Parallel transformer control system based on new energy flexible DC system

Abstract

The invention relates to the technical field of power supply intelligent dispatching systems, in particular to a control system of a parallel transformer based on a new energy flexible direct current system, which comprises a data acquisition module used for acquiring monitoring data of the new energy flexible direct current system; the data processing module is used for processing the monitoring data of the new energy flexible direct current system to obtain model training data; the model construction module is used for training and outputting a system monitoring model; the model monitoring module is used for carrying out real-time monitoring; the scheduling decision module is used for obtaining a scheduling decision; the scheduling execution module is used for executing; and the execution feedback module is used for performing execution feedback and changing an execution decision, and is also used for judging an adjustment condition and performing adjustment and alarm. The parallel transformer is controlled through the scheduling decision, intelligent scheduling of the new energy flexible direct current system is realized, and the power transmission control efficiency of the new energy flexible direct current system is improved.

Description

Technical Field

[0001] This invention relates to the field of intelligent power supply dispatching system technology, and in particular to a control system for a parallel transformer based on a new energy flexible DC system. Background Technology

[0002] With the rapid development of new energy sources and the transformation of the energy structure, traditional power dispatching methods can no longer meet the demand, and intelligent dispatching has become an inevitable choice. The modern power grid structure is becoming increasingly complex, and the requirements for power dispatching are also getting higher and higher. Intelligent dispatching can cope with complex power grid environments, ensure the safe and stable operation of the power grid, monitor the operation status of the power grid in real time, promptly detect and deal with potential safety hazards, and improve the reliability and security of power supply.

[0003] Chinese Patent Publication No. CN115189390A discloses a control method and system for parallel transformers based on a new energy flexible DC system. The method includes: selecting transformers and transformer lines to be put into operation from target transformers according to target requirements, and confirming whether the flexible DC system has charging capabilities; controlling the transformers to be put into operation to perform a startup operation according to a preset startup control strategy, while simultaneously controlling the transformer lines to be put into operation to perform an operation; controlling the transformers to be put into operation to perform an online parallel operation step according to a preset online parallel control strategy; and controlling the onshore station parallel transformers to perform a tap changer self-synchronization step according to a preset tap changer self-synchronization control strategy. However, this scheme lacks intelligent scheduling, resulting in low power transmission control efficiency. Summary of the Invention

[0004] Therefore, this invention provides a control system for a parallel transformer based on a new energy flexible DC system to overcome the problem of low power transmission control efficiency in the prior art.

[0005] To achieve the above objectives, the present invention provides a control system for a parallel transformer based on a new energy flexible DC system, comprising:

[0006] The data acquisition module is used to collect monitoring data from the new energy flexible DC system;

[0007] The data processing module is used to process the collected monitoring data of the new energy flexible DC system and obtain model training data.

[0008] The model building module is used to train and output the system monitoring model based on the model training data;

[0009] The model monitoring module is used to perform real-time monitoring based on the system monitoring model;

[0010] The scheduling decision module is used to make decisions based on real-time monitoring results and obtain scheduling decisions;

[0011] The scheduling and execution module is used to execute tasks based on scheduling decisions.

[0012] The execution feedback module is used to provide execution feedback based on the duration of the anomaly, modify the execution decision based on the feedback results, judge the adjustment situation based on the frequency of the anomaly, and adjust and issue alarms in the execution feedback process based on the judgment results.

[0013] Furthermore, the data processing module performs data cleaning and preprocessing on the collected monitoring data of the new energy flexible DC system, detects outliers in the new energy flexible DC system monitoring data after data cleaning and preprocessing, and deletes the outliers. The data processing module obtains the new energy flexible DC system monitoring data after deleting outliers and the corresponding reasons for the anomalies in the new energy flexible DC system monitoring data after deleting outliers within the training period, and uses them as model training data to divide them into training set, validation set and test set.

[0014] Furthermore, the model building module inputs the training set of model training data into the neural network model to initialize the neural network model, trains the initialized neural network model, and tests the trained neural network model based on the test set of the model training data. When the accuracy U of the test result is less than or equal to the preset test accuracy U0, the model training data of the next training cycle is obtained to train the initialized neural network model. When the accuracy U of the test result is greater than the preset test accuracy U0, the tested neural network model is verified based on the validation set of the model training data. When the accuracy W of the validation result is less than or equal to the preset validation accuracy W0, the model training data of the next training cycle is obtained to train the initialized neural network model. When the accuracy W of the validation result is greater than the preset validation accuracy W0, the validated neural network model is output as the system monitoring model.

