High-voltage power grid voltage quality compensation method based on passive control strategy

Abstract

The invention discloses a high-voltage power grid voltage quality compensation method based on a passive control strategy, and relates to the technical field of power grid voltage control, and the method comprises the steps: reading the data of an energy recovery type passive control system, and constructing a digital twinborn model; performing control fitting evaluation based on a twinborn model, distributing monitoring sensors and establishing mapping in the model; establishing time sequence monitoring data through a monitoring sensor, synchronizing the time sequence monitoring data to a short-term power grid state prediction network in the twin model, and establishing a prediction data set; analyzing the automatic compensation strategy, performing regulation analysis on the dynamic voltage regulator, and establishing a regulation data set; acquiring energy storage data, and establishing a compensation objective function; and inputting the prediction data set, the adjustment data set, the time sequence monitoring data and the energy storage data into the compensation objective function, executing control optimization, and performing power grid voltage quality compensation. Therefore, the technical effects of simple control, control cost reduction and compensation response efficiency improvement are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid voltage control, and in particular to a high-voltage power grid voltage quality compensation method based on a passive control strategy. Background Art

[0002] The voltage quality of high-voltage power grid is one of the key factors affecting the stability and reliability of power supply. With the widespread application of renewable energy, the volatility and uncertainty problems faced by the power grid are becoming increasingly prominent. The existing compensation control methods are mostly based on PID, which requires a large number of controllers and complex control parameter settings. There are technical problems such as difficult parameter selection, high control cost, and slow compensation response. Summary of the invention

[0003] The present invention provides a high-voltage power grid voltage quality compensation method based on a passive control strategy to solve the technical problems of difficult parameter selection, high control cost and slow compensation response in the prior art, and achieves the technical effects of simple control, reduced control cost and improved compensation response efficiency.

[0004] The high-voltage power grid voltage quality compensation method based on the passive control strategy provided by the present invention is applied to an energy recovery passive control system, wherein the energy recovery passive control system includes an energy storage unit, a dynamic voltage regulator, and a multi-mode controller. The method includes:

[0005] Digital communication with an energy recovery type passive control system is established, system data of the energy recovery type passive control system is read based on the digital communication, and a digital twin model is constructed.

[0006] A control fitting evaluation of the energy recovery passive control system is performed based on the digital twin model, monitoring sensors are distributed according to the control fitting evaluation results, and a mapping of the monitoring sensors within the digital twin model is established.

[0007] Time series monitoring data is established through the monitoring sensor, and the time series monitoring data is synchronized to the short-term power grid state prediction network in the digital twin model through mapping to establish a prediction data set. The prediction data of the short-term power grid state prediction network includes the behavior data of the power grid and the output data of renewable energy.

[0008] The multi-mode controller is analyzed to obtain an automatic compensation strategy of the multi-mode controller, and a regulation analysis of the dynamic voltage regulator is performed based on the automatic compensation strategy to establish a regulation data set.

[0009] The energy storage data of the energy storage unit is obtained, and a compensation objective function is established.

[0010] The prediction data set, the adjustment data set, the time series monitoring data, and the energy storage data are input into the compensation objective function, control optimization is performed, the optimization result is established, and the grid voltage quality compensation is performed according to the optimization result.

[0011] In a feasible implementation, the step of inputting the prediction data set, the adjustment data set, the time series monitoring data, and the energy storage data into the compensation objective function further includes:

[0012] The prediction data set is parsed to obtain time series behavior prediction data and time series output prediction data.

[0013] The regulation data set is parsed to obtain a regulation voltage of a dynamic voltage regulator.

[0014] The time series monitoring data is analyzed to obtain the power demand of the power grid, and the time series behavior prediction data and time series output prediction data, adjustment voltage, power demand of the power grid, and energy storage data are used as extracted data features and input into the compensation objective function.

[0015] In a feasible implementation, the compensation objective function is as follows:

[0016]

[0017]

[0018] Among them, J represents the compensation objective function, T is the monitoring and evaluation period, and u ref Characterizes the voltage level maintained under ideal conditions of the power grid, u comp (t) is the regulated voltage at time t, P renew,pred (t) is the time series output prediction data at time t, P renew,actual (t) is the actual output data of renewable energy at time t, B represents the closed cycle within the monitoring and evaluation period T, E pred (t) is the predicted energy storage data at time t, E actual (t) is the actual energy storage data at time t, P demand (t) represents the power demand of the power grid at time t, is the time series behavior prediction data, and α, β, γ, δ, and δ are weight factors, which are used to adjust voltage stability, renewable energy data prediction accuracy, energy storage prediction accuracy, the matching degree between grid demand and actual renewable energy output, and the matching degree between predicted demand and predicted renewable energy output, respectively.

