A on-load tap-changer control system and method for a converter transformer

Abstract

The present invention discloses a on-load tap-changer control system and method for a converter transformer, which relates to the technical field of tap-changer control of a converter transformer, and includes the following steps: Through high-precision sensors and a data acquisition system, the key operating parameters of the converter transformer are monitored in real time at each moment. Based on the collected operating parameter data, a time series model of the operating state of the converter transformer is constructed, and the model is used to describe the law of change of the operating parameters over time and capture the dynamic characteristics of the operating state; The constructed time series model is used to predict the change trend of the operating state of the converter transformer in a future period of time. The present invention reasonably arranges the tap-changing timing of the converter transformer through the time series model and the dynamic programming optimization method, reduces operations and mechanical wear, and extends the service life of the equipment. At the same time, a power loss quantization index is introduced into the cost function to optimize the switching strategy and screen out the scheme with the minimum power loss, thereby improving the economy and stability of the system.

Description

Technical Field

[0001] The present invention relates to the technical field of converter transformer tap-changer control, and particularly to a on-load tap-changer control system and method for a converter transformer. Background Art

[0002] In the prior art, the specific principle of on-load tap-changer (OLTC) control of a converter transformer is usually based on a closed-loop feedback control system. By real-time monitoring the output voltage and load current of the transformer, when it is detected that the voltage deviates from the set target value (for example, higher or lower than the set threshold range), the control system will trigger the tap-changer to switch to an appropriate tap to adjust the output voltage of the transformer and ensure that the voltage is stable within a predetermined range.

[0003] The prior art has the following deficiencies:

[0004] In the actual use process, the on-load tap-changer (OLTC) control system of the prior art often lacks the ability to predict future changes in the operating state, mainly relying on the voltage and load data monitored in real time for simple feedback control. This method is difficult to fully optimize the tap-changing strategy, easily leads to frequent switching, increases the mechanical wear of the equipment, and shortens the service life. In addition, improper switching timing may lead to an increase in power loss, reducing the overall efficiency and economy of the system. These problems are particularly serious when the load fluctuates frequently or the grid conditions are complex, which may endanger the stability of the power system and affect the long-term reliability of the equipment.

[0005] The above information disclosed in the background art section is only used to enhance the understanding of the background of the present disclosure, and therefore it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0006] The object of the present invention is to provide a on-load tap-changer control system and method for a converter transformer. By constructing a time series model and a dynamic programming optimization method, the tap-changing timing of the converter transformer is predicted in advance and reasonably arranged, unnecessary operations are reduced, mechanical wear is reduced, the equipment life is extended, and the long-term reliability of the system is improved. At the same time, by introducing a power loss quantification index into the cost function and globally optimizing the switching strategy, the switching scheme with the minimum power loss is selected, the economy and energy efficiency of the system are improved, and the stability of the power system is enhanced to solve the problems in the above background art.

[0007] To achieve the above object, the present invention provides the following technical solution: A on-load tap-changer control method for a converter transformer, comprising the following steps:

[0008] Through high-precision sensors and data acquisition systems, the key operating parameters of the converter transformer are monitored in real time. Based on the collected operating parameter data, a time series model of the operating state of the converter transformer is constructed to describe the law of change of the operating parameters over time through the model and capture the dynamic characteristics of the operating state;

[0009] Using the constructed time series model, predict the change trend of the operating state of the converter transformer over a period of time in the future;

[0010] According to the predicted change trend of the operating state, determine the optimal time point for switching the tap of the converter transformer;

[0011] For each determined switching time point, obtain the power loss difference before and after the tap change and define it as the switching cost;

[0012] Construct a cost function to evaluate the power loss at each switching time point. By considering the minimization of power loss, the power loss difference before and after the switch is used as the optimization objective, calculate the power loss for all switching time points, and screen out the switching scheme with the minimum power loss;

[0013] Using the dynamic programming optimization method, solve the cost function to obtain the optimal tap-changing strategy. By decomposing the power loss problem into multiple sub-problems, gradually optimize the switching strategy at each time point to ensure that the power loss of the system is minimized at each moment;

[0014] According to the optimal tap-changing strategy obtained by dynamic programming optimization, control the switching action of the on-load tap-changer in real time and convert the optimal strategy into actual operation instructions.