[0015] Furthermore, the model monitoring module inputs the monitoring data of the new energy flexible DC system after deleting outliers into the system monitoring model and outputs real-time monitoring results to monitor the monitoring data of the new energy flexible DC system in real time.

[0016] Furthermore, the scheduling decision module performs a text comparison between the real-time monitoring results and the preset anomaly causes in the expert decision database, and makes a decision based on the comparison results, wherein:

[0017] When no preset abnormal cause consistent with the real-time monitoring results is found in the expert decision database, a decision alarm is issued to the user, and the real-time monitoring results and decision window are pushed to the user terminal to obtain the decision entered by the user in the decision window and use it as the scheduling decision.

[0018] When a preset anomaly cause consistent with the real-time monitoring results exists in the expert decision database, the scheduling decision is judged based on the number of decisions n corresponding to the preset anomaly cause, where:

[0019] If n=1, the scheduling decision module determines the decision corresponding to the preset abnormal cause as the scheduling decision;

[0020] If n > 1, the scheduling decision module will sequentially input the decisions corresponding to the preset abnormal causes into the decision simulation model for simulation, obtain simulation data, calculate the effective coefficient a based on the simulation data, and set a = 0.3 × R / R0 + 0.7 × Y / Y0, where R is the simulation execution time, R0 is the preset simulation execution time, Y is the simulation abnormal duration, and Y0 is the preset simulation abnormal duration. The scheduling decision module will sort the decisions according to the effective coefficient a in ascending order to obtain the decision ranking, and take the decision ranked first as the scheduling decision.

[0021] Furthermore, the scheduling execution module obtains the scheduling decision obtained by the scheduling decision module and executes it as the execution decision.

[0022] Further, the execution feedback module obtains the abnormal duration t, compares the abnormal duration t with the preset abnormal duration t0, and performs execution feedback based on the comparison result, wherein:

[0023] When t≤t0, the execution feedback module reports that the execution is normal;

[0024] When t > t0, the execution feedback module reports an execution exception and modifies the execution decision.

[0025] Furthermore, when the execution feedback module reports an execution exception, it modifies the execution decision, wherein:

[0026] If the scheduling decision corresponding to the execution decision is the decision entered by the user in the decision window and the number of decisions n=1, the execution feedback module pushes a change decision window to the user, takes the decision entered by the user as the changed execution decision, and sends the changed execution decision to the scheduling execution module for execution until the execution feedback module reports that the execution is normal.

[0027] If the scheduling decision corresponding to the execution decision is a decision with n > 1, the execution feedback module will take the next decision in the decision ranking as the modified execution decision. The execution feedback module will send the modified execution decision to the scheduling execution module for execution until the execution feedback module reports that the execution is normal. When there is no next decision in the decision ranking, the execution feedback module will push a change decision window to the user and take the decision entered by the user as the modified execution decision.

[0028] Furthermore, the execution feedback module obtains the anomaly frequency p, compares the anomaly frequency p with each preset anomaly frequency, and judges the adjustment situation based on the comparison result, wherein:

[0029] When p≤p1, the execution feedback module determines that no adjustment should be made;

[0030] When p1 < p ≤ p2, the execution feedback module determines to make adjustments;

[0031] When p2 < p, the execution feedback module determines that no adjustment should be made, the execution feedback module connects to the redundant device, and sends an abnormal frequency alarm to the user;

[0032] p1 is the first preset abnormal frequency, p2 is the second preset abnormal frequency, and 0 < p1 < p2.

[0033] Furthermore, when the execution feedback module determines that an adjustment is needed, it sets the adjustment coefficient Z = 0.85 + e. -(p-p1)-2 e is the base of the natural logarithm. The execution feedback module adjusts the preset abnormal duration t0 in the execution feedback process according to the adjustment coefficient. The adjusted preset abnormal duration is tz0, where tz0 = Z × t0.