[0019] In a feasible implementation, the step of synchronizing the time series monitoring data to a short-term power grid state prediction network in a digital twin model through mapping to establish a prediction data set also includes:

[0020] The renewable energy analysis subnetwork of the short-term power grid state prediction network is called, and weather data is read through the renewable energy analysis subnetwork to establish auxiliary input data.

[0021] Through the renewable energy analysis subnetwork, auxiliary input data and time series monitoring data are analyzed and predicted, the output data of renewable energy is established, and a prediction data set is established based on the output data of renewable energy.

[0022] In a feasible implementation, the step of establishing a prediction data set according to the output data of renewable energy further includes:

[0023] The behavior prediction subnetwork of the short-term power grid state prediction network is called, and after receiving the time series monitoring data through the behavior prediction subnetwork, the time series monitoring data is time-series segmented to establish a time series segmentation result.

[0024] An adaptive feature extraction window is configured based on the time series segmentation result, long and short time series features of the time series segmentation result are extracted through the adaptive feature extraction window, behavior prediction is performed based on the long and short time series feature extraction results, the behavior data of the power grid is established, and a prediction data set is established based on the behavior data of the power grid and the output data of renewable energy.

[0025] In a feasible implementation, the method further includes:

[0026] A time-series electricity price set is established, and a balancing model is configured, wherein the balancing model is used to balance voltage stability and cost efficiency.

[0027] The time-series electricity price set and the optimization result are used as input data, and a balance analysis is performed through a balance model to establish a balance analysis result.

[0028] The optimization results are adjusted according to the balance analysis results, and the adjusted optimization results are used to compensate for the grid voltage quality.

[0029] In a feasible implementation, the method further includes:

[0030] The optimization result is monitored after compensation to establish a monitoring result.

[0031] The monitoring results are optimized and controlled for authentication, and control feedback data are established through the optimized control authentication results.

[0032] Self-learning feedback optimization of the multi-modal controller is performed through the control feedback data.

[0033] In a feasible implementation, establishing the monitoring result further includes:

[0034] Determine whether any of the monitoring results meets a preset abnormal threshold.

[0035] If any of the monitoring results meets a preset abnormal threshold, a control abnormality is reported.

[0036] The present invention discloses a high-voltage power grid voltage quality compensation method based on a passive control strategy, including: reading data of an energy recovery passive control system through digital communication to construct a digital twin model. Based on the digital twin model, a control fitting evaluation is performed, and a monitoring sensor is distributed according to the result, and the sensor position is mapped in the model. Time series monitoring data is established by monitoring sensors, and synchronized to a short-term power grid state prediction network of a digital twin model to generate a prediction data set, including power grid behavior data and renewable energy output data. A multi-modal controller is analyzed to obtain an automatic compensation strategy, and a dynamic voltage regulator analysis is performed based on this to establish a regulation data set. Storage data of an energy storage unit is obtained to establish a compensation objective function. The prediction data set, the regulation data set, the time series monitoring data and the energy storage data are input into the compensation objective function, control optimization is performed, an optimization result is generated, and power grid voltage quality compensation is performed. The high-voltage power grid voltage quality compensation method based on a passive control strategy disclosed by the present invention solves the technical problems of difficult parameter selection, high control cost and slow compensation response, and achieves the technical effects of simple control, reduced control cost and improved compensation response efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a flow chart of a high-voltage power grid voltage quality compensation method based on a passive control strategy of the present invention;

[0038] Figure 2 The present invention is a flow chart of establishing a prediction data set in a high-voltage power grid voltage quality compensation method based on a passive control strategy. DETAILED DESCRIPTION

[0039] The technical solution provided in the embodiments of the present invention is to solve the technical problems of difficult parameter selection, high control cost and slow compensation response in the prior art. The overall idea adopted is as follows:

[0040] First, digital communication with an energy recovery type passive control system is established, system data of the energy recovery type passive control system is read based on the digital communication, and a digital twin model is constructed.