[0015] Preferably, based on the collected operating parameter data of the converter transformer, the specific steps for constructing the time series model of the operating state of the converter transformer are as follows:

[0016] Select a high-precision sensor system suitable for monitoring the key parameters of the converter transformer;

[0017] Construct a data acquisition and transmission system to capture the dynamic changes of the operating state of the converter transformer in real time;

[0018] After the data is transmitted to the central control system, monitor and preprocess the data in real time;

[0019] After the preprocessing is completed, construct a time series model based on the cleaned data. The time series model is used to describe the law of change of the key operating parameters of the converter transformer over time and capture the dynamic characteristics therein.

[0020] Preferably, the specific steps for determining the optimal time point for switching the tap of the converter transformer are as follows:

[0021] Use the constructed time series model to perform multi-step forward prediction on the operating state in the future for a period of time. Let X t represent the key operating parameters, and the predicted values for the next k steps in the future are expressed as: In the formula, is the predicted value at time point t + k, g(·) is the prediction function based on the time series model, θ is the model parameter set, n is the length of the historical data window, and k is the index of the time step;

[0022] Based on the predicted operating state values, calculate the predicted state change rate at each time point to capture the moments of drastic changes in the future operating state. Define the predicted state change rate V t+k as:

[0023]

[0024] , where p is the non-linear expansion exponent and λ is the time decay factor;

[0025] Calculate the predicted state change acceleration A t+k , and the predicted state change acceleration can capture more subtle dynamic change characteristics. The calculation expression is:

[0026]

[0027] , where Δt is the time step, and A t+k represents the predicted state change acceleration at time point t + k;

[0028] Combine the predicted state change rate and the predicted state change acceleration to construct a discriminant function Q t+k for the key time points. The expression of the discriminant function is:

[0029]

[0030] , where represents the maximum value of the product of the predicted state change rate and the predicted state change acceleration at all future time points t + j. j represents the time point index and is used to traverse each future time point. β is the weight coefficient of the predicted state change acceleration. The function Q t+k standardizes the state change characteristics at each time point;

[0031] Determine the optimal switching time point through the maximum value of the discriminant function Q t+k . The calculation expression of the optimal switching time point is: where t * represents the optimal switching time point.

[0032] Preferably, the power loss difference at each determined switching time point is defined as the switching cost, and the specific steps are as follows:

[0033] Before the determined switching time point, obtain the power loss P at the current tap position pre , and the power loss is calculated from the electrical parameters monitored in real time during system operation. The calculation expression is: In the formula, I t is the current at the current tap position, R pre is the resistance at the current tap position, V t is the voltage at the current tap position, cos(φ pre ) is the power factor at the current tap position;

[0034] Predict the power loss R after switching to the new tap position post , and after switching, the calculation expression for the power loss after switching to the new tap position is: In the formula, R post and cos(φ post ) are the resistance and power factor corresponding to the new tap position after switching, I t+k and V t+k are the current and voltage corresponding to the new tap position after switching;

[0035] After obtaining the power loss data before and after switching, calculate the difference between the two to quantify the impact of the switching operation. The power loss difference ΔP is expressed as: ΔP = P post - P pre ;

[0036] Finally, define the calculated power loss difference ΔP as the switching cost.

[0037] Preferably, construct a cost function to evaluate the power loss at each switching time point. The specific steps are as follows:

[0038] Define the power loss function P for each switching time point t + k t+k , and the power loss function P t+k represents the power loss of the converter transformer in the future time series. Considering the changes in current, voltage, resistance, and power factor, the expression of the power loss function P t+k is: In the formula, I t is the current at time t, R t is the resistance at time t, V t is the voltage at time t, cos(φ t ) is the power factor at time t;

[0039] Calculate the power loss difference ΔP before and after a specific switching time point t + k t+k , and this difference directly reflects the impact of the switching operation on the system energy efficiency. The calculation expression is: ΔP t+k = P post，t+l - P pre，t+k , where P pre，t+k is the power loss before switching, and P pre，t+k is the power loss after switching.

