Rectifier intelligent control method and system suitable for power electronic equipment

Abstract

The invention discloses a rectifier intelligent control method and system suitable for power electronic equipment, and relates to the technical field of rectifier control. The method comprises the following steps: configuring a rectifier for target equipment; performing demand analysis on the equipment information of the target equipment to obtain a target demand identifier; performing cluster centralized analysis according to the target demand identifier to obtain a target device cluster; respectively inputting the equipment information of the target equipment in the target equipment cluster into a demand control unit for mapping learning, and generating a demand control subunit set; monitoring and analyzing the target equipment cluster to generate a monitoring data cluster; performing equilibrium deviation analysis on the monitoring data set in the monitoring data cluster to generate a deviation monitoring data cluster; and performing mapping intelligent control on the corresponding rectifiers by using the demand control subunit set. The technical problem of rectifier control deviation caused by insufficient analysis of different application scenes in the prior art is solved, and the technical effect of eliminating rectifier control deviation is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of rectifier control, and in particular to a rectifier intelligent control method and system suitable for power electronic equipment. Background Art

[0002] Power electronic equipment will face various complex operating environments and conditions in different application scenarios, which will have different effects on the performance of the rectifier. For example, in the field of industrial production, power electronic equipment usually needs to operate for a long time and at high load. In this case, the rectifier needs to have good heat dissipation performance and stable current output capability to ensure the continuous and stable operation of the equipment. In addition, in some special application scenarios, such as aerospace, medical equipment and other fields, the rectifier needs to have high-precision current and voltage control capabilities, as well as extremely high reliability and stability to ensure the normal operation and safety of the equipment. There is a technical problem in the prior art that insufficient analysis of different application scenarios leads to deviations in rectifier control. Summary of the invention

[0003] The embodiments of the present application provide a rectifier intelligent control method and system applicable to power electronic equipment, which solves the technical problem in the prior art that insufficient analysis of different application scenarios leads to deviations in rectifier control.

[0004] In view of the above problems, an embodiment of the present application provides a rectifier intelligent control method and system applicable to power electronic equipment.

[0005] A first aspect of an embodiment of the present application provides a rectifier intelligent control method applicable to a power electronic device, the method comprising: The interactive rectifier intelligent control platform extracts N target devices in the target control area, and configures N rectifiers for the N target devices, wherein the N rectifiers have N position coordinate identifiers; Performing demand analysis on N device information of the N target devices to obtain N target demand identifiers; Performing clustering and centralized analysis according to the N target demand identifiers to obtain M target device clusters, wherein the M target device clusters have M target centralized demand identifiers; Based on the M target centralized demand identifiers, M demand control units in the rectifier intelligent control platform are called, and device information of target devices in the M target device clusters is respectively input into the M demand control units for mapping learning to generate M demand control sub-unit sets; Traversing the M demand control subunit sets to perform resource configuration, and using the configured M demand control subunit sets to monitor and analyze the M target device clusters to generate M monitoring data clusters, wherein each monitoring data cluster includes multiple monitoring data sets; Performing a balanced deviation analysis on the monitoring data sets in the M monitoring data clusters to generate M deviation monitoring data clusters; Based on the M deviation monitoring data clusters, the M demand control sub-unit sets are utilized to perform mapping intelligent control on the corresponding rectifiers.