[0041] Then, a control fitting evaluation of the energy recovery passive control system is performed based on the digital twin model, monitoring sensors are distributed according to the control fitting evaluation results, and a mapping of the monitoring sensors in the digital twin model is established.

[0042] Then, the monitoring sensor is used to establish time series monitoring data, and the time series monitoring data is synchronized to the short-term power grid state prediction network in the digital twin model through mapping to establish a prediction data set. The prediction data of the short-term power grid state prediction network includes the behavior data of the power grid and the output data of renewable energy.

[0043] Next, the multi-mode controller is analyzed to obtain an automatic compensation strategy of the multi-mode controller, and a regulation analysis of the dynamic voltage regulator is performed based on the automatic compensation strategy to establish a regulation data set.

[0044] Furthermore, energy storage data of the energy storage unit is obtained, and a compensation objective function is established.

[0045] Finally, the prediction data set, the adjustment data set, the time series monitoring data, and the energy storage data are input into the compensation objective function, the control optimization is performed, the optimization result is established, and the grid voltage quality compensation is performed according to the optimization result.

[0046] The above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods of the specification to better understand the above technical solution. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments of the present invention. It should be understood that the present invention is not limited to the example embodiments used only to explain the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. In addition, it should be noted that, for the convenience of description, only the parts related to the present invention are shown in the drawings, rather than all of them.

[0047] Embodiment 1

[0048] Figure 1 The present invention is a flow chart of a high-voltage power grid voltage quality compensation method based on a passive control strategy, the method is applied to an energy recovery passive control system, the energy recovery passive control system includes an energy storage unit, a dynamic voltage regulator, and a multi-mode controller, the method includes:

[0049] Digital communication with an energy recovery type passive control system is established, system data of the energy recovery type passive control system is read based on the digital communication, and a digital twin model is constructed.

[0050] Specifically, digital communication with the energy recovery passive control system is established, including determining the communication protocol and interface type (such as Modbus, CAN, Ethernet, etc.), and configuring communication parameters (such as baud rate, data bit, stop bit, check method, etc.), so as to establish a communication connection with the energy recovery passive control system.

[0051] Specifically, the system data of the energy recovery passive control system is read, including hardware configuration, operating parameters, system status, historical performance, etc. Exemplarily, the system data is obtained from sources such as sensors, controllers, and log files of the energy recovery passive control system.

[0052] Among them, the energy storage unit includes multiple energy storage elements (such as capacitors) for storing and releasing electrical energy. When the power grid experiences voltage fluctuations or short-term voltage interruptions, the energy storage unit can quickly release electrical energy to maintain the stability of the voltage on the load side. The dynamic voltage regulator (DVR) can dynamically adjust the voltage to offset the voltage disturbance in the power grid, and is used to improve the voltage quality in the power system and ensure the stability of the voltage on the load side. The multimodal controller is the control center of the energy recovery passive control system, which is used to control the operation of the entire system. The multimodal controller adjusts the action of the energy storage unit and the DVR according to the real-time status of the power grid and the load to ensure voltage stability and efficient use of electrical energy.

[0053] Specifically, to build a digital twin model, first, create a physical model of the energy recovery passive control system to accurately describe the physical characteristics, connection relationships, and component configurations of the system. Then, based on data analysis and machine learning techniques, build a data-driven model to update the state of the physical model according to real-time data. Next, integrate the above physical model and data-driven model to form a digital twin model that can update the model state according to the received real-time data.

[0054] A control fitting evaluation of the energy recovery passive control system is performed based on the digital twin model, monitoring sensors are distributed according to the control fitting evaluation results, and a mapping of the monitoring sensors within the digital twin model is established.

[0055] Specifically, the control fitting evaluation of the energy recovery passive control system is carried out based on the digital twin model. First, the actual operation of the control system is simulated by the digital twin model, and the predicted system state and performance are generated. Then, these prediction results are compared with the system state and performance under ideal conditions (such as the voltage amplitude and harmonic content of the power grid), and the fitting error is calculated. The fitting error reflects the performance of the evaluated control system and the accuracy of the digital twin model.

[0056] Furthermore, a control fitting evaluation is performed based on the fitting error, and based on the results of the control fitting evaluation, it is determined which parts of the system have good performance and which parts have poor performance. Then, the distribution of sensors is adaptively adjusted to more effectively and more specifically monitor the state and performance of the system. Exemplary methods include increasing the density of monitoring sensors for most of the voltage fluctuations, increasing the acquisition frequency of monitoring sensors, etc.