[0040] Preferably, introduce the power loss difference ΔP t+k into the cost function C t+k to evaluate the power loss at each switching time point. The expression of the cost function is:

[0041]

[0042] , where T t+k is the time weight factor used to adjust the importance of the switching time point, and α is the exponential parameter that controls the nonlinear degree of the function;

[0043] Use the defined cost function C t+k to perform cost evaluation on all switching time points t + k. Specifically, traverse each existing switching time point within the future time period and calculate the corresponding cost values;

[0044] Establish a data set for the cost values generated by all switching time points, and sort the cost values within the data set in order. Select the time point with the smallest cost value from all switching time points as the optimal switching timing.

[0045] Preferably, use the dynamic programming optimization method to solve the cost function and obtain the optimal tap - changing strategy. The specific steps are as follows:

[0046] First, define the state - transition function S(t + k, u), where u represents the tap - changing operation at the time point t + k. S(t + k, u) describes the transfer of the current state to the next time point by selecting u. The state - transition function is expressed as: S(t + k, u) = P t+k + f(S(t + k - 1, u - 1)), where P t+k is the power loss at the time point t + k, and the function f(S(t + k - 1, u - 1)) represents the influence of the state at the previous time point on the current state;

[0047] In dynamic programming, the stage cost function C(t + k, u) is used to represent the power loss cost when operation u is selected at time point t + k. The construction expression of the stage cost function C(t + k, u) is: C(t + k, u) = r(S(t + k, u), u), where r(S, u) is the functional relationship between power loss and operation, representing the total power loss after selecting u.

[0048] Preferably, using the idea of dynamic programming, recursively solve the optimal sub-problems at each time point. Define the optimal value function M(t + k) as the minimum cumulative power loss after time point t + k, which is obtained through recursive calculation. The calculation expression is: M(t + k) = min u∈U {C(t + k, u) + M(t + k + 1)}, where U represents the set of all existing operations;

[0049] In dynamic programming, boundary conditions are used to determine the starting or ending point of recursive calculation. Set the optimal value function M(t + k) at the last time point t + N as the known final value. Here, the optimal value function M(t + k) is defined as: M(t + k) = C(t + k, u * )), where u * is the operation selected at the last time point, and N represents the last time step within the prediction time range;

[0050] Through the backward tracing method, starting from the boundary conditions, gradually trace back to construct the optimal switching strategy at each time point. The optimal strategy Φ(t + k) is defined as the optimal operation selected at each time point. The specific expression is: Φ(t + k) = argmin u∈U {C(t + k, u) + M(t + k + 1)}. Through backward tracing calculation, obtain the optimal strategy Φ(t + k) for the entire time series, and apply this strategy in actual operation for tap changer switching to ensure that the power loss at each moment of the system is minimized.

[0051] Preferably, after the optimal tap changer switching strategy is obtained through dynamic programming optimization, the specific process of real-time controlling the switching action of the on-load tap changer is as follows:

[0052] First, generate specific operation instructions according to the switching time points and corresponding operations determined in the optimal strategy;

[0053] Subsequently, the control system executes these instructions through the actuator to ensure that the tap is switched at the specified time point.

[0054] A on-load tap-changer control system for a converter transformer, comprising a data acquisition and model construction module, a state prediction module, an optimal switching timing determination module, a switching cost calculation module, a cost function construction and evaluation module, a dynamic programming optimization module, and a real-time control execution module;

[0055] Through high-precision sensors and a data acquisition system, the key operating parameters of the converter transformer are monitored in real time at each moment. Based on the collected operating parameter data, a time series model of the operating state of the converter transformer is constructed, and the model describes the law of change of the operating parameters over time to capture the dynamic characteristics of the operating state;

[0056] Using the constructed time series model, the change trend of the operating state of the converter transformer in a future period of time is predicted;

[0057] According to the predicted change trend of the operating state, the optimal time point for switching the tap of the converter transformer is determined;

[0058] For each determined switching time point, the power loss difference before and after the tap change is obtained and defined as the switching cost;

[0059] A cost function is constructed to evaluate the power loss at each switching time point. By considering the minimization of power loss, the power loss difference before and after the switch is used as the optimization target, the power loss is calculated for all switching time points, and the switching scheme with the minimum power loss is selected;

[0060] Using the dynamic programming optimization method, the cost function is solved to obtain the optimal tap-changing strategy. By decomposing the power loss problem into multiple sub-problems and gradually optimizing the switching strategy at each time point, it is ensured that the power loss of the system is minimized at each moment;

[0061] According to the optimal tap-changing strategy obtained by dynamic programming optimization, the switching action of the on-load tap-changer is controlled in real time, and the optimal strategy is converted into actual operation instructions.