[0034] Compared with existing technologies, the beneficial effects of this invention are as follows: The system processes the collected monitoring data of the new energy flexible DC system through a data processing module to obtain model training data, ensuring the accuracy and completeness of the monitoring data to facilitate subsequent feature extraction and model training. The system also trains and outputs a system monitoring model based on the model training data through a model building module, allowing for model training according to actual conditions and improving the effectiveness and accuracy of model training. Furthermore, the system performs real-time monitoring based on the system monitoring model through a model monitoring module to monitor the power transmission status of the new energy flexible DC system in real time. Finally, the system makes decisions based on the real-time monitoring results through a scheduling decision module, obtaining scheduling decisions to optimize the scheduling of the new energy flexible DC system. The system controls the parallel transformers of the flexible DC system to further improve power transmission control efficiency. The system executes scheduling decisions through a scheduling execution module to control the parallel transformers of the flexible DC system, further improving power transmission control efficiency. The system also uses an execution feedback module to provide execution feedback based on the duration of abnormalities and modifies execution decisions based on the feedback results to achieve intelligent scheduling and optimized control of the flexible DC system. Furthermore, the system uses the execution feedback module to judge the adjustment status based on the frequency of abnormalities and adjusts and alarms the execution feedback process based on the judgment results to achieve intelligent scheduling and alarm prompts for dangerous situations, thereby improving the safety of the flexible DC system and enhancing power transmission control efficiency. Attached Figure Description

[0035] Figure 1 This is a schematic diagram of the control system of the parallel transformer based on the new energy flexible DC system in this embodiment. Detailed Implementation

[0036] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0037] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0038] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0039] Please see Figure 1 As shown, this is a schematic diagram of the control system for the parallel transformer based on the new energy flexible DC system in this embodiment. The system includes:

[0040] The data acquisition module is used to collect monitoring data from the new energy flexible DC system;

[0041] The data processing module is used to process the collected monitoring data of the new energy flexible DC system and obtain model training data. The data processing module is connected to the data acquisition module.

[0042] The model building module is used to train and output the system monitoring model based on the model training data. The model building module is connected to the data processing module.

[0043] The model monitoring module is used to perform real-time monitoring based on the system monitoring model. The model monitoring module is connected to the model building module.

[0044] The scheduling decision module is used to make decisions based on real-time monitoring results and obtain scheduling decisions. The scheduling decision module is connected to the model monitoring module.

[0045] The scheduling execution module is used to execute according to the scheduling decision, and the scheduling execution module is connected to the scheduling decision module;

[0046] The execution feedback module is used to provide execution feedback based on the duration of the anomaly, and to modify the execution decision based on the feedback results. It is also used to judge the adjustment situation based on the frequency of the anomaly, and to adjust and alarm the execution feedback process based on the judgment results. The execution feedback module is connected to the scheduling execution module.

[0047] Specifically, the system is applied to a new energy flexible DC system to monitor the parallel transformers of the system and control them based on the monitoring data. This enables intelligent scheduling of the new energy flexible DC system, thereby improving power transmission control efficiency. The system collects monitoring data from the new energy flexible DC system through a data acquisition module for subsequent analysis. A data processing module processes the collected monitoring data to obtain model training data, ensuring the accuracy and completeness of the monitoring data for subsequent feature extraction and model training. A model building module trains and outputs the system monitoring model based on the training data, allowing for model training according to actual conditions and improving the effectiveness and accuracy of model training. Finally, a model monitoring module performs real-time monitoring based on the system monitoring model to monitor the new energy flexible DC system. The system monitors power transmission in real time. Based on the real-time monitoring results, the system's scheduling decision module makes decisions to control the parallel transformers of the new energy flexible DC system, further improving power transmission control efficiency. The system's scheduling execution module executes these decisions to control the parallel transformers of the new energy flexible DC system, further improving power transmission control efficiency. The system's execution feedback module provides feedback based on the duration of abnormalities and modifies the execution decisions based on the feedback results, achieving intelligent scheduling and optimized control of the new energy flexible DC system. Furthermore, the system's execution feedback module judges the adjustment situation based on the frequency of abnormalities and adjusts and alarms the execution feedback process based on the judgment results, achieving intelligent scheduling and alarm prompts for dangerous situations, thereby improving the safety of the new energy flexible DC system and enhancing power transmission control efficiency.