[0057] Specifically, after adjusting the sensor distribution, a sensor mapping is established to update the location and function of multiple sensors in the digital twin model. The sensor mapping describes the location and function of each sensor in the model. Exemplarily, the sensor mapping is determined based on the unique device identifier of the monitoring sensor (such as identity ID, SN, IMEI, etc.), the deployment location of the monitoring sensor (physical address and logical address), etc., and the sensor mapping is represented as a database or structured text information. The establishment of sensor mapping helps to efficiently find relevant sensor information when maintaining and debugging the system, analyzing data, and updating the model.

[0058] Time series monitoring data is established through the monitoring sensor, and the time series monitoring data is synchronized to the short-term power grid state prediction network in the digital twin model through mapping to establish a prediction data set. The prediction data of the short-term power grid state prediction network includes the behavior data of the power grid and the output data of renewable energy.

[0059] Specifically, multiple monitoring sensors distributed in the target power grid are activated to collect information such as voltage, frequency, phase, etc. of the target power grid, as well as data related to renewable energy output, such as wind speed, light intensity, temperature, etc., and store them as time series monitoring data. Then, the collected time series monitoring data is associated with the corresponding part of the model through the constructed sensor mapping, and mapped and synchronized to the short-term power grid status prediction network in the digital twin model.

[0060] Among them, the prediction network is used to predict the short-term state of the power grid, and the output prediction data includes the behavior data of the power grid (such as the predicted values ​​of voltage, current, frequency, etc.) and the output data of renewable energy (such as the predicted output of wind turbines or solar panels). Exemplarily, the prediction network is a lightweight prediction network built based on neural networks and knowledge distillation. It learns the historical time series monitoring data of the target power grid through a large, highly complex neural network model, learns the behavior pattern of the target power grid and the output characteristics of renewable energy, and builds a smaller and lighter neural network model, so that the lightweight model imitates the output of the large model to learn some internal representations and decision-making processes of the teacher model, and then obtain short-term prediction capabilities. In this way, the accuracy of prediction can be achieved on devices with limited resources.

[0061] In some embodiments, Figure 2 As shown, the time series monitoring data is synchronized to the short-term power grid state prediction network in the digital twin model by mapping to establish a prediction data set, and also includes:

[0062] The renewable energy analysis subnetwork of the short-term power grid state prediction network is called, and weather data is read through the renewable energy analysis subnetwork to establish auxiliary input data.

[0063] Through the renewable energy analysis sub-network, auxiliary input data and time series monitoring data are analyzed and predicted, the output data of renewable energy is established, and a prediction data set is established based on the output data of renewable energy.

[0064] Specifically, the renewable energy analysis subnetwork in the short-term power grid state prediction network is first called. The renewable energy analysis subnetwork is used to read and analyze weather data, such as wind speed, light intensity, temperature, etc., to establish auxiliary input data. Exemplarily, the renewable energy analysis subnetwork is connected to a weather data source or a weather data platform to obtain real-time weather data of the environment where the target power grid is located.

[0065] Furthermore, the renewable energy analysis subnetwork uses the auxiliary input data and the time series monitoring data as input data for analysis and prediction, and establishes the output data of renewable energy, which reflects the expected output of renewable energy equipment (such as wind turbines or solar panels) under given weather conditions. The renewable energy analysis subnetwork includes an analysis model based on a neural network and a mathematical prediction model based on the principle of renewable energy regeneration.

[0066] Exemplarily, the neural network-based analytical model learns complex nonlinear relationships from a large amount of training data to predict future energy output. Specifically, the training data includes weather data, equipment status data, historical energy output data, etc. The mathematical prediction model based on the principle of renewable energy regeneration is based on the physical principles of renewable energy, such as the photovoltaic effect of solar cells, the aerodynamics of wind turbines, etc., and can provide physically explainable predictions and use weather data (such as light intensity, wind speed, etc.) to calculate the expected energy output.

[0067] In some implementations, establishing a prediction data set based on the output data of the renewable energy source further includes:

[0068] The behavior prediction subnetwork of the short-term power grid state prediction network is called, and after receiving the time series monitoring data through the behavior prediction subnetwork, the time series monitoring data is time-series segmented to establish a time series segmentation result.