[0062] In the above technical solution, the technical effects and advantages provided by the present invention:

[0063] Through the above on-load tap-changer control method for a converter transformer, by constructing a time series model and using the dynamic programming optimization method, the switching timing is predicted in advance and reasonably arranged, unnecessary operations are reduced, mechanical wear is reduced, the service life of the equipment is extended, and the long-term reliability of the system is improved.

[0064] By optimizing the switching strategy, introducing a quantization index of power loss into the cost function, and globally optimizing the switching strategy using dynamic programming, the present invention can screen out the switching scheme with the minimum power loss, enabling each tap change to be carried out with the minimum power loss, improving the economic efficiency and energy efficiency of system operation, and enhancing the stability of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required in the embodiments. Obviously, the drawings described below are only some embodiments recorded in the present invention, and those of ordinary skill in the art can also obtain other drawings based on these drawings.

[0066] Figure 1 It is a method flow chart of a method for controlling an on-load tap-changer of a converter transformer according to the present invention.

[0067] Figure 2 It is a module schematic diagram of a control system for an on-load tap-changer of a converter transformer according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0068] Now, the exemplary embodiments will be described more fully with reference to the accompanying drawings. However, the exemplary embodiments can be implemented in various forms and should not be construed as limited to the examples set forth herein; rather, these exemplary embodiments are provided so that the present disclosure will be more thorough and complete, and will fully convey the concept of the exemplary embodiments to those skilled in the art.

[0069] The present invention provides a method for controlling an on-load tap-changer of a converter transformer as shown in Figure 1 and includes the following steps:

[0070] Through high-precision sensors and a data acquisition system, the key operating parameters of the converter transformer are monitored in real time. Based on the collected operating parameter data, a time series model of the operating state of the converter transformer is constructed to describe the law of change of the operating parameters over time through the model and capture the dynamic characteristics of the operating state.

[0071] The specific steps for constructing a time series model of the operating state of the converter transformer based on the collected operating parameter data of the converter transformer are as follows:

[0072] Select a high-precision sensor system suitable for monitoring the key parameters of the converter transformer.

[0073] Sensors in a high-precision sensor system should be able to measure and record key operating parameters in real time, including voltage, current, temperature, frequency, power factor, etc. The selection of sensors needs to consider measurement accuracy, response speed, and anti-interference ability to ensure accurate and stable data can be obtained under various operating conditions. The installation location of the sensors is also crucial and should be as close as possible to key components to capture subtle changes in the operating state in real time. In addition, to ensure data integrity, the sensor system should be seamlessly integrated with the control and monitoring system of the transformer and have a redundant design to cope with possible equipment failures.

[0074] Build a data acquisition and transmission system to capture the dynamic changes in the operating state of the converter transformer in real time;

[0075] The core of this system is the data acquisition device (such as a data logger or PLC), which needs to have a high sampling rate and a large-capacity data storage ability to capture the parameter data collected by the sensors in real time. The data acquisition system should set a reasonable sampling frequency to ensure that the dynamic changes in the operating state of the converter transformer can be captured while avoiding data redundancy. The collected data is transmitted to the central control system or cloud server in real time through wired or wireless networks for centralized processing and analysis. During the data transmission process, the integrity and security of the data should be ensured, and encryption technology and data verification mechanisms are used to prevent data loss or tampering.

[0076] After the data is transmitted to the central control system, perform real-time monitoring and preprocessing on the data;

[0077] Real-time monitoring includes continuously tracking the sensor data to detect possible anomalies or faults. Preprocessing includes steps such as data cleaning, denoising, and normalization to ensure that the data input into the model is accurate and consistent. For example, it may be necessary to filter out abnormal data points caused by sensor noise or faults and normalize different types of data for unified analysis. At this stage, simple statistical analysis can also be performed on the preprocessed data, such as calculating the average value, standard deviation, etc., to preliminarily understand the operating state and trend of the converter transformer.

