the first feature vector; determining a second correction coefficient corresponding to the first eigenvector based on first correction coefficients corresponding to a plurality of third eigenvectors, so as to correct the first photovoltaic power generation power based on the second correction coefficient to obtain a second photovoltaic power generation power; Distributed energy in the industrial park is dispatched and managed based on the load demand data corresponding to the industrial park in the forecast period and the second photovoltaic power generation power.

2. The distributed energy management method for an industrial park according to claim 1, characterized in that: The process of constructing the target policy library includes: Acquire multiple sets of historical deviation data; each set of historical deviation data includes historical fault data, historical meteorological data, and historical correction coefficients; Performing cluster analysis on the multiple groups of historical deviation data to obtain cluster centers corresponding to the multiple cluster areas; Based on the size of the historical correction coefficient corresponding to each cluster center, multiple feature data points are selected from each cluster area; Determine the cluster center and multiple feature data points of each cluster area as multiple target data points of the corresponding cluster area; For each cluster region, determining the second eigenvector and the first correction coefficient corresponding to each target data point in the cluster region, thereby obtaining multiple second eigenvectors and multiple first correction coefficients for the cluster region; wherein any second eigenvector is determined based on historical fault data and historical meteorological data in the corresponding target data point, and any first correction coefficient is a historical correction coefficient in the corresponding target data point; The target strategy library is constructed based on a plurality of second feature vectors and a plurality of first correction coefficients of each cluster area.

3. The distributed energy management method for an industrial park according to claim 2, characterized in that: The step of performing feature matching on the first feature vector and a plurality of second feature vectors in the target strategy library to obtain a plurality of third feature vectors includes: For each cluster region, respectively calculating the distance between the first eigenvector and each second eigenvector in the cluster region to obtain multiple first distances, and calculating the average of the multiple first distances to obtain the second distance corresponding to the cluster region; Selecting a target cluster area from the plurality of cluster areas based on the second distances corresponding to the respective cluster areas; The plurality of second eigenvectors of the target cluster region are determined as the plurality of third eigenvectors.

4. The distributed energy management method for an industrial park according to claim 2, characterized in that: For each cluster area, based on the size of the historical correction coefficient corresponding to each cluster center, multiple feature data points are selected from the cluster area, including: If the historical correction coefficient corresponding to the cluster center of the cluster area is greater than a first threshold, selecting the plurality of characteristic data points from the cluster area based on the standard deviation of each data point in the cluster area; If the historical correction coefficient corresponding to the cluster center of the cluster area is between a first threshold and a second threshold, performing secondary clustering on all data points in the cluster area, and determining the cluster center determined by the secondary clustering as the multiple feature data points; the first threshold is greater than the second threshold; If the historical correction coefficient corresponding to the cluster center of the cluster area is less than the second threshold, multiple edge points of the cluster area are selected based on the convex hull algorithm, and the multiple edge points are determined as the multiple feature data points.

5. The distributed energy management method for an industrial park according to claim 2, characterized in that: Before performing cluster analysis on the multiple groups of historical deviation data, the method further includes: Determine a detection threshold corresponding to each set of historical deviation data based on the historical fault data and the historical meteorological data in each set of historical deviation data; Based on the detection threshold corresponding to each set of historical deviation data, target historical deviation data is selected from the multiple sets of historical deviation data; wherein the historical correction coefficient included in the target historical deviation data is greater than or equal to the detection threshold; The cluster analysis of the multiple groups of historical deviation data includes: Perform cluster analysis on the target historical deviation data.

6. The distributed energy management method for an industrial park according to claim 5, characterized in that: For any set of historical deviation data, determining a detection threshold corresponding to the set of historical deviation data based on the historical fault data and the historical meteorological data in the set of historical deviation data includes: Obtaining a reference value for the detection threshold; determining a first adjustment coefficient based on historical fault data in the set of historical deviation data; determining a second adjustment coefficient based on the historical meteorological data in the set of historical deviation data; The reference value of the detection threshold is adjusted based on the first adjustment coefficient and the second adjustment coefficient to obtain the detection threshold corresponding to the set of historical deviation data.

7. The distributed energy management method for an industrial park according to claim 3, characterized in that: The determining the second correction coefficient corresponding to the first eigenvector based on the first correction coefficients corresponding to the plurality of third eigenvectors includes: Based on the first distance corresponding to each third eigenvector, the first correction coefficients corresponding to the multiple third eigenvectors are weighted and summed to obtain the second correction coefficient corresponding to the first eigenvector; wherein the weight of each first correction coefficient is negatively correlated with the corresponding first distance.

8. A distributed energy management device for an industrial park, characterized in that: include: a photovoltaic prediction module for predicting the first photovoltaic power generation power of the industrial park during the prediction period using a physical model based on the physical parameters of the photovoltaic power generation equipment and the solar irradiance during the prediction period, wherein the solar irradiance during the prediction period is obtained based on the meteorological data during the prediction period; a feature extraction module, configured to determine a first feature vector based on fault prediction data of the photovoltaic power generation equipment during a prediction period and meteorological data of the industrial park during the prediction period; a feature matching module, configured to perform feature matching on the first feature vector with a plurality of second feature vectors in a target strategy library to obtain a plurality of third feature vectors; wherein each second feature vector in the target strategy library has a corresponding first correction coefficient, and the third feature vector is a vector among the plurality of second feature vectors that matches the first feature vector; a parameter determination module, configured to determine a second correction coefficient corresponding to the first eigenvector based on first correction coefficients corresponding to the plurality of third eigenvectors, so as to correct the first photovoltaic power generation power based on the second correction coefficient to obtain a second photovoltaic power generation power; The scheduling management module is used to schedule and manage the distributed energy of the industrial park based on the load demand data corresponding to the industrial park in the forecast period and the second photovoltaic power generation power.

9. An industrial park distributed energy management system, comprising a controller, the controller comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images