[0069] An adaptive feature extraction window is configured based on the time series segmentation result, long and short time series features of the time series segmentation result are extracted through the adaptive feature extraction window, behavior prediction is performed based on the long and short time series feature extraction results, the behavior data of the power grid is established, and a prediction data set is established based on the behavior data of the power grid and the output data of renewable energy.

[0070] Specifically, the behavior prediction subnetwork of the short-term power grid state prediction network reflects the periodic variation law of the target power grid's power consumption behavior characteristics. In other words, the behavior prediction subnetwork is used to describe how the target power grid's power consumption behavior fluctuates over time. Exemplarily, the behavior prediction subnetwork learns the periodic pattern of power grid power consumption behavior by analyzing historical power grid power consumption data. This includes daily cycles (e.g., power consumption behavior during the day and at night may be very different), weekly cycles (e.g., power consumption behavior during weekdays and weekends may be very different), and annual cycles (e.g., power consumption behavior in summer and winter may be very different).

[0071] Specifically, after receiving the time series monitoring data, the behavior prediction sub-network performs time series segmentation on the time series monitoring data, dividing the continuous data stream into a series of time windows, where each window contains data within a period of time. Then, an adaptive feature extraction window is configured according to the time series segmentation result. The adaptive feature extraction window can automatically adjust its size to best adapt to data changes.

[0072] Exemplarily, the size of the adaptive feature extraction window is determined based on the information entropy of the time series monitoring data. First, the information entropy of the time series monitoring data is calculated based on the voltage distribution or the distribution of the harmonic proportion. Then, the size of the adaptive feature extraction window is determined based on the information entropy. If the information entropy is high (the data complexity or uncertainty is large), a larger window size is selected to capture more data features. On the contrary, if the information entropy is low (the data complexity or uncertainty is small), a smaller window size is selected to adapt to data changes more quickly.

[0073] Specifically, through the adaptive feature extraction window, the long and short time series feature extraction is performed on the time series segmentation results to extract the key features in the data. The long and short time series feature extraction results can reflect the long-term and short-term change trends of the data. Then, based on the long and short time series feature extraction results, the behavior prediction subnetwork is input for behavior prediction to obtain the behavior data of the power grid. Through the behavior data of the power grid, the future behavior and state of the target power grid can be predicted, including the load, voltage, harmonics, frequency, etc. of the power grid.

[0074] Exemplarily, long and short time series feature extraction is performed on the results of time series segmentation, including extraction of various statistical features (such as mean, variance, etc.), trend features (such as rising or falling trends), periodic features (such as daily cycles, weekly cycles, annual cycles), etc.

[0075] The multi-mode controller is analyzed to obtain an automatic compensation strategy of the multi-mode controller, and a regulation analysis of the dynamic voltage regulator is performed based on the automatic compensation strategy to establish a regulation data set.

[0076] Specifically, the multi-modal controller is analyzed, including obtaining its internal control logic and strategy, such as analyzing the controller's control algorithm, parameter settings, mode switching logic, etc., and then determining the automatic compensation strategy.

[0077] Exemplarily, the automatic compensation strategy is a passive control strategy based on an EL model, wherein the EL model is an energy and delay model used to describe the energy consumption and delay of a passive control system. Factors involved in the EL model include the system's operating mode, workload, environmental conditions, etc.

[0078] Furthermore, the regulation analysis of the dynamic voltage regulator is performed based on the automatic compensation strategy. This includes simulating the response of the controller, analyzing its compensation effect on voltage fluctuations, evaluating its performance under different conditions, and then establishing a regulation data set, which contains multiple sets of associated stored regulation effects, regulation control parameters, and system status data. By establishing a regulation data set, a set of compensation schemes with clear compensation performance that can be quickly called is provided for subsequent compensation, which helps to improve the response speed and compensation quality of subsequent compensation.

[0079] The energy storage data of the energy storage unit is obtained, and a compensation objective function is established.

[0080] Specifically, the energy storage data of the energy storage unit refers to the size and distribution of the amount of electric energy stored in multiple capacitors in the energy storage unit. The compensation objective function is an objective function for optimizing the voltage quality compensation effect, including multiple optimization objectives. Exemplarily, the optimization objectives include maximizing energy utilization, minimizing energy consumption, minimizing delay, etc.