[0048] Specifically, the monitoring data of the new energy flexible DC system includes system voltage, system current, system power, and equipment temperature. This embodiment does not limit the method of acquiring system voltage; those skilled in the art can freely set it according to actual conditions, as long as the requirement for accurate acquisition of system voltage is met. For example, it can be set to acquire the voltage at the output terminal of the new energy flexible DC system through a voltage sensor. This embodiment does not limit the method of acquiring system current; those skilled in the art can freely set it according to actual conditions, as long as the requirement for accurate acquisition of system current is met. For example, it can be set to acquire the current at the output terminal of the new energy flexible DC system through an ammeter. This embodiment does not limit the method of acquiring system power; those skilled in the art can freely set it according to actual conditions, as long as the requirement for accurate acquisition of system power is met. For example, it can be set to acquire the power at the output terminal of the new energy flexible DC system through a power meter. This embodiment does not limit the method of acquiring equipment temperature; those skilled in the art can freely set it according to actual conditions, as long as the requirement for accurate acquisition of equipment temperature is met. For example, it can be set to acquire the temperature at the output terminal of the new energy flexible DC system through a temperature sensor.

[0049] Specifically, the data processing module performs data cleaning and preprocessing on the collected monitoring data of the new energy flexible DC system, detects outliers in the new energy flexible DC system monitoring data after data cleaning and preprocessing, and deletes the outliers. The data processing module obtains the new energy flexible DC system monitoring data after deleting outliers and the corresponding reasons for the anomalies in the new energy flexible DC system monitoring data after deleting outliers within the training period, and uses them as model training data to divide them into training set, validation set and test set.

[0050] Specifically, the data processing module ensures the accuracy and completeness of the monitoring data of the new energy flexible DC system through data cleaning and data preprocessing, so as to facilitate subsequent feature extraction and model training. The data processing module uses the monitoring data of the new energy flexible DC system after deleting outliers during the training period and the corresponding reasons for the outliers as model training data to train the model according to the actual situation, thereby improving the effectiveness and accuracy of model training.

[0051] Specifically, the training period refers to a preset time interval for obtaining monitoring data of the new energy flexible DC system after removing outliers and the corresponding causes of the outliers in the monitoring data of the new energy flexible DC system after removing outliers. This embodiment does not limit the data cleaning method; those skilled in the art can freely set it according to actual conditions, as long as it meets the requirement of accurate data extraction. For example, the data cleaning method can be set to remove duplicate, erroneous, and invalid data from the monitoring data of the new energy flexible DC system. This embodiment does not limit the data preprocessing method; those skilled in the art can freely set it according to actual conditions, as long as it meets the requirement of accurate data extraction. For example, the data preprocessing method can be set to preprocess the new energy flexible DC system. The monitoring data of the flexible DC system is standardized and normalized. This embodiment does not limit the outlier monitoring method. Those skilled in the art can freely set it according to the actual situation, as long as the requirement of accurate data extraction is met. For example, the outlier monitoring method can be set to monitor outliers in the monitoring data of the new energy flexible DC system through machine learning algorithms. This embodiment does not limit the segmentation method. Those skilled in the art can freely set it according to the actual situation, as long as the subsequent model training requirements are met. For example, the segmentation method can be set to randomly select 70% of the model training data as the training set, randomly select 20% of the model training data as the test set, and randomly select 10% of the model training data as the validation set.

[0052] Specifically, the model building module inputs the training set of model training data into the neural network model to initialize the neural network model, trains the initialized neural network model, tests the trained neural network model based on the test set of the model training data, and obtains the model training data for the next training cycle to train the initialized neural network model when the accuracy U of the test result is less than or equal to the preset test accuracy U0. When the accuracy U of the test result is greater than the preset test accuracy U0, the tested neural network model is verified based on the validation set of the model training data. When the accuracy W of the validation result is less than or equal to the preset validation accuracy W0, the model training data for the next training cycle is obtained to train the initialized neural network model. When the accuracy W of the validation result is greater than the preset validation accuracy W0, the validated neural network model is output as the system monitoring model.

[0053] Specifically, initialization refers to setting the hyperparameters of the neural network model. This embodiment does not limit the initialization method; those skilled in the art can freely set them according to actual conditions, as long as the training requirements of the model are met. For example, the learning rate can be set to 0.01, the number of iterations to 5, and the number of hidden layer nodes to 10. The accuracy of the test result is the ratio of the amount of data in the test set that are consistent with the abnormal causes output by the trained neural network model to the total amount of data in the test set. The preset test accuracy rate is a preset value representing the accuracy rate of the trained neural network model passing the test, such as 95%. The accuracy of the verification result is the ratio of the amount of data in the verification set that are consistent with the abnormal causes output by the tested neural network model to the total amount of data in the verification set. The preset verification accuracy rate is a preset value representing the accuracy rate of the trained neural network model passing the verification, such as 98%.