[0006] A second aspect of the embodiment of the present application provides a rectifier intelligent control system applicable to power electronic equipment, the system comprising: An extraction module, the extraction module is used to interact with the rectifier intelligent control platform to extract N target devices in a target control area, and configure N rectifiers for the N target devices, wherein the N rectifiers have N position coordinate identifiers; A demand analysis module, the demand analysis module is used to perform demand analysis on N device information of the N target devices to obtain N target demand identifiers; A cluster analysis module, the cluster analysis module is used to perform cluster centralized analysis according to the N target demand identifiers to obtain M target device clusters, wherein the M target device clusters have M target centralized demand identifiers; A mapping learning module, wherein the mapping learning module is used to call M demand control units in the rectifier intelligent control platform based on the M target centralized demand identifiers, input device information of target devices in the M target device clusters into the M demand control units for mapping learning, and generate M demand control sub-unit sets; A monitoring and analysis module, the monitoring and analysis module is used to traverse the M demand control subunit sets to perform resource configuration, and use the configured M demand control subunit sets to monitor and analyze the M target device clusters to generate M monitoring data clusters, wherein each monitoring data cluster includes multiple monitoring data sets; A deviation analysis module, the deviation analysis module is used to perform a balanced deviation analysis on the monitoring data sets in the M monitoring data clusters to generate M deviation monitoring data clusters; A control module is used to map and intelligently control corresponding rectifiers based on the M deviation monitoring data clusters and using the M demand control sub-unit sets.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages: The interactive rectifier intelligent control platform extracts N target devices in the target control area, configures N rectifiers for the N target devices, wherein the N rectifiers have N position coordinate identifiers. Then, the demand analysis is performed on the N device information of the N target devices to obtain N target demand identifiers. Next, clustering and centralized analysis is performed based on the N target demand identifiers to obtain M target device clusters, wherein the M target device clusters have M target centralized demand identifiers. Based on the M target centralized demand identifiers, the M demand control units in the rectifier intelligent control platform are called, and the device information of the target devices in the M target device clusters is respectively input into the M demand control units for mapping learning to generate M demand control sub-unit sets. The M demand control sub-unit sets are traversed to perform resource configuration, and the M target device clusters are monitored and analyzed using the configured M demand control sub-unit sets to generate M monitoring data clusters, wherein each monitoring data cluster includes multiple monitoring data sets. Balanced deviation analysis is performed on the monitoring data sets in the M monitoring data clusters to generate M deviation monitoring data clusters. Finally, based on M deviation monitoring data clusters, M demand control subunit sets are used to map and intelligently control the corresponding rectifiers. This solves the technical problem of rectifier control deviation caused by insufficient analysis of different application scenarios in the prior art, and achieves the technical effect of eliminating rectifier control deviation. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0009] Figure 1 A schematic flow chart of a rectifier intelligent control method applicable to power electronic equipment provided in an embodiment of the present application; Figure 2 A schematic diagram of the structure of a rectifier intelligent control system suitable for power electronic equipment provided in an embodiment of the present application.

[0010] Explanation of the accompanying drawings: extraction module 11, demand analysis module 12, clustering analysis module 13, mapping learning module 14, monitoring analysis module 15, deviation analysis module 16, control module 17. DETAILED DESCRIPTION

[0011] The embodiments of the present application solve the technical problem in the prior art that insufficient analysis of different application scenarios leads to deviations in rectifier control by providing a rectifier intelligent control method and system suitable for power electronic equipment.

[0012] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0013] It should be noted that the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules that are not explicitly listed or are inherent to these processes, methods, products or devices.

[0014] Embodiment 1 like Figure 1 As shown, an embodiment of the present application provides a rectifier intelligent control method applicable to power electronic equipment, and the method is applied to a rectifier intelligent control platform, wherein the method includes: Interacting with the rectifier intelligent control platform to extract N target devices in a target control area, and configuring N rectifiers for the N target devices, wherein the N rectifiers have N position coordinate identifiers; The rectifier intelligent control platform is used to control the rectifier. N target devices in the target control area are extracted through the interactive rectifier intelligent control platform, and N rectifiers are configured for the N target devices. Among them, the N rectifiers have N position coordinate identifiers, and the position coordinate identifiers reflect the specific positions of the rectifiers in the target control area.

[0015] Performing demand analysis on N device information of the N target devices to obtain N target demand identifiers; Perform demand analysis on N device information of N target devices to obtain N target demand identifiers, wherein each target device corresponds to a different application scenario and thus corresponds to a different target demand identifier, and the target demand identifier reflects the actual operation requirements and characteristics of the device.

[0016] Furthermore, performing demand analysis on N device information of the N target devices to obtain N target demand identifiers includes: Traversing the N target devices to extract information and generate N device information, wherein the N device information includes N device models, N device working fields and N device load fluctuation information; Based on the N device models, the N device working fields and the N device load fluctuation information, matching is performed in the demand analysis space to obtain N matching demand sets; The first n requirements in the N matching requirement sets are extracted respectively to generate the N target requirement identifiers.

[0017] By traversing N target devices, N device information is extracted, including N device models, N device working fields, and N device load fluctuation information. The device model reflects the basic properties and performance characteristics of the device, the working field reveals the application scenario and environmental conditions of the device, and the load fluctuation information is directly related to the power demand and operating stability of the device. Based on the matching of N device models, N device working fields, and N device load fluctuation information in the demand analysis space, the specific requirements of each device in a specific working scenario can be accurately identified, and the corresponding matching requirement set can be generated, and finally N matching requirement sets are obtained. The top n requirements in the N matching requirement sets are extracted respectively to generate N target requirement identifiers, among which the top n requirements usually represent the main requirements.

[0018] Further, the method comprises: Collecting multiple sample device information and multiple sample demand identifiers, wherein the multiple sample device information includes multiple sample device models, multiple sample device working fields, and multiple sample device load fluctuation information; Constructing a plurality of sample points in the demand analysis space based on the plurality of sample device information including a plurality of sample device models, a plurality of sample device working fields and a plurality of sample device load fluctuation information, wherein the demand analysis space is a three-dimensional space; Inputting the N equipment models, the N equipment working fields and the N equipment load fluctuation information into the demand analysis space to obtain multiple target points; Taking the multiple target points as the center respectively, the m sample demand identifiers of the m sample points whose distances to the multiple target points in the demand analysis space are in the first m positions are summarized to generate the N matching demand sets, where m is an integer greater than or equal to 3.

[0019] The multiple sample device information includes multiple sample device models, multiple sample device working fields and multiple sample device load fluctuation information, which reflects the actual operating conditions and requirements of different devices in different application scenarios. Through the collected multiple sample device information, multiple sample points in the demand analysis space are constructed, wherein the demand analysis space is a three-dimensional space, and each dimension corresponds to the device model, working field and load fluctuation information. The N device models, N device working fields and N device load fluctuation information of the N target devices are input into the demand analysis space to obtain multiple target points, which represent the position of the target device in the demand analysis space and reflect the unique requirements and characteristics of each device. With each target point as the center, the distance to the target point in the demand analysis space is calculated, and the m sample demand identifiers of the m sample points with the first m distances are selected for aggregation to generate N matching demand sets, wherein m is an integer greater than or equal to 3.

[0020] Performing clustering and centralized analysis according to the N target demand identifiers to obtain M target device clusters, wherein the M target device clusters have M target centralized demand identifiers; The N target demand identifiers are clustered and analyzed. Specifically, clustering algorithms such as K-means and hierarchical clustering are used to process the N target demand identifiers. That is, the devices are divided into different clusters according to the similarities and differences of the device requirements to obtain M target device clusters. The devices in the target device cluster have similar demand characteristics, so that a unified control strategy can be formulated for each target device cluster. At the same time, each target device cluster has a target centralized demand identifier, which reflects the common requirements of the devices in the device cluster. By forming target device clusters and target centralized demand identifiers, the centralized requirements of devices can be more accurately identified and managed.

[0021] Based on the M target centralized demand identifiers, M demand control units in the rectifier intelligent control platform are called, and device information of target devices in the M target device clusters is respectively input into the M demand control units for mapping learning to generate M demand control sub-unit sets; After obtaining M target set demand identifiers, the rectifier intelligent control platform will use these M target set demand identifiers to call the corresponding M demand control units. Each target set demand identifier corresponds to a specific demand control unit. The demand control unit refers to a pre-trained unit that meets the target set demand identifier. The pre-trained unit is fine-tuned according to the actual situation of different target devices to obtain each demand control sub-unit set. The device information of the target devices in the M target device clusters is input into the M demand control units for mapping learning. That is to say, the training data retrieved from the big data by the target device is used to perform fine training on the demand control units respectively to generate demand control sub-units that meet the requirements of each target device. Through mapping learning, M demand control sub-unit sets are generated. Each set contains a set of control sub-units for a specific device cluster. These sub-units have the ability to handle the requirements of the device cluster.

[0022] Further, based on the M target centralized demand identifiers, calling the M demand control units in the rectifier intelligent control platform, the method includes: Acquire multiple sample control unit parameter sets, multiple monitoring data sets, and multiple rectifier control parameter sets as multiple training data sets, respectively, and pre-train a framework built based on a convolutional neural network, wherein the multiple rectifier control parameter sets have multiple sample set requirement identifiers; Use the pre-training loss function to perform loss certification on the pre-training process, and update and adjust multiple model parameters based on the loss certification results; After multiple updates, when the loss authentication result meets the requirements and / or the number of iterations meets the preset number of iterations, multiple demand control units of the rectifier intelligent control platform are generated with updated adjustment results; Using the M target concentrated demand identifiers as indexes, searching is performed among multiple demand control units of the rectifier intelligent control platform to obtain the M demand control units.