[0081] In some embodiments, the compensation objective function is as follows:

[0082]

[0083] Among them, J represents the compensation objective function, T is the monitoring and evaluation period, and u ref Characterizes the voltage level maintained under ideal conditions of the power grid, u comp (t) is the regulated voltage at time t, P renew,pred (t) is the time series output prediction data at time t, P renew,actual (t) is the actual output data of renewable energy at time t, B represents the closed cycle within the monitoring and evaluation period T, E pred (t) is the predicted energy storage data at time t, E actual (t) is the actual energy storage data at time t, P demand (t) represents the power demand of the power grid at time t, is the time series behavior prediction data, α, β, γ, δ, ∈ are weight factors, which are used to adjust voltage stability, renewable energy data prediction accuracy, energy storage prediction accuracy, the matching degree between grid demand and actual renewable energy output, and the matching degree between predicted demand and predicted renewable energy output, respectively.

[0084] Specifically, the compensation objective function is a composite objective function that takes into account multiple factors such as voltage stability, renewable energy data prediction accuracy, energy storage prediction accuracy, the matching degree between grid demand and actual renewable energy output, and the matching degree between predicted demand and predicted renewable energy output, so as to evaluate and optimize the performance of voltage quality compensation from multiple perspectives.

[0085] Specifically, the compensation objective function optimizes the performance of voltage quality compensation by minimizing the sum of squares of various errors (errors between multiple grid voltage indicators and ideal values). Each error corresponds to a specific performance indicator, such as voltage stability, prediction accuracy, etc. Moreover, each error is configured with a corresponding weight factor to reflect the importance of different performance indicators.

[0086] The prediction data set, the adjustment data set, the time series monitoring data, and the energy storage data are input into the compensation objective function, control optimization is performed, the optimization result is established, and the grid voltage quality compensation is performed according to the optimization result.

[0087] In some embodiments, the step of inputting the prediction data set, the adjustment data set, the time series monitoring data, and the energy storage data into the compensation objective function further includes:

[0088] The prediction data set is parsed to obtain time series behavior prediction data and time series output prediction data.

[0089] The regulation data set is parsed to obtain a regulation voltage of a dynamic voltage regulator.

[0090] The time series monitoring data is analyzed to obtain the power demand of the power grid, and the time series behavior prediction data and time series output prediction data, adjustment voltage, power demand of the power grid, and energy storage data are used as extracted data features and input into the compensation objective function.

[0091] Specifically, the control strategy can be adjusted to make the value of the compensation objective function as small as possible, which includes, for example, adjusting the energy storage unit charging / discharging strategy, adjusting the system's operating mode, adjusting the parameters of the prediction model, and the like.

[0092] Optionally, the regulation data set is used as the optimized parameter space, and through an optimization algorithm (such as gradient descent, genetic algorithm, simulated annealing, etc.), the regulation data set is traversed to iterate the regulation voltages of multiple dynamic voltage regulators, and combined with the timing behavior prediction data and timing output prediction data, grid demand power, and energy storage data, and the compensation objective function is input to calculate the function value, and the control strategy that minimizes the compensation objective function is found, and the output is the optimization result.

[0093] Furthermore, the method further comprises:

[0094] A time-series electricity price set is established, and a balancing model is configured, wherein the balancing model is used to balance voltage stability and cost efficiency.

[0095] The time-series electricity price set and the optimization result are used as input data, and a balance analysis is performed through a balance model to establish a balance analysis result.

[0096] The optimization results are adjusted according to the balance analysis results, and the adjusted optimization results are used to compensate for the grid voltage quality.

[0097] Specifically, in actual grid operation, it is necessary to ensure voltage stability and consider the cost of operation, so a balance is found between voltage stability and cost efficiency by configuring the balance model. Among them, the time-series electricity price set is obtained through the grid operator or the market, including the electricity price for each period.

[0098] Exemplarily, the configuration balance model is a multi-objective optimization model, in which one objective is voltage stability and the other objective is cost efficiency. The two objectives are configured with corresponding weight coefficients for adjusting the importance of the two objectives. Specifically, the weight coefficients are determined based on the operation requirements of the target power grid. For example, if the target power grid is more concerned with reducing the operation cost, a higher level of weight coefficient is configured for cost efficiency.

[0099] Specifically, the time-series electricity price set and the optimization result are used as input data, analyzed through the balance model, and the balance analysis result is established according to the output of the balance model. The balance analysis result is a numerical analysis result, which is used to quantitatively evaluate the degree of balance of the optimization result in achieving the target grid voltage stability and cost efficiency requirements.