[0054] Specifically, the model monitoring module inputs the monitoring data of the new energy flexible DC system after deleting outliers into the system monitoring model and outputs real-time monitoring results to monitor the monitoring data of the new energy flexible DC system in real time.

[0055] Specifically, the real-time monitoring result refers to the cause of the anomaly corresponding to the monitoring data of the new energy flexible DC system after removing outliers. The cause of the anomaly refers to the root cause of the fault in the new energy flexible DC system as represented by the monitoring data of the new energy flexible DC system after removing outliers.

[0056] Specifically, the scheduling decision module performs a text comparison between the real-time monitoring results and preset anomaly causes in the expert decision database, and makes a decision based on the comparison results, wherein:

[0057] When no preset abnormal cause consistent with the real-time monitoring results is found in the expert decision database, a decision alarm is issued to the user, and the real-time monitoring results and decision window are pushed to the user terminal to obtain the decision entered by the user in the decision window and use it as the scheduling decision.

[0058] When a preset anomaly cause consistent with the real-time monitoring results exists in the expert decision database, the scheduling decision is judged based on the number of decisions n corresponding to the preset anomaly cause, where:

[0059] If n=1, the scheduling decision module determines the decision corresponding to the preset abnormal cause as the scheduling decision;

[0060] If n > 1, the scheduling decision module will sequentially input the decisions corresponding to the preset abnormal causes into the decision simulation model for simulation, obtain simulation data, calculate the effective coefficient a based on the simulation data, and set a = 0.3 × R / R0 + 0.7 × Y / Y0, where R is the simulation execution time, R0 is the preset simulation execution time, Y is the simulation abnormal duration, and Y0 is the preset simulation abnormal duration. The scheduling decision module will sort the decisions according to the effective coefficient a in ascending order to obtain the decision ranking, and take the decision ranked first as the scheduling decision.

[0061] Specifically, the scheduling decision module compares the real-time monitoring results with the preset anomaly causes in the expert decision database to make a decision. The scheduling decision module judges the scheduling decision based on the number of decisions corresponding to the preset anomaly causes in order to select the optimal scheduling decision, improve the efficiency of intelligent scheduling, and control the parallel transformer of the new energy flexible DC system through the optimal scheduling decision, thereby further improving the efficiency of power transmission control.

[0062] Specifically, the expert decision database refers to a database that stores data in a preset anomaly cause-decision format. This preset anomaly cause-decision format refers to the correspondence between preset anomaly causes and decisions, including a format where one preset anomaly cause corresponds to one decision and a format where one preset anomaly cause corresponds to multiple decisions. This embodiment does not specifically limit the setup method of the expert decision database; those skilled in the art can freely set it according to actual conditions, as long as it meets the preset requirements for decisions. For example, it can be set to collect anomaly causes and decisions from big data and preset them manually. The text comparison refers to comparing the anomaly causes from real-time monitoring results with the preset anomaly causes. The decision alarm refers to scheduling for users with management responsibilities. The decision-making alarm prompts the user to make a decision. This embodiment does not limit the decision-making alarm method; those skilled in the art can freely set it according to actual conditions, as long as it meets the user's alarm prompting needs. For example, it can be set to set a pop-up alarm on the user terminal. The decision window refers to the window pushed to the terminal where the user inputs the decision content. This embodiment does not limit the form of the decision window; those skilled in the art can freely set it according to actual conditions, as long as it meets the needs of obtaining the decision content. For example, the decision window can be set to a blank window for user customization, or it can be set to a window including basic decision options for user selection. The number of decisions corresponding to the preset abnormal reasons refers to the preset abnormal reasons minus the decisions. In this format, the number of decisions corresponding to a preset abnormal cause is defined as follows: n=1 when one preset abnormal cause corresponds to one decision, and n=3 when one preset abnormal cause corresponds to three decisions. The decision simulation model refers to a model established based on the new energy flexible DC system. During simulation, the decision simulation model sets edge conditions according to the abnormal cause and runs, acquiring the simulation data output after running. The simulation data includes the simulation execution time and the simulation abnormality duration. The simulation execution time refers to the time required for the decision simulation model to execute all the steps included in the decision during simulation. This embodiment does not limit the method of obtaining the simulation execution time; those skilled in the art can freely set it according to actual needs, as long as it meets the requirements of simulation execution. The requirement for precise acquisition of the simulation duration is sufficient. For example, the start and end times of the simulation can be collected via a built-in clock, and the time difference between these times can be calculated as the simulation execution duration. The decision refers to the steps set based on the cause of the anomaly to resolve it. The cause of the anomaly is the reason leading to the anomaly in the new energy flexible DC system. The duration of the simulation anomaly refers to the time elapsed from the appearance of the cause of the anomaly to its disappearance during the simulation by the decision-making simulation model. This embodiment does not limit the method for acquiring the duration of the simulation anomaly; those skilled in the art can freely set it according to the actual situation, as long as the requirement for precise acquisition of the duration of the simulation anomaly is met.For example, the duration of simulated anomalies can be collected via a built-in clock.