[0078] After the preprocessing is completed, build a time series model based on the cleaned data. The time series model is used to describe the variation law of the key operating parameters of the converter transformer over time and capture the dynamic characteristics therein;

[0079] Common time series modeling methods include the ARIMA model, long short-term memory network (LSTM), autoregressive conditional heteroskedasticity model (GARCH), etc. The model selection should be based on the characteristics of the data and specific analysis requirements. When constructing the model, historical data needs to be trained and the model parameters tuned to improve the prediction accuracy and generalization ability of the model. In this way, the time series model can effectively identify the periodic trends, sudden anomalies, and long-term evolutions in the operating state of the converter transformer.

[0080] Using the constructed time series model, predict the changing trend of the operating state of the converter transformer over a period of time in the future. Based on the predicted changing trend of the operating state, determine the optimal time point for switching the tap of the converter transformer;

[0081] The specific steps for determining the optimal time point for switching the tap of the converter transformer are as follows:

[0082] Use the constructed time series model (such as LSTM, GRU, or extended Kalman filter EKF) to perform multi-step forward prediction on the operating state over a period of time in the future. Let X t represent the key operating parameters (such as voltage, load, etc.), and the predicted values for the next k steps are expressed as: In the formula, is the predicted value at time point t + k, g(·) is the prediction function based on the time series model, θ is the set of model parameters, n is the length of the historical data window, and k is the index of the time step;

[0083] The goal of this step is to generate the predicted parameter values at multiple future time points for subsequent analysis.

[0084] Based on the predicted operating state values, calculate the predicted state change rate at each time point to capture the moments of significant changes in the future operating state. Define the predicted state change rate V t+k as:

[0085]

[0086] , where p is the non-linear expansion exponent and λ is the time decay factor;

[0087] The non-linear expansion exponent p is an exponential factor used to enhance or weaken the change amplitude of the predicted parameters at different time points. It is used to introduce non-linear effects when calculating the predicted state change rate, so that the impact of parameter changes can be amplified or reduced according to specific requirements. By adjusting the value of p, the sensitivity of state changes can be flexibly controlled.

[0088] When p > 1, larger changes will be further amplified, which is suitable for situations where strong fluctuations are emphasized.

[0089] When 0 < p < 1, small changes are amplified and large changes are compressed. This setting is suitable for smoothing out drastic changes and enhancing the response to minor changes.

[0090] When p = 1, the formula remains linear and the change amplitude is proportional to the change in the original data.

[0091] The specific value can be set according to the operating characteristics of the system and actual requirements. For example, when being particularly sensitive to drastic fluctuations, p > 1 can be selected; while in the case of smoothing out drastic fluctuations, p < 1 can be chosen. By adjusting p, different types of operating state changes can be flexibly addressed.

[0092] The predicted state change rate is used to measure the state change rate at each future time point. A high rate means that the state of the system will experience drastic fluctuations in the future, and tap changer switching may be required.

[0093] To further quantify the severity of state changes, the predicted state change acceleration A is calculated t+k , and the predicted state change acceleration can capture more subtle dynamic change characteristics. The calculation expression is:

[0094]

[0095] , where Δt is the time step, and A t+k represents the predicted state change acceleration at time point t + k;

[0096] By calculating the predicted state change acceleration, the time points where drastic changes may occur can be identified, and these points need to be considered with emphasis.

[0097] Combining the predicted state change rate and the predicted state change acceleration, a discriminant function Q for key time points is constructed t+k , and the expression of the discriminant function is:

[0098]

[0099] , where represents the maximum value of the product of the predicted state change rate and the predicted state change acceleration at all future time points t + j. j represents the time point index, which is used to traverse each future time point. β is the weight coefficient of the predicted state change acceleration. The function Q t+k normalizes the state change characteristics of each time point. Through this function, the severity of changes at future time points can be ranked, and the time point corresponding to the maximum value here is the optimal switching time.