[0100] Specifically, if the balance analysis results show that the optimization results do not meet the balance analysis threshold corresponding to the target power grid demand, the optimization results are adjusted according to the contribution ratio of the power grid voltage stability and cost efficiency to the balance analysis results in the analysis results. For example, if the contribution ratio of cost efficiency to the balance analysis results is low, it means that the operating cost is much lower than the expected value of the target power grid, and then the voltage regulation strategy is adjusted to reduce the cost. Exemplary methods include reducing the operating frequency of the voltage regulator, optimizing the working mode of the voltage regulator, etc. The method of adjusting the optimization results according to the balance analysis results can ensure that the power grid can effectively control the operating costs while ensuring voltage stability.

[0101] Furthermore, the method further comprises:

[0102] The optimization result is monitored after compensation to establish a monitoring result.

[0103] The monitoring results are optimized and controlled for authentication, and control feedback data are established through the optimized control authentication results.

[0104] Specifically, the target power grid is continuously monitored based on the monitoring sensor, the compensation effect data of the target power grid is obtained, and the monitoring results are established. Exemplarily, the monitoring results include the voltage amplitude, voltage stability, harmonic proportion, etc. of the target power grid after compensation. Then, through in-depth analysis and evaluation of the monitoring results, the effectiveness and accuracy of the optimization results are verified to perform optimization control certification, and control feedback data is established. Optimization control certification is used to quantitatively evaluate the voltage quality compensation efficiency of the optimization results. Through the above-mentioned method steps, the quality of the power grid voltage can be further ensured, and possible problems in compensation can be discovered in a timely manner.

[0105] In some implementations, establishing the monitoring results further includes:

[0106] Determine whether any of the monitoring results meets a preset abnormal threshold.

[0107] If any of the monitoring results meets a preset abnormal threshold, a control abnormality is reported.

[0108] Further, it is determined whether there are any monitoring results that meet the preset abnormal threshold. If there are any monitoring results that meet the preset abnormal threshold, then a control abnormality is reported, and a control abnormality response is performed according to the reported control abnormality, including suspending compensation, adjusting compensation strategy, etc.

[0109] In summary, the high-voltage power grid voltage quality compensation method based on the passive control strategy provided by the present invention has the following technical effects:

[0110] The data of the energy recovery passive control system is read through digital communication to build a digital twin model. The control fitting evaluation is performed based on the digital twin model, and the monitoring sensors are distributed according to the results, and the sensor positions are mapped in the model. The time series monitoring data is established by monitoring the sensors, and synchronized to the short-term grid state prediction network of the digital twin model to generate a prediction data set, including grid behavior data and renewable energy output data. The multi-modal controller is analyzed to obtain the automatic compensation strategy, and the dynamic voltage regulator is analyzed based on this to establish a regulation data set. The storage data of the energy storage unit is obtained to establish a compensation objective function. The prediction data set, regulation data set, time series monitoring data and energy storage data are input into the compensation objective function, control optimization is performed, optimization results are generated, and grid voltage quality compensation is performed. Thereby, the technical effects of simple control, reduced control cost and improved compensation response efficiency are achieved.

[0111] It should be understood that the embodiments disclosed in the present invention and the above description can enable those skilled in the art to use the present invention to implement the present invention. At the same time, the present invention is not limited to the above-mentioned embodiments. It should be understood that those skilled in the art can still modify the technical solutions recorded in the above-mentioned embodiments, or replace some of the technical features therein by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention.

Claims

1. A high voltage power grid voltage quality compensation method based on a passive control strategy, characterized in that: The method is applied to an energy recovery type passive control system, the energy recovery type passive control system includes an energy storage unit, a dynamic voltage regulator, and a multi-mode controller, and the method includes: Establishing digital communication with the energy recovery type passive control system, reading system data of the energy recovery type passive control system based on the digital communication, and building a digital twin model; Performing a control fitting evaluation of the energy recovery passive control system based on the digital twin model, distributing monitoring sensors according to the control fitting evaluation results, and establishing a mapping of the monitoring sensors within the digital twin model; Establishing time series monitoring data through the monitoring sensor, synchronizing the time series monitoring data to the short-term power grid state prediction network in the digital twin model through mapping, and establishing a prediction data set, wherein the prediction data of the short-term power grid state prediction network includes behavior data of the power grid and output data of renewable energy; Analyzing the multi-mode controller, obtaining an automatic compensation strategy of the multi-mode controller, and performing regulation analysis of the dynamic voltage regulator based on the automatic compensation strategy to establish a regulation data set; Acquire energy storage data of the energy storage unit and establish a compensation objective function; The prediction data set, the adjustment data set, the time series monitoring data, and the energy storage data are input into the compensation objective function, control optimization is performed, the optimization result is established, and the grid voltage quality compensation is performed according to the optimization result.

2. The high voltage power grid voltage quality compensation method based on passive control strategy according to claim 1, characterized in that: The step of inputting the prediction data set, the adjustment data set, the time series monitoring data, and the energy storage data into the compensation objective function further includes: Parsing the prediction data set to obtain time series behavior prediction data and time series output prediction data; parsing the regulation data set to obtain a regulation voltage of a dynamic voltage regulator; The time series monitoring data is analyzed to obtain the power demand of the power grid, and the time series behavior prediction data and time series output prediction data, adjustment voltage, power demand of the power grid, and energy storage data are used as extracted data features and input into the compensation objective function.

3. The high voltage power grid voltage quality compensation method based on passive control strategy according to claim 2, characterized in that: The compensation objective function is as follows: Among them, J represents the compensation objective function, T is the monitoring and evaluation period, and u ref Characterizes the voltage level maintained under ideal conditions of the power grid, u comp (t) is the regulated voltage at time t, P renew,pred (t) is the time series output prediction data at time t, P renew,actual (t) is the actual output data of renewable energy at time t, B represents the closed cycle within the monitoring and evaluation period T, E pred (t) is the predicted energy storage data at time t, E actual (t) is the actual energy storage data at time t, P demand (t) represents the power demand of the power grid at time t, is the time series behavior prediction data, α, β, γ, δ, ∈ are weight factors, which are used to adjust voltage stability, renewable energy data prediction accuracy, energy storage prediction accuracy, the matching degree between grid demand and actual renewable energy output, and the matching degree between predicted demand and predicted renewable energy output, respectively.

4. The high voltage power grid voltage quality compensation method based on passive control strategy according to claim 1, characterized in that: The step of synchronizing the time series monitoring data to a short-term power grid state prediction network in the digital twin model through mapping to establish a prediction data set also includes: Calling the renewable energy analysis subnetwork of the short-term power grid state prediction network, reading weather data through the renewable energy analysis subnetwork, and establishing auxiliary input data; Through the renewable energy analysis sub-network, auxiliary input data and time series monitoring data are analyzed and predicted, the output data of renewable energy is established, and a prediction data set is established based on the output data of renewable energy.

5. The high voltage power grid voltage quality compensation method based on passive control strategy according to claim 4, characterized in that: The step of establishing a prediction data set based on the output data of renewable energy further includes: Calling the behavior prediction subnetwork of the short-term power grid state prediction network, after receiving the time series monitoring data through the behavior prediction subnetwork, performing time series segmentation on the time series monitoring data and establishing a time series segmentation result; An adaptive feature extraction window is configured based on the time series segmentation result, long and short time series features of the time series segmentation result are extracted through the adaptive feature extraction window, behavior prediction is performed based on the long and short time series feature extraction results, the behavior data of the power grid is established, and a prediction data set is established based on the behavior data of the power grid and the output data of renewable energy.

6. The high voltage power grid voltage quality compensation method based on passive control strategy according to claim 1, characterized in that: The method further comprises: Establishing a time-series electricity price set and configuring a balancing model, wherein the balancing model is used to balance voltage stability and cost efficiency; Taking the time-series electricity price set and the optimization result as input data, performing a balance analysis through a balance model, and establishing a balance analysis result; The optimization results are adjusted according to the balance analysis results, and the adjusted optimization results are used to compensate for the grid voltage quality.

7. The high voltage power grid voltage quality compensation method based on passive control strategy according to claim 1, characterized in that: The method further comprises: Monitoring the optimization result after compensation to establish a monitoring result; Performing optimized control authentication on the monitoring results, and establishing control feedback data through the optimized control authentication results; Self-learning feedback optimization of the multi-modal controller is performed through the control feedback data.

8. The high voltage power grid voltage quality compensation method based on passive control strategy according to claim 7, characterized in that: The establishing of monitoring results also includes: Determine whether the monitoring results meet a preset abnormal threshold; If any of the monitoring results meets a preset abnormal threshold, a control abnormality is reported.

Images