[0063] Specifically, the scheduling execution module obtains the scheduling decision from the scheduling decision module and executes it as the execution decision.

[0064] It is understood that this embodiment does not limit the execution method of the execution decision. Those skilled in the art can freely set it according to the actual situation, as long as the execution requirements of the execution decision are met. For example, it can be set to preset execution instructions in the decision, and control the parallel transformer of the new energy flexible DC system according to the preset execution instructions as the execution decision, so as to execute the execution decision.

[0065] Specifically, the execution feedback module obtains the abnormal duration t, compares the abnormal duration t with the preset abnormal duration t0, and performs execution feedback based on the comparison result, wherein:

[0066] When t≤t0, the execution feedback module reports that the execution is normal;

[0067] When t > t0, the execution feedback module reports an execution exception and modifies the execution decision.

[0068] Specifically, when the execution feedback module reports an execution exception, it modifies the execution decision, wherein:

[0069] If the scheduling decision corresponding to the execution decision is the decision entered by the user in the decision window and the number of decisions n=1, the execution feedback module pushes a change decision window to the user, takes the decision entered by the user as the changed execution decision, and sends the changed execution decision to the scheduling execution module for execution until the execution feedback module reports that the execution is normal.

[0070] If the scheduling decision corresponding to the execution decision is a decision with n > 1, the execution feedback module will take the next decision in the decision ranking as the modified execution decision. The execution feedback module will send the modified execution decision to the scheduling execution module for execution until the execution feedback module reports that the execution is normal. When there is no next decision in the decision ranking, the execution feedback module will push a change decision window to the user and take the decision entered by the user as the modified execution decision.

[0071] Specifically, when the execution feedback module reports an execution anomaly, it modifies the execution decision to achieve intelligent scheduling of the new energy flexible DC system, optimize and modify the control of the new energy flexible DC system, thereby improving the accuracy of controlling the parallel transformers of the new energy flexible DC system and further improving the power transmission control efficiency.

[0072] Specifically, the abnormal duration refers to the time elapsed from the occurrence of the abnormal cause to its disappearance. This embodiment does not limit the method of obtaining the abnormal duration; those skilled in the art can freely set it according to actual conditions, as long as the requirement for accurate acquisition of the abnormal duration is met. For example, the abnormal duration can be obtained through system logs. The preset abnormal duration refers to a preset value reflecting the abnormal duration of the execution abnormality, such as 30 seconds. The change decision window refers to a window pushed by the user, where the user inputs the changed execution decision content. This embodiment does not limit the form of the change decision window; those skilled in the art can freely set it according to actual conditions, as long as the requirement for acquisition of the changed execution decision content is met. For example, it can be set to a blank window for user-defined input. The next decision refers to the next scheduling decision corresponding to the current execution decision in the decision ranking. The absence of a next decision in the decision ranking means that the scheduling decision corresponding to the current execution decision is the last scheduling decision in the decision ranking.

[0073] Specifically, the execution feedback module obtains the abnormal frequency p, compares the abnormal frequency p with each preset abnormal frequency, and judges the adjustment situation based on the comparison result, wherein:

[0074] When p≤p1, the execution feedback module determines that no adjustment should be made;

[0075] When p1 < p ≤ p2, the execution feedback module determines to make adjustments;

[0076] When p2 < p, the execution feedback module determines that no adjustment should be made, the execution feedback module connects to the redundant device, and sends an abnormal frequency alarm to the user;

[0077] p1 is the first preset abnormal frequency, p2 is the second preset abnormal frequency, and 0 < p1 < p2.

[0078] Specifically, when the execution feedback module determines that an adjustment is needed, it sets the adjustment coefficient Z = 0.85 + e. -(p-p1)-2 e is the base of the natural logarithm. The execution feedback module adjusts the preset abnormal duration t0 in the execution feedback process according to the adjustment coefficient. The adjusted preset abnormal duration is tz0, where tz0 = Z × t0.

[0079] Specifically, the execution feedback module adjusts the preset abnormal duration during the execution feedback process according to the adjustment coefficient to reduce the preset abnormal duration, thereby increasing the judgment strength of the feedback execution abnormality, improving the sensitivity of the feedback of execution abnormality, and further improving the accuracy of the feedback, thereby improving the accuracy of intelligent scheduling and improving the efficiency of power transmission control. When the abnormal frequency is greater than the second preset abnormal frequency, the execution feedback module connects to redundant equipment and sends an abnormal frequency alarm to the user to realize intelligent scheduling and alarm prompts for dangerous situations, thereby improving the safety of the new energy flexible DC system and improving the efficiency of power transmission control.

[0080] Specifically, the abnormal frequency refers to the number of times an abnormal cause occurs within a preset feedback period. The preset feedback period is a preset time interval for judging the adjustment based on the abnormal frequency. For example, if the preset feedback period is 7 days, the adjustment is judged every 7 days based on the abnormal frequency within 7 days. This embodiment does not limit the method of obtaining the abnormal frequency. Those skilled in the art can freely set it according to the actual situation. For example, it can be set to count the number of abnormal causes by a built-in counter. The first preset abnormal frequency refers to the preset value of the abnormal frequency for adjustment. For example, p1 can be set to 5. The second preset abnormal frequency refers to the preset value of the abnormal frequency for alarm. For example, p2 can be set to 20. The redundant equipment refers to the backup equipment in the new energy flexible DC system. This embodiment does not limit the access method of the redundant equipment. Those skilled in the art can freely set it according to the actual situation, as long as the replacement requirements of the redundant equipment are met. For example, the redundant equipment can be replaced by a parallel transformer. This embodiment does not limit the alarm method of the abnormal frequency. Those skilled in the art can freely set it according to the actual situation, as long as the alarm prompt requirements of the user are met. For example, an alarm sound prompt can be set on the user terminal.

[0081] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. A control system for a parallel transformer based on a new energy flexible DC system, characterized in that, include: The data acquisition module is used to collect monitoring data from the new energy flexible DC system; The data processing module is used to process the collected monitoring data of the new energy flexible DC system and obtain model training data. The model building module is used to train and output the system monitoring model based on the model training data; The model monitoring module is used to perform real-time monitoring based on the system monitoring model; The scheduling decision module is used to make decisions based on real-time monitoring results and obtain scheduling decisions; The scheduling and execution module is used to execute tasks based on scheduling decisions. The execution feedback module is used to provide execution feedback based on the duration of the anomaly, modify the execution decision based on the feedback results, judge the adjustment situation based on the frequency of the anomaly, and adjust and issue alarms in the execution feedback process based on the judgment results.

2. The control system for parallel transformers based on a new energy flexible DC system according to claim 1, characterized in that, The data processing module performs data cleaning and preprocessing on the collected monitoring data of the new energy flexible DC system, detects outliers in the monitoring data after data cleaning and preprocessing, and deletes the outliers. The data processing module obtains the monitoring data of the new energy flexible DC system after deleting outliers and the corresponding reasons for the anomalies in the monitoring data of the new energy flexible DC system after deleting outliers within the training period, and uses them as model training data to divide them into training set, validation set and test set.

3. The control system for parallel transformers based on a new energy flexible DC system according to claim 2, characterized in that, The model building module inputs the training set of model training data into the neural network model to initialize the neural network model, trains the initialized neural network model, and tests the trained neural network model based on the test set of the model training data. When the accuracy U of the test result is less than or equal to the preset test accuracy U0, the model training data of the next training cycle is obtained to train the initialized neural network model. When the accuracy U of the test result is greater than the preset test accuracy U0, the tested neural network model is verified based on the validation set of the model training data. When the accuracy W of the validation result is less than or equal to the preset validation accuracy W0, the model training data of the next training cycle is obtained to train the initialized neural network model. When the accuracy W of the validation result is greater than the preset validation accuracy W0, the validated neural network model is output as the system monitoring model.