[0100] Through the discriminant function Q t+kDetermine the optimal switching time point based on the maximum value, and the calculation expression for the optimal switching time point is: In the formula, t * represents the optimal switching time point;

[0101] In this step, directly select the time point with the largest discriminant function value as the optimal time to switch the tap, ensuring that the tap is switched at the moment when the state changes most drastically and appropriately, thereby optimizing the stability and operating efficiency of the system.

[0102] For each determined switching time point, obtain the power loss difference before and after the tap is switched, and define it as the switching cost;

[0103] Define the power loss difference at each determined switching time point as the switching cost, and the specific steps are as follows:

[0104] Before the determined switching time point, obtain the power loss P pre under the current tap position. The power loss is calculated by real-time monitoring of the electrical parameters (such as voltage, current, load) of the system operation, and the calculation expression is: In the formula, I t is the current at the current tap position, R pre is the resistance at the current tap position, V t is the voltage at the current tap position, cos(φ pre ) is the power factor at the current tap position;

[0105] The core of this step lies in accurately capturing and calculating the power loss at the current tap position, laying a foundation for subsequent switching cost calculation.

[0106] Predict the power loss P post after switching to the new tap position. After switching, due to the change in the tap position, electrical parameters such as resistance and power factor will also change accordingly. At this time, the calculation expression for the power loss after switching to the new tap position is: In the formula, R post and cos(φ post ) are the resistance and power factor corresponding to the new tap position after switching, I t+k and V t+k are the current and voltage corresponding to the new tap position after switching;

[0107] To accurately predict the changes in these parameters, an existing time series model or calculation based on the physical characteristics of the system can be used. The goal of this step is to obtain the power loss data of the new tap position by analyzing the operating state after switching.

[0108] After obtaining the power loss data before and after the switch, calculate the difference between the two to quantify the impact of the switching operation. The power loss difference ΔP is expressed as: ΔP = P post - P pre ;

[0109] The power loss difference ΔP directly reflects the impact of the switching operation on the system energy efficiency. If ΔP is positive, it means that the power loss increases after the switch; otherwise, it indicates that the power loss decreases. This calculation can not only quantify the immediate impact of the switch but also provide a basis for defining the switching cost in the next step.

[0110] Finally, define the calculated power loss difference ΔP as the switching cost C switch , and the defined expression is: C switch = |ΔP|;

[0111] The switching cost is expressed as the absolute change in power loss and is a measure of the impact of the switching operation on energy consumption.

[0112] Construct a cost function to evaluate the power loss at each switching time point. By considering the minimization of power loss, the power loss difference before and after the switch is used as the optimization objective, calculate the power loss for all switching time points, and select the switching scheme with the minimum power loss;

[0113] Define the power loss function P for each switching time point t + k t+k , and the power loss function P t+k represents the power loss of the converter transformer in the future time series. Considering the changes in current, voltage, resistance, and power factor, the expression of the power loss function P t+k is: In the formula, I t is the current at time t, R t is the resistance at time t, V t is the voltage at time t, cos(φ t ) is the power factor at time t;

[0114] This integral formula represents the cumulative power loss in the time range from t to t + k. By calculating P t+k at different time points, the power loss situation at each time point can be understood.

[0115] Calculate the power loss difference ΔP before and after the specific switching time point t + k t+k , and this difference directly reflects the impact of the switching operation on the system energy efficiency. The calculation expression is: ΔP t+k = P post，t+l - P pre，t+k , in the formula, Ppre，t+k is the power loss before switching, P pre，t+k is the power loss after switching;

[0116] By calculating the power loss difference ΔP t+k , the change in power loss caused by the switching operation at each time point can be quantified.

[0117] Introduce the power loss difference ΔP t+k into the cost function C t+k to evaluate the power loss at each switching time point. The expression of the cost function is:

[0118]

[0119] , where T t+k is the time weight factor, which is used to adjust the importance of the switching time point. Its function is to highlight or weaken the relative importance of certain time points by assigning different weights to different time points when calculating the cost function. By adjusting T t+k , the selection of the switching time point can be better optimized to ensure switching at the most appropriate moment, thereby improving the overall operation efficiency and stability of the system. The specific setting standard of the time weight factor is usually based on the operating characteristics and strategic objectives of the system. For example, the time weight factor T t+k can be set according to factors such as historical operation data, load change rules, equipment aging conditions, or power grid fluctuation characteristics. If the power demand or load fluctuates frequently at certain time points, higher weights can be assigned to these time points, so as to give priority to tap switching at these moments. On the contrary, for periods with stable load or stable system operation status, lower weights can be set to reduce unnecessary switching operations. Through such settings, the switching strategy can be optimized to improve the efficiency and reliability of the system. α is an exponential parameter that controls the nonlinear degree of the function. This cost function aims to minimize the power loss difference and at the same time consider the priority of the switching time point.

[0120] Using the defined cost function C t+k , conduct cost evaluation for all switching time points t + k. Specifically, it is necessary to traverse each existing switching time point within the future time period and calculate the corresponding cost value;

[0121] The purpose of this step is to systematically evaluate the power loss at all time points so as to screen out the time point with the minimum cost in the subsequent steps.

[0122] Establish a data set for the cost values generated by all switching time points, and sort the cost values in the data set in order. Select the time point with the minimum cost value from all switching time points as the optimal switching timing;

[0123] This selection criterion ensures that the system performs tap changing with the lowest power loss and the best energy efficiency. By minimizing the cost function C t+k , the system can perform the switching at the appropriate time point, thereby optimizing the power loss and improving the operating efficiency.

[0124] Using the dynamic programming optimization method, the cost function is solved to obtain the optimal tap-changing strategy. By decomposing the power loss problem into multiple sub-problems, the switching strategy at each time point is gradually optimized to ensure that the power loss of the system is minimized at each moment;

[0125] Using the dynamic programming optimization method, the cost function is solved to obtain the optimal tap-changing strategy. The specific steps are as follows:

[0126] First, define the state transition function S(t + k, u), where u represents the tap-changing operation at time point t + k. S(t + k, u) describes the transfer of the current state to the next time point by selecting u. The state transition function is expressed as: S(t + k, u) = P t+k + f(S(t + k - 1, u - 1)), where P t+k is the power loss at time point t + k, and the function f(S(t + k - 1, u - 1)) represents the influence of the state at the previous time point on the current state;

[0127] The function f(S(t + k - 1, u - 1)) reflects how the power loss at time point t + k is affected by the state transition or operation decision at the previous time point t + k - 1. The determination of the function f(S(t + k - 1, u - 1)) is usually based on the physical characteristics and dynamic behavior of the system. For example, in the case of a converter transformer, the function f(S(t + k - 1, u - 1)) can be modeled according to the change rates of current and voltage, the adjustment of power factor, and the system stability after tap changing. Usually, the function can be determined by fitting experimental data, analyzing historical operation data, or based on the physical model of the power system, so as to accurately describe how the operation at the previous time point affects the power loss and system state at the current time point. This transfer function is used to describe the change of the system power loss after selecting different tap-changing schemes at each time step.

[0128] In dynamic programming, the stage cost function C(t + k, u) is constructed to represent the power loss cost when selecting operation u at time point t + k. The construction expression of the stage cost function C(t + k, u) is: C(t + k, u) = r(S(t + k, u), u), where r(S, u) is the functional relationship between power loss and operation, representing the total power loss after selecting u;

[0129] The stage cost function is used to evaluate the immediate cost when different operations are selected at different time points. Combining with the state transition function, the optimal strategy can be recursively solved.

[0130] Using the idea of dynamic programming, the optimal value problem at each time point is recursively solved. The optimal value function M(t + k) is defined as the minimum cumulative power loss after time point t + k, which is obtained through recursive calculation. The calculation expression is: M(t + k) = min u∈U {C(t + k, u) + M(t + k + 1)}, where U represents the set of all existing operations;

[0131] This formula represents the minimum cumulative cost after selecting the optimal operation u at time point t + k. Through recursive calculation, the optimal power loss from the current time point to future time points can be obtained.

[0132] In dynamic programming, setting boundary conditions is very important. Boundary conditions are used to determine the starting or ending point of recursive calculation. Usually, the optimal value function M(t + k) at the last time point t + N is set as the known final value. Here, the optