Distributed photovoltaic power quality intelligent control system

Abstract

The invention discloses an intelligent power quality control system for distributed photovoltaic. The intelligent power quality control system comprises a data acquisition module, a self-adaptive clustering parameter acquisition module, a clustering detection module and an intelligent regulation and control module. The invention belongs to the field of electric power energy control, and particularly relates to an intelligent power quality control system for distributed photovoltaic. According to the scheme, time sequence smoothing processing is introduced, and instantaneous sudden change interference is restrained; the key electric energy quality state is highlighted through a weighted distance function, and the grouping judgment precision is improved; determining a maximum scale parameter by analyzing a spatial aggregation effect; by introducing a sample distribution stability constraint, analyzing neighbor distance distribution and selecting a minimum sample number parameter; a fuzzy logic mechanism and local density evaluation are introduced, for data in a critical state, the interference of local noise on overall clustering judgment is relieved by giving weights, the overall misjudgment risk caused by the local noise is relieved, and then the electric energy quality intelligent control effect is improved.

Description

Technical Field

[0001] The present invention relates to the field of electric energy control, and in particular to an intelligent power quality control system for distributed photovoltaics. Background Art

[0002] The intelligent power quality control system ensures the high quality and stable operation of the power grid through real-time data monitoring, intelligent analysis and automatic regulation, and prevents equipment damage or power supply interruption caused by power quality problems. However, general intelligent power quality control systems are easily affected by instantaneous interference, and distance measurement cannot highlight the key state, which will cause occasional short-term interference to be misjudged as abnormal state; general intelligent power quality control systems will increase the overall risk of misjudgment when improperly processing critical state noise data. Summary of the invention

[0003] In view of the above situation, in order to overcome the defects of the prior art, the present invention provides an intelligent power quality control system for distributed photovoltaics. In view of the problem that the general intelligent power quality control system is easily affected by instantaneous interference, the distance measurement cannot highlight the key state, and occasional short-term interference is misjudged as an abnormal state, the scheme introduces time series smoothing processing to suppress the interference of instantaneous mutations; through the weighted distance function, the key power quality state is highlighted and the determination accuracy of the grouping is improved; the maximum scale parameter is determined by analyzing the spatial aggregation effect of each monitoring node at different scales; by introducing the sample distribution stability constraint, the minimum sample number parameter is selected by analyzing the neighbor distance distribution in the feature space of the monitoring node; ensure that the real anomaly and noise are distinguished under different working conditions, thereby improving the accuracy of intelligent power quality control; in view of the problem that the general intelligent power quality control system will increase the overall misjudgment risk when the critical state noise data is improperly processed, the scheme introduces fuzzy logic mechanism and local density evaluation, for the data in the critical state, by assigning weights to alleviate the interference of local noise on the overall clustering judgment, alleviate the overall misjudgment risk caused by local noise, thereby improving the intelligent power quality control effect.

[0004] The technical solution adopted by the present invention is as follows: The present invention provides an intelligent power quality control system for distributed photovoltaics; including a data acquisition module, an adaptive clustering parameter acquisition module, a clustering detection module and an intelligent control module;

[0005] The data acquisition module collects historical power quality control data and working status; generates power quality characteristics through time series data smoothing;

[0006] The adaptive clustering parameter acquisition module adaptively determines the maximum scale and minimum sample number parameters required by DBSCAN by constructing a weighted distance function and statistical analysis based on spatial aggregation effect and neighbor distance distribution;

[0007] The cluster detection module uses the constructed power quality characteristics and the adaptively acquired parameters to perform clustering processing through DBSCAN, which refers to a density-based spatial clustering algorithm; fuzzy logic filtering is performed on the noise data in the cluster, and the corresponding working state cluster label is determined according to the cluster center;

[0008] The intelligent control module maps the pre-designed intelligent control strategy to generate specific control instructions according to the working state cluster label corresponding to the real-time collected data.

[0009] Furthermore, the data acquisition module uses the working state as a data label, and the data label is only used for cluster label selection; the collected data is subjected to feature engineering processing to obtain static power quality features; based on the static power quality features, time series data smoothing processing is introduced to construct power quality features with spatiotemporal characteristics, which are expressed as: ;in, is the power quality characteristic at time t after smoothing; It is the original power quality characteristics collected; is the smoothing parameter; t is the sampling time.

[0010] Furthermore, the adaptive clustering parameter acquisition module specifically includes:

[0011] Construct the Chebyshev distance function, expressed as: ;in, is the Chebyshev distance; and are the power quality feature vectors of the i-th monitoring node and the j-th monitoring node respectively; and are the voltage fluctuations of the ith monitoring node and the jth monitoring node respectively; and are the total harmonic distortion of the ith monitoring node and the jth monitoring node respectively; and are the flicker indicators of the i-th monitoring node and the j-th monitoring node respectively; and are the power factors of the ith monitoring node and the jth monitoring node respectively; and are the current indicators of the i-th monitoring node and the j-th monitoring node respectively; construct the Manhattan distance function , expressed as: ; Construct weighted distance function , highlighting the key power quality, can clearly distinguish the state of severe voltage flicker from the state of severe harmonic distortion; the weighted distance function is expressed as: ; Where a and b are distance weight parameters; L is the normalization factor;

[0012] Determine the maximum scale parameter; analyze the spatial aggregation effect between monitoring nodes based on power quality characteristics, and calculate the aggregation degree of adjacent feature points at different scales , taking the upper limit of the scale range that exhibits clustering characteristics as the maximum scale parameter, expressed as: ; ; Where n is the number of monitoring nodes; r is the analysis scale; is the monitoring node density estimate; It is the power quality difference determination;

[0013] Determine the minimum sample number parameter; analyze the distribution of neighbor distances in the feature space of each monitoring node, and convert the actual data With the assumption that the data is random based on the conditional space The histogram of is compared, and the sample distribution stability constraint is introduced. The histogram information content is selected The maximum k value is used as the minimum sample number parameter; the amount of histogram information is expressed as: ;in, is the amount of information in the histogram; is the mth histogram bin In, the probability density distribution of actual monitoring data under k-neighborhood conditions; is the mth histogram bin Inside, the probability density of the assumed data distribution based on conditional spatial randomness; N is the number of bins split in the histogram; is the stability penalty coefficient; It is the standard deviation of the distribution of neighbor distances in the feature space of the monitoring node under the selected k value; is the maximum standard deviation observed within the range of possible k values.

[0014] Furthermore, the cluster detection module specifically includes:

[0015] Clustering settings; clustering is performed based on DBSCAN according to the constructed power quality characteristics, maximum scale and minimum sample number parameters. Clustering is completed when clustering converges or reaches the maximum number of iterations. Different clusters composed of high-density areas are identified. Each cluster represents a working state. Cluster labels are selected based on the labels of cluster centers.

[0016] Noise filtering; adding fuzzy logic mechanism to assign weights to critical data; expressed as: ;in, It is the comprehensive local filtering factor after noise filtering of sample q in a given cluster Ms, which is used to measure the relative density of the current data point relative to its neighborhood; is the set of samples that belong to the same cluster as sample q; It is the weight assigned based on the fuzzy logic mechanism, reflecting the similarity between samples o and q; is the local reachable distance of the sample within the cluster;

[0017] Clustering determination: Determine the clustering effect based on the silhouette coefficient. If the clustering is not good, adjust the parameters and re-cluster.

[0018] Furthermore, after obtaining the clustering results, the intelligent control module performs data allocation on the power quality control data collected in real time, and uses the cluster label corresponding to the data as the working state corresponding to the data; each control cycle re-updates the clustering based on the existing historical data, calculates the current state characteristics of the real-time monitoring data, and determines the cluster category to which it belongs; based on the determination result, calls the control strategy mapping function corresponding to the cluster category, generates specific control instructions, and then issues dispatch commands to the distributed resources in the distribution network to optimize the power quality of the distribution network; the control strategy mapping function is expressed as: ;in, It is the pre-designed intelligent control strategy mapping function for the u-th working state; It is the active power output control instruction of the inverter; It is the reactive power output control instruction of the inverter; It is the active power dispatch of the energy storage system; It is the voltage regulation instruction; It is the frequency control instruction; It is a load classification dispatch instruction; It is the photovoltaic power reduction instruction.

[0019] The beneficial effects achieved by the present invention using the above scheme are as follows:

[0020] (1) In view of the problem that general power quality intelligent control systems are easily affected by instantaneous interference and distance measurement cannot highlight key states, which will cause occasional short-term interference to be misjudged as abnormal states, this scheme introduces time series smoothing processing to suppress the interference of instantaneous mutations; through the weighted distance function, the key power quality state is highlighted and the determination accuracy of the grouping is improved; the maximum scale parameter is determined by analyzing the spatial aggregation effect of each monitoring node at different scales; by introducing the sample distribution stability constraint, the minimum sample number parameter is selected by analyzing the distribution of neighbor distances in the feature space of the monitoring node; it ensures the distinction between real anomalies and noise under different working conditions, thereby improving the accuracy of power quality intelligent control.

[0021] (2) In order to solve the problem that the general power quality intelligent control system will increase the overall misjudgment risk when improperly processing critical state noise data, this scheme introduces fuzzy logic mechanism and local density evaluation. For data in a critical state, weights are assigned to mitigate the interference of local noise on the overall clustering judgment, mitigate the overall misjudgment risk caused by local noise, and thus improve the power quality intelligent control effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 A schematic diagram of a flow chart of an intelligent power quality control system for distributed photovoltaics provided by the present invention.

[0023] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention. DETAILED DESCRIPTION

[0024] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0025] In the description of the present invention, it should be understood that terms such as “upper”, “lower”, “front”, “back”, “left”, “right”, “top”, “bottom”, “inside” and “outside” indicating directions or positional relationships are based on the directions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore should not be understood as limiting the present invention.

[0026] Example 1, see Figure 1 , the present invention provides an intelligent power quality control system for distributed photovoltaics, including a data acquisition module, an adaptive clustering parameter acquisition module, a clustering detection module and an intelligent control module;

[0027] The data acquisition module collects historical power quality control data and working status; generates power quality characteristics through time series data smoothing; and sends the data to the adaptive clustering parameter acquisition module;

[0028] The adaptive clustering parameter acquisition module adaptively determines the maximum scale and minimum sample number parameters required by DBSCAN by constructing a weighted distance function and statistical analysis based on spatial aggregation effect and neighbor distance distribution; and sends the data to the cluster detection module;

[0029] The cluster detection module uses the constructed power quality characteristics and adaptively acquired parameters to perform clustering processing through DBSCAN, which refers to a density-based spatial clustering algorithm; performs fuzzy logic filtering on the noise data in the cluster, and determines the corresponding working state cluster label based on the cluster center; and sends the data to the intelligent control module;

[0030] The intelligent control module maps the pre-designed intelligent control strategy to generate specific control instructions according to the working status cluster label corresponding to the real-time collected data, and then issues dispatch commands to the distributed resources in the distribution network to optimize the power quality of the distribution network;.

[0031] Example 2, see Figure 1 , this embodiment is based on the above embodiment, the data acquisition module uses the working state as the data label, and the data label is only used for cluster label selection; in the distributed photovoltaic system, the power quality monitor is deployed on the inverter side, the distribution end and the key load; the historical power quality control data, including voltage, current, harmonics, flicker, power factor and frequency, is collected; the collected data is subjected to feature engineering processing to obtain static power quality characteristics; and on the basis of the static power quality characteristics, the time series data smoothing processing is introduced to construct the power quality characteristics with time and space characteristics, which is expressed as: ;in, is the power quality characteristic at time t after smoothing; It is the original power quality characteristics collected; is the smoothing parameter; t is the sampling time; based on time series smoothing, the interference of instantaneous mutations on the clustering results is suppressed to avoid misjudging instantaneous interference as an abnormal state.

[0032] Example 3, see Figure 1 , this embodiment is based on the above embodiment, and the adaptive clustering parameter acquisition module specifically includes:

[0033] Construct the Chebyshev distance function, expressed as: ;in, is the Chebyshev distance; and are the power quality feature vectors of the i-th monitoring node and the j-th monitoring node respectively; and are the voltage fluctuations of the ith monitoring node and the jth monitoring node respectively; and are the total harmonic distortion of the ith monitoring node and the jth monitoring node respectively; and are the flicker indicators of the i-th monitoring node and the j-th monitoring node respectively; and are the power factors of the ith monitoring node and the jth monitoring node respectively; and are the current indicators of the i-th monitoring node and the j-th monitoring node respectively; construct the Manhattan distance function , expressed as: ; Construct weighted distance function , highlighting the key power quality, can clearly distinguish the state of severe voltage flicker from the state of severe harmonic distortion; the weighted distance function is expressed as: ; Where a and b are distance weight parameters; L is the normalization factor;

[0034] Determine the maximum scale parameter; analyze the spatial aggregation effect between monitoring nodes based on power quality characteristics, and calculate the aggregation degree of adjacent feature points at different scales , taking the upper limit of the scale range that exhibits clustering characteristics as the maximum scale parameter to avoid over-clustering or too many isolated clusters, expressed as: ; ; Where n is the number of monitoring nodes; r is the analysis scale; is the monitoring node density estimate; It is the power quality difference determination;

[0035] Determine the minimum sample number parameter; by analyzing the distribution of neighbor distances in the feature space of each monitoring node, the actual data With the assumption that the data is random based on the conditional space The histogram of is compared, and the sample distribution stability constraint is introduced. The histogram information content is selected The maximum k value is used as the minimum sample number parameter; the amount of histogram information is expressed as: ;in, is the amount of information in the histogram; is the mth histogram bin In, the probability density distribution of actual monitoring data under k-neighborhood conditions; is the mth histogram bin Inside, the probability density of the assumed data distribution based on conditional spatial randomness; N is the number of bins split in the histogram; is the stability penalty coefficient; It is the standard deviation of the distribution of neighbor distances in the feature space of the monitoring node under the selected k value; It is the maximum standard deviation observed within the range of k values ​​to be selected; it automatically identifies the cluster boundaries under different working conditions to avoid misjudging power quality anomalies as noise due to improper settings.

[0036] By performing the above operations, in view of the problem that general power quality intelligent control systems are easily affected by instantaneous interference, and distance measurement cannot highlight key states, which will cause occasional short-term interference to be misjudged as abnormal states, this scheme introduces time series smoothing processing to suppress the interference of instantaneous mutations; through the weighted distance function, the key power quality states are highlighted and the determination accuracy of grouping is improved; the maximum scale parameter is determined by analyzing the spatial aggregation effect of each monitoring node at different scales; by introducing the sample distribution stability constraint, the minimum sample number parameter is selected by analyzing the neighbor distance distribution in the feature space of the monitoring node; ensure the distinction between real anomalies and noise under different working conditions, thereby improving the accuracy of intelligent power quality control.

[0037] Example 4, see Figure 1 , this embodiment is based on the above embodiment, and the cluster detection module specifically includes:

[0038] Clustering settings; clustering is performed based on DBSCAN according to the constructed power quality characteristics, maximum scale and minimum sample number parameters. Clustering is completed when clustering converges or reaches the maximum number of iterations. Different clusters composed of high-density areas are identified. Each cluster represents a working state. Cluster labels are selected based on the labels of cluster centers.

[0039] Noise filtering: remove noise data mixed into clustering, add fuzzy logic mechanism, give weight to critical data instead of simple binary elimination, and alleviate the risk of misjudging the overall clustering due to partial data noise; expressed as: ;in, It is the comprehensive local filtering factor after noise filtering of sample q in a given cluster Ms, which is used to measure the relative density of the current data point relative to its neighborhood; is the set of samples that belong to the same cluster as sample q; It is the weight assigned based on the fuzzy logic mechanism, reflecting the similarity between samples o and q; It is the local reachable distance of the sample within the cluster; a high comprehensive local filter factor indicates that the local density of the data point is significantly lower than the average density of other points in the neighborhood, which may be a boundary point or noise; a local threshold is set, and data points with a comprehensive local filter factor lower than the local threshold are regarded as noise points and ignored;

[0040] Clustering determination: Determine the clustering effect based on the silhouette coefficient. If the clustering is not good, adjust the parameters and re-cluster.

[0041] Example 5, see Figure 1, this embodiment is based on the above embodiment. After obtaining the clustering result, the intelligent control module distributes the power quality control data collected in real time, and uses the cluster label corresponding to the data as the working state corresponding to the data. The system will match the pre-designed intelligent control strategy for each working state, and issue control instructions in real time to form a closed-loop control; each control cycle re-updates the clustering based on the existing historical data, calculates the current state characteristics of the real-time monitoring data, and determines the cluster category to which it belongs; according to the determination result, the control strategy mapping function corresponding to the cluster category is called to generate specific control instructions, and then the dispatching command is issued to the distributed resources in the distribution network to optimize the power quality of the distribution network; the control strategy mapping function is expressed as: ;in, It is the pre-designed intelligent control strategy mapping function for the u-th working state; It is the active power output control instruction of the inverter; It is the reactive power output control instruction of the inverter; It is the active power dispatch of the energy storage system; It is the voltage regulation instruction; It is the frequency control instruction; It is a load classification dispatch instruction; It is the photovoltaic power reduction instruction.

[0042] By performing the above operations, in order to solve the problem that the general power quality intelligent control system will increase the overall misjudgment risk when improperly processing critical state noise data, this scheme introduces fuzzy logic mechanism and local density evaluation. For data in a critical state, weights are assigned to alleviate the interference of local noise on the overall clustering judgment, alleviate the overall misjudgment risk caused by local noise, and thus improve the power quality intelligent control effect.

[0043] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0044] While the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that many changes, modifications, substitutions and variations can be made to the embodiments without departing from the principles and spirit of the invention.

[0045] The present invention and its embodiments are described above, and such description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if ordinary technicians in the field are inspired by it, without departing from the purpose of the invention, they can design a structure and embodiment similar to the technical solution without creativity, which should belong to the protection scope of the present invention.

Claims

1. An intelligent power quality control system for distributed photovoltaics, characterized in that: The system includes a data acquisition module, an adaptive clustering parameter acquisition module, a clustering detection module and an intelligent control module; The data acquisition module collects historical power quality control data and working status; generates power quality characteristics through time series data smoothing; The adaptive clustering parameter acquisition module adaptively determines the maximum scale and minimum sample number parameters required by DBSCAN by constructing a weighted distance function and statistical analysis based on spatial aggregation effect and neighbor distance distribution; The cluster detection module uses the constructed power quality characteristics and adaptively acquired parameters to perform clustering processing through DBSCAN, performs fuzzy logic filtering on the noise data in the cluster, and determines the corresponding working state cluster label according to the cluster center; The intelligent control module maps the pre-designed intelligent control strategy to generate specific control instructions according to the working status cluster label corresponding to the real-time collected data, and then issues dispatch commands to the distributed resources in the distribution network to optimize the power quality of the distribution network.

2. According to claim 1, an intelligent power quality control system for distributed photovoltaics is characterized by: The data acquisition module uses the working state as a data label, and the data label is only used for cluster label selection; the collected data is subjected to feature engineering processing to obtain static power quality features; based on the static power quality features, time series data smoothing processing is introduced to construct power quality features with spatiotemporal characteristics, which are expressed as: ;in, is the power quality characteristic at time t after smoothing; It is the original power quality characteristics collected; is the smoothing parameter; t is the sampling time.

3. The intelligent power quality control system for distributed photovoltaic according to claim 2, characterized in that: The adaptive clustering parameter acquisition module specifically includes: Construct the Chebyshev distance function, expressed as: ;in, is the Chebyshev distance; and are the power quality feature vectors of the i-th monitoring node and the j-th monitoring node respectively; and are the voltage fluctuations of the ith monitoring node and the jth monitoring node respectively; and are the total harmonic distortion of the ith monitoring node and the jth monitoring node respectively; and are the flicker indicators of the i-th monitoring node and the j-th monitoring node respectively; and are the power factors of the ith monitoring node and the jth monitoring node respectively; and are the current indicators of the i-th monitoring node and the j-th monitoring node respectively; construct the Manhattan distance function , expressed as: ; Construct weighted distance function , highlighting the key power quality, can clearly distinguish the state of severe voltage flicker from the state of severe harmonic distortion; the weighted distance function is expressed as: ; Where a and b are distance weight parameters; L is the normalization factor; Determine the maximum scale parameter; Determine the minimum number of samples parameter.

4. The intelligent power quality control system for distributed photovoltaic according to claim 3, characterized in that: In the adaptive clustering parameter acquisition module, the maximum scale parameter is determined by analyzing the spatial aggregation effect between each monitoring node based on the power quality characteristics, and calculating the aggregation degree of adjacent feature points at different scales. , taking the upper limit of the scale range that exhibits clustering characteristics as the maximum scale parameter, expressed as: ; ; Where n is the number of monitoring nodes; r is the analysis scale; is the monitoring node density estimate; It is the power quality difference judgment.

5. The intelligent power quality control system for distributed photovoltaic according to claim 4, characterized in that: In the adaptive clustering parameter acquisition module, the minimum sample number parameter is determined by analyzing the distribution of neighbor distances in the feature space of each monitoring node, and the actual data With the assumption that the data is random based on the conditional space The histogram of is compared, and the sample distribution stability constraint is introduced. The histogram information content is selected The maximum k value is used as the minimum sample number parameter; the amount of histogram information is expressed as: ;in, is the amount of information in the histogram; is the mth histogram bin In, the probability density distribution of actual monitoring data under k-neighborhood conditions; is the mth histogram bin Inside, the probability density of the assumed data distribution based on conditional spatial randomness; N is the number of bins split in the histogram; is the stability penalty coefficient; It is the standard deviation of the distribution of neighbor distances in the feature space of the monitoring node under the selected k value; is the maximum standard deviation observed within the range of possible k values.

6. The intelligent power quality control system for distributed photovoltaic power generation according to claim 5, characterized in that: The cluster detection module specifically includes: Clustering settings; clustering is performed based on DBSCAN according to the constructed power quality characteristics, maximum scale and minimum sample number parameters; clustering is completed when clustering converges or reaches the maximum number of iterations; different clusters composed of high-density areas are identified, each cluster represents a working state, and cluster labels are selected based on the labels of cluster centers; Noise filtering; adding fuzzy logic mechanism to assign weights to critical data; expressed as: ;in, It is the comprehensive local filtering factor after noise filtering of sample q in a given cluster Ms, which is used to measure the relative density of the current data point relative to its neighborhood; is the set of samples that belong to the same cluster as sample q; It is the weight assigned based on the fuzzy logic mechanism, reflecting the similarity between samples o and q; is the local reachable distance of the sample within the cluster; Clustering determination: Determine the clustering effect based on the silhouette coefficient. If the clustering is not good, adjust the parameters and re-cluster.

7. The intelligent power quality control system for distributed photovoltaic power generation according to claim 6, characterized in that: The intelligent control module distributes the power quality control data collected in real time after obtaining the clustering results, and uses the cluster label corresponding to the data as the working state corresponding to the data; each control cycle re-updates the clustering based on the existing historical data, calculates the current state characteristics of the real-time monitoring data, and determines the cluster category to which it belongs; According to the judgment results, the control strategy mapping function of the corresponding clustering category is called to generate specific control instructions, and then dispatch commands are issued to the distributed resources in the distribution network to optimize the power quality of the distribution network; the control strategy mapping function is expressed as: ;in, It is the pre-designed intelligent control strategy mapping function for the u-th working state; It is the active power output control instruction of the inverter; It is the reactive power output control instruction of the inverter; It is the active power dispatch of the energy storage system; It is the voltage regulation instruction; It is the frequency control instruction; It is a load classification dispatch instruction; It is the photovoltaic power reduction instruction.

Images