4. The control system for parallel transformers based on a new energy flexible DC system according to claim 3, characterized in that, The model monitoring module inputs the monitoring data of the new energy flexible DC system after deleting outliers into the system monitoring model and outputs real-time monitoring results to monitor the monitoring data of the new energy flexible DC system in real time.

5. The control system for the parallel transformer based on a new energy flexible DC system according to claim 4, characterized in that, The scheduling decision module performs a text comparison between the real-time monitoring results and the preset anomaly causes in the expert decision database, and makes a decision based on the comparison results, wherein: When no preset abnormal cause consistent with the real-time monitoring results is found in the expert decision database, a decision alarm is issued to the user, and the real-time monitoring results and decision window are pushed to the user terminal to obtain the decision entered by the user in the decision window and use it as the scheduling decision. When a preset anomaly cause consistent with the real-time monitoring results exists in the expert decision database, the scheduling decision is judged based on the number of decisions n corresponding to the preset anomaly cause, where: If n=1, the scheduling decision module determines the decision corresponding to the preset abnormal cause as the scheduling decision; If n > 1, the scheduling decision module will sequentially input the decisions corresponding to the preset abnormal causes into the decision simulation model for simulation, obtain simulation data, calculate the effective coefficient a based on the simulation data, and set a = 0.3 × R / R0 + 0.7 × Y / Y0, where R is the simulation execution time, R0 is the preset simulation execution time, Y is the simulation abnormal duration, and Y0 is the preset simulation abnormal duration. The scheduling decision module will sort the decisions according to the effective coefficient a in ascending order to obtain the decision ranking, and take the decision ranked first as the scheduling decision.

6. The control system for the parallel transformer based on a new energy flexible DC system according to claim 5, characterized in that, The scheduling execution module obtains the scheduling decision from the scheduling decision module and executes it as the execution decision.

7. The control system for parallel transformers based on a new energy flexible DC system according to claim 1, characterized in that, The execution feedback module obtains the abnormal duration t, compares the abnormal duration t with the preset abnormal duration t0, and performs execution feedback based on the comparison result, wherein: When t≤t0, the execution feedback module reports that the execution is normal; When t > t0, the execution feedback module reports an execution exception and modifies the execution decision.

8. The control system for the parallel transformer based on the new energy flexible DC system according to claim 7, characterized in that, When an execution exception is reported, the execution feedback module modifies the execution decision, wherein: If the scheduling decision corresponding to the execution decision is the decision entered by the user in the decision window and the number of decisions n=1, the execution feedback module pushes a change decision window to the user, takes the decision entered by the user as the changed execution decision, and sends the changed execution decision to the scheduling execution module for execution until the execution feedback module reports that the execution is normal. If the scheduling decision corresponding to the execution decision is a decision with n > 1, the execution feedback module will take the next decision in the decision ranking as the modified execution decision. The execution feedback module will send the modified execution decision to the scheduling execution module for execution until the execution feedback module reports that the execution is normal. When there is no next decision in the decision ranking, the execution feedback module will push a change decision window to the user and take the decision entered by the user as the modified execution decision.

9. The control system for parallel transformers based on a new energy flexible DC system according to claim 8, characterized in that, The execution feedback module obtains the anomaly frequency p, compares p with each preset anomaly frequency, and judges the adjustment situation based on the comparison result, wherein: When p≤p1, the execution feedback module determines that no adjustment should be made; When p1 < p ≤ p2, the execution feedback module determines to make adjustments; When p2 < p, the execution feedback module determines that no adjustment should be made, the execution feedback module connects to the redundant device, and sends an abnormal frequency alarm to the user; p1 is the first preset abnormal frequency, p2 is the second preset abnormal frequency, and 0 < p1 < p2.

10. The control system for parallel transformers based on a new energy flexible DC system according to claim 9, characterized in that, When the execution feedback module determines that an adjustment is needed, it sets the adjustment coefficient Z = 0.85 + e. -(p-p1)-2 e is the base of the natural logarithm. The execution feedback module adjusts the preset abnormal duration t0 in the execution feedback process according to the adjustment coefficient. The adjusted preset abnormal duration is tz0, where tz0 = Z × t0.

Images