[0023] The monitoring data set contains various real-time data during the operation of the equipment, reflecting the actual operation status of the equipment, and the rectifier control parameter set represents the rectifier control parameters. Optionally, multiple sample control unit parameter sets, monitoring data sets, and rectifier control parameter sets are obtained as training data sets to pre-train the framework based on the convolutional neural network. In the pre-training process, the pre-training loss function is used to perform loss certification on the training process. The loss function is an indicator that measures the difference between the model prediction value and the true value. By minimizing the loss function, the parameters of the model can be continuously optimized to make it closer to the actual situation. In each iteration, the model parameters are updated and adjusted according to the loss certification results to gradually improve the performance of the model. After multiple iterative updates, when the loss certification results meet the preset requirements or the number of iterations reaches the preset number of iterations, it is considered that the model has been fully learned and optimized. At this time, multiple demand control units of the rectifier intelligent control platform are generated according to the update and adjustment results. Using the M target set demand identifiers as indexes, multiple demand control units of the rectifier intelligent control platform are searched to quickly and accurately find the demand control unit corresponding to each target device cluster.

[0024] Furthermore, the pre-training loss function is: ; in, is the loss value, To monitor data collection, is the set of rectifier control parameters, is the learning rate in the model, , is an empirical parameter, is the verification rectifier control parameter set output by the model, is the weight.

[0025] The pre-training loss function is used to measure the difference between the model's predicted results and the actual results, and to guide the model's parameter update and optimization accordingly. is the loss value, To monitor data collection, is the set of rectifier control parameters, is the learning rate in the model, , is an empirical parameter, is the verification rectifier control parameter set output by the model, is the weight. As the learning rate in the model, it determines the step size of the model parameter update in each iteration. , It is an empirical parameter and is set according to the actual application scenario and data characteristics. The model is based on the input data Predicted control parameters. As weights, it is used to adjust the importance of different items in the loss function by adjusting The value of can be used to fine-tune the emphasis of the loss function to suit a specific optimization goal.

[0026] Traversing the M demand control subunit sets to perform resource configuration, and using the configured M demand control subunit sets to monitor and analyze the M target device clusters to generate M monitoring data clusters, wherein each monitoring data cluster includes multiple monitoring data sets; After obtaining M demand control subunit sets, the rectifier intelligent control platform will configure resources for these M demand control subunit sets. Resource configuration involves the reasonable allocation of computing resources, storage resources and communication resources to provide sufficient support for the demand control subunit sets to ensure that the monitoring and analysis tasks can be performed efficiently and stably. After completing the resource configuration, the M configured demand control subunit sets will be used to monitor and analyze the corresponding M target device clusters. The monitoring and analysis mainly collects the operating status data of the equipment in real time, including key indicators such as voltage, current, power, and temperature. During the monitoring and analysis process, M monitoring data clusters are generated, each of which corresponds to a target device cluster. Each monitoring data cluster consists of multiple monitoring data sets, which contain various monitoring data collected from the target device cluster. These monitoring data reflect the real-time operating status of the equipment.

[0027] Performing a balanced deviation analysis on the monitoring data sets in the M monitoring data clusters to generate M deviation monitoring data clusters; Balanced deviation analysis is performed on the monitoring data sets in M ​​monitoring data clusters to generate M deviation monitoring data clusters. Each data point in the M deviation monitoring data clusters reflects the degree of deviation in the original monitoring data, and the differences between different data sets are eliminated after balanced processing. Balanced deviation analysis aims to identify and quantify anomalies or deviation trends in monitoring data in order to obtain a more accurate assessment of the operating status and performance of the target device cluster.

[0028] Furthermore, performing balanced deviation analysis on the monitoring data sets in the M monitoring data clusters to generate M deviation monitoring data clusters, the method includes: Extracting a first monitoring data cluster from the M monitoring data clusters, and then extracting a first monitoring data set from the first monitoring data cluster; Performing mean calculation on the first monitoring data set to generate a first monitoring data mean; Taking multiple mean straight lines in any direction passing through the mean of the first monitoring data as the midpoint of the region, diffusing to both sides of the straight line according to the preset balanced deviation step, counting the amount of data in multiple diffusion areas, and retaining the mean straight line corresponding to the maximum amount of data as the target mean straight line; Taking the target mean straight line as a starting point, performing a balanced deviation analysis in the first monitoring data set to generate first deviation monitoring data; A balanced deviation analysis is performed on the M monitoring data clusters to generate M deviation monitoring data clusters.

[0029] A monitoring data cluster is extracted from the M monitoring data clusters as the first monitoring data cluster, and a monitoring data set is extracted from the first monitoring data cluster as the first monitoring data set. The first monitoring data set is averaged to obtain the first monitoring data mean, which represents the central trend of the data set. A plurality of mean straight lines in any direction passing through the first monitoring data mean are used as the regional midline to start the balanced deviation analysis. By diffusing to both sides of the straight line according to the preset balanced deviation step, the amount of data in different diffusion areas is counted to determine which direction of the diffusion area contains the most data points, thereby reflecting the main distribution trend of the data. After the amount of data in multiple diffusion areas is counted, the mean straight line corresponding to the maximum amount of data is selected to be retained as the target mean straight line. The target mean straight line not only represents the main distribution direction of the data, but also provides a benchmark for the balanced deviation analysis. Taking the target mean straight line as the starting point, a balanced deviation analysis is performed in the first monitoring data set. By calculating the degree of deviation of each data point relative to the target mean straight line, the first deviation monitoring data can be generated, and the first deviation monitoring data reflects the deviation of each point in the data set relative to the main distribution trend. In the same way, the remaining monitoring data clusters are subjected to balanced deviation analysis to generate M deviation monitoring data clusters.

[0030] Further, taking the target mean straight line as a starting point, performing a balanced deviation analysis in the first monitoring data set to generate first deviation monitoring data, the method includes: Taking the target mean straight line as the starting point, moving to both sides of the straight line according to the preset equilibrium deviation step length to generate two stage straight lines; The data volumes in the two diffusion regions corresponding to the two stage straight lines are counted respectively, the stage straight line corresponding to the maximum data volume is retained, and it is determined whether the data volume of the retained stage straight line is greater than the data volume of the target mean straight line. If so, the retained stage straight line is updated as the starting point, and the balanced deviation analysis is continued until the data volume difference after two adjacent moves is less than the preset difference gain, and the stage straight line corresponding to the maximum data volume in the analysis process is used as the target stage straight line; The first deviation monitoring data is generated by performing mean calculation on a plurality of monitoring data within a target diffusion region of the target phase straight line.

[0031] Taking the target mean straight line as the starting point, move to both sides of the straight line according to the preset balanced deviation step to generate two stage straight lines. These two stage straight lines represent the deviation ranges in two different directions from the mean straight line. Then, the data volume in the two diffusion areas corresponding to the two stage straight lines is counted respectively. The size of the data volume reflects the density of the data points in the area, so as to determine which direction the deviation is more significant. The stage straight line corresponding to the maximum data volume is retained, and further check whether the data volume of the retained stage straight line is greater than the data volume of the target mean straight line. If so, it means that the deviation in this direction is more representative. The retained stage straight line is updated as a new starting point, and the balanced deviation analysis is continued. Iterate continuously until the difference in the data volume after two adjacent moves is less than the preset difference gain, and the stage straight line corresponding to the maximum data volume in the analysis process is used as the target stage straight line. The target stage straight line not only represents the main deviation direction of the data, but also the corresponding data volume is the largest locally, thereby ensuring the accuracy of the analysis. The mean value of multiple monitoring data in the target diffusion area of ​​the target stage straight line is calculated. By calculating the average value of the data in this area, the first deviation monitoring data is obtained. The first deviation monitoring data reflects the average performance of the target device in a specific deviation direction.

[0032] Based on the M deviation monitoring data clusters, the M demand control sub-unit sets are utilized to perform mapping intelligent control on the corresponding rectifiers.

[0033] According to the M deviation monitoring data clusters, the corresponding rectifiers are mapped and intelligently controlled using the M demand control subunit sets corresponding to the M deviation monitoring data clusters, that is, a personalized control plan is formulated according to the specific situation of each rectifier.

[0034] In summary, the embodiments of the present application have at least the following technical effects: The interactive rectifier intelligent control platform extracts N target devices in the target control area, configures N rectifiers for the N target devices, wherein the N rectifiers have N position coordinate identifiers. Then, the demand analysis is performed on the N device information of the N target devices to obtain N target demand identifiers. Next, clustering and centralized analysis is performed based on the N target demand identifiers to obtain M target device clusters, wherein the M target device clusters have M target centralized demand identifiers. Based on the M target centralized demand identifiers, the M demand control units in the rectifier intelligent control platform are called, and the device information of the target devices in the M target device clusters is respectively input into the M demand control units for mapping learning to generate M demand control sub-unit sets. The M demand control sub-unit sets are traversed to perform resource configuration, and the M target device clusters are monitored and analyzed using the configured M demand control sub-unit sets to generate M monitoring data clusters, wherein each monitoring data cluster includes multiple monitoring data sets. Balanced deviation analysis is performed on the monitoring data sets in the M monitoring data clusters to generate M deviation monitoring data clusters. Finally, based on M deviation monitoring data clusters, M demand control subunit sets are used to map and intelligently control the corresponding rectifiers. This solves the technical problem of rectifier control deviation caused by insufficient analysis of different application scenarios in the prior art, and achieves the technical effect of eliminating rectifier control deviation.

[0035] Embodiment 2 Based on the same inventive concept as the rectifier intelligent control method applicable to power electronic equipment in the aforementioned embodiment, Figure 2 As shown, the present application provides a rectifier intelligent control system applicable to power electronic equipment, and the system and method embodiments in the present application are based on the same inventive concept. The system includes: An extraction module 11, the extraction module 11 is used to interact with the rectifier intelligent control platform to extract N target devices in a target control area, and configure N rectifiers for the N target devices, wherein the N rectifiers have N position coordinate identifiers; A demand analysis module 12, the demand analysis module 12 is used to perform demand analysis on the N device information of the N target devices to obtain N target demand identifiers; A cluster analysis module 13, the cluster analysis module 13 is used to perform cluster centralized analysis according to the N target demand identifiers to obtain M target device clusters, wherein the M target device clusters have M target centralized demand identifiers; A mapping learning module 14, wherein the mapping learning module 14 is used to call M demand control units in the rectifier intelligent control platform based on the M target centralized demand identifiers, input device information of target devices in the M target device clusters into the M demand control units for mapping learning, and generate M demand control sub-unit sets; A monitoring and analysis module 15, wherein the monitoring and analysis module 15 is used to traverse the M demand control subunit sets to perform resource configuration, and use the configured M demand control subunit sets to monitor and analyze the M target device clusters to generate M monitoring data clusters, wherein each monitoring data cluster includes multiple monitoring data sets; A deviation analysis module 16, the deviation analysis module 16 is used to perform a balanced deviation analysis on the monitoring data sets in the M monitoring data clusters to generate M deviation monitoring data clusters; The control module 17 is used to map and intelligently control the corresponding rectifiers based on the M deviation monitoring data clusters and using the M demand control sub-unit sets.

[0036] Furthermore, the demand analysis module 12 is used to perform the following method: Traversing the N target devices to extract information and generate N device information, wherein the N device information includes N device models, N device working fields and N device load fluctuation information; Based on the N device models, the N device working fields and the N device load fluctuation information, matching is performed in the demand analysis space to obtain N matching demand sets; The first n requirements in the N matching requirement sets are extracted respectively to generate the N target requirement identifiers.

[0037] Furthermore, the demand analysis module 12 is used to perform the following method: Collecting multiple sample device information and multiple sample demand identifiers, wherein the multiple sample device information includes multiple sample device models, multiple sample device working fields, and multiple sample device load fluctuation information; Constructing a plurality of sample points in the demand analysis space based on the plurality of sample device information including a plurality of sample device models, a plurality of sample device working fields and a plurality of sample device load fluctuation information, wherein the demand analysis space is a three-dimensional space; Inputting the N equipment models, the N equipment working fields and the N equipment load fluctuation information into the demand analysis space to obtain multiple target points; Taking the multiple target points as the center respectively, the m sample demand identifiers of the m sample points whose distances to the multiple target points in the demand analysis space are in the first m positions are summarized to generate the N matching demand sets, where m is an integer greater than or equal to 3.

[0038] Furthermore, the mapping learning module 14 is used to perform the following method: Acquire multiple sample control unit parameter sets, multiple monitoring data sets, and multiple rectifier control parameter sets as multiple training data sets, respectively, and pre-train a framework built based on a convolutional neural network, wherein the multiple rectifier control parameter sets have multiple sample set requirement identifiers; Use the pre-training loss function to perform loss certification on the pre-training process, and update and adjust multiple model parameters based on the loss certification results; After multiple updates, when the loss authentication result meets the requirements and / or the number of iterations meets the preset number of iterations, multiple demand control units of the rectifier intelligent control platform are generated with updated adjustment results; Using the M target concentrated demand identifiers as indexes, searching is performed among multiple demand control units of the rectifier intelligent control platform to obtain the M demand control units.

[0039] Furthermore, the mapping learning module 14 is used to perform the following method: The pre-training loss function is: ; in, is the loss value, To monitor data collection, is the set of rectifier control parameters, is the learning rate in the model, , is an empirical parameter, is the verification rectifier control parameter set output by the model, is the weight.

[0040] Furthermore, the deviation analysis module 16 is used to perform the following method: Extracting a first monitoring data cluster from the M monitoring data clusters, and then extracting a first monitoring data set from the first monitoring data cluster; Performing mean calculation on the first monitoring data set to generate a first monitoring data mean; Taking multiple mean straight lines in any direction passing through the mean of the first monitoring data as the midpoint of the region, diffusing to both sides of the straight line according to the preset balanced deviation step, counting the amount of data in multiple diffusion areas, and retaining the mean straight line corresponding to the maximum amount of data as the target mean straight line; Taking the target mean straight line as a starting point, performing a balanced deviation analysis in the first monitoring data set to generate first deviation monitoring data; A balanced deviation analysis is performed on the M monitoring data clusters to generate M deviation monitoring data clusters.

[0041] Furthermore, the deviation analysis module 16 is used to perform the following method: Taking the target mean straight line as the starting point, moving to both sides of the straight line according to the preset equilibrium deviation step length to generate two stage straight lines; The data volumes in the two diffusion regions corresponding to the two stage straight lines are counted respectively, the stage straight line corresponding to the maximum data volume is retained, and it is determined whether the data volume of the retained stage straight line is greater than the data volume of the target mean straight line. If so, the retained stage straight line is updated as the starting point, and the balanced deviation analysis is continued until the data volume difference after two adjacent moves is less than the preset difference gain, and the stage straight line corresponding to the maximum data volume in the analysis process is used as the target stage straight line; The first deviation monitoring data is generated by performing mean calculation on a plurality of monitoring data within a target diffusion region of the target phase straight line.

[0042] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0043] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0044] This specification and drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. A rectifier intelligent control method applicable to power electronic equipment, characterized in that: The method is applied to a rectifier intelligent control platform, and the method comprises: Interacting with the rectifier intelligent control platform to extract N target devices in a target control area, and configuring N rectifiers for the N target devices, wherein the N rectifiers have N position coordinate identifiers; Performing demand analysis on N device information of the N target devices to obtain N target demand identifiers; Performing clustering and centralized analysis according to the N target demand identifiers to obtain M target device clusters, wherein the M target device clusters have M target centralized demand identifiers; Based on the M target centralized demand identifiers, M demand control units in the rectifier intelligent control platform are called, and device information of target devices in the M target device clusters is respectively input into the M demand control units for mapping learning to generate M demand control sub-unit sets; Traversing the M demand control subunit sets to perform resource configuration, and using the configured M demand control subunit sets to monitor and analyze the M target device clusters to generate M monitoring data clusters, wherein each monitoring data cluster includes multiple monitoring data sets; Performing a balanced deviation analysis on the monitoring data sets in the M monitoring data clusters to generate M deviation monitoring data clusters; Based on the M deviation monitoring data clusters, the M demand control sub-unit sets are utilized to perform mapping intelligent control on the corresponding rectifiers.

2. The method according to claim 1, characterized in that Performing demand analysis on N device information of the N target devices to obtain N target demand identifiers, the method comprising: Traversing the N target devices to extract information and generate N device information, wherein the N device information includes N device models, N device working fields and N device load fluctuation information; Based on the N device models, the N device working fields and the N device load fluctuation information, matching is performed in the demand analysis space to obtain N matching demand sets; The first n requirements in the N matching requirement sets are extracted respectively to generate the N target requirement identifiers.

3. The method according to claim 2, characterized in that The method comprises: Collecting a plurality of sample device information and a plurality of sample demand identifiers, wherein the plurality of sample device information includes a plurality of sample device models, a plurality of sample device working fields and a plurality of sample device load fluctuation information; Constructing a plurality of sample points in the demand analysis space based on the plurality of sample device information including a plurality of sample device models, a plurality of sample device working fields and a plurality of sample device load fluctuation information, wherein the demand analysis space is a three-dimensional space; Inputting the N equipment models, the N equipment working fields and the N equipment load fluctuation information into the demand analysis space to obtain multiple target points; Taking the multiple target points as the center respectively, the m sample demand identifiers of the m sample points whose distances to the multiple target points in the demand analysis space are in the first m positions are summarized to generate the N matching demand sets, where m is an integer greater than or equal to 3.

4. The method according to claim 1, characterized in that Based on the M target centralized demand identifiers, calling M demand control units in the rectifier intelligent control platform, the method includes: Acquire multiple sample control unit parameter sets, multiple monitoring data sets, and multiple rectifier control parameter sets as multiple training data sets, respectively, and pre-train a framework built based on a convolutional neural network, wherein the multiple rectifier control parameter sets have multiple sample set requirement identifiers; Use the pre-training loss function to perform loss certification on the pre-training process, and update and adjust multiple model parameters based on the loss certification results; After multiple updates, when the loss authentication result meets the requirements and / or the number of iterations meets the preset number of iterations, multiple demand control units of the rectifier intelligent control platform are generated with updated adjustment results; Using the M target concentrated demand identifiers as indexes, searching is performed among multiple demand control units of the rectifier intelligent control platform to obtain the M demand control units.

5. The method according to claim 4, characterized in that The pre-training loss function is: ； in, is the loss value, To monitor data collection, is the set of rectifier control parameters, is the learning rate in the model, , is an empirical parameter, is the verification rectifier control parameter set output by the model, is the weight.

6. The method according to claim 1, characterized in that Performing a balanced deviation analysis on the monitoring data sets in the M monitoring data clusters to generate M deviation monitoring data clusters, the method comprising: Extracting a first monitoring data cluster from the M monitoring data clusters, and then extracting a first monitoring data set from the first monitoring data cluster; Performing mean calculation on the first monitoring data set to generate a first monitoring data mean; Taking multiple mean straight lines in any direction passing through the mean of the first monitoring data as the midpoint of the region, diffusing to both sides of the straight line according to the preset balanced deviation step, counting the amount of data in multiple diffusion areas, and retaining the mean straight line corresponding to the maximum amount of data as the target mean straight line; Taking the target mean straight line as a starting point, performing a balanced deviation analysis in the first monitoring data set to generate first deviation monitoring data; A balanced deviation analysis is performed on the M monitoring data clusters to generate M deviation monitoring data clusters.

7. The method according to claim 6, characterized in that Taking the target mean straight line as a starting point, performing a balanced deviation analysis in the first monitoring data set to generate first deviation monitoring data, the method comprising: Taking the target mean straight line as the starting point, moving to both sides of the straight line according to the preset equilibrium deviation step length to generate two stage straight lines; The data volumes in the two diffusion regions corresponding to the two stage straight lines are counted respectively, the stage straight line corresponding to the maximum data volume is retained, and it is determined whether the data volume of the retained stage straight line is greater than the data volume of the target mean straight line. If so, the retained stage straight line is updated as the starting point, and the balanced deviation analysis is continued until the data volume difference after two adjacent moves is less than the preset difference gain, and the stage straight line corresponding to the maximum data volume in the analysis process is used as the target stage straight line; The first deviation monitoring data is generated by performing mean calculation on a plurality of monitoring data within a target diffusion region of the target phase straight line.

8. A rectifier intelligent control system suitable for power electronic equipment, characterized in that: A rectifier intelligent control method applicable to power electronic equipment for implementing any one of claims 1 to 7, the system comprising: An extraction module, the extraction module is used to interact with the rectifier intelligent control platform to extract N target devices in a target control area, and configure N rectifiers for the N target devices, wherein the N rectifiers have N position coordinate identifiers; A demand analysis module, the demand analysis module is used to perform demand analysis on N device information of the N target devices to obtain N target demand identifiers; A cluster analysis module, the cluster analysis module is used to perform cluster centralized analysis according to the N target demand identifiers to obtain M target device clusters, wherein the M target device clusters have M target centralized demand identifiers; A mapping learning module, wherein the mapping learning module is used to call M demand control units in the rectifier intelligent control platform based on the M target centralized demand identifiers, input device information of target devices in the M target device clusters into the M demand control units for mapping learning, and generate M demand control sub-unit sets; A monitoring and analysis module, the monitoring and analysis module is used to traverse the M demand control subunit sets to perform resource configuration, and use the configured M demand control subunit sets to monitor and analyze the M target device clusters to generate M monitoring data clusters, wherein each monitoring data cluster includes multiple monitoring data sets; A deviation analysis module, the deviation analysis module is used to perform a balanced deviation analysis on the monitoring data sets in the M monitoring data clusters to generate M deviation monitoring data clusters; A control module is used to map and intelligently control corresponding rectifiers based on the M deviation monitoring data clusters and using the M demand control sub-unit sets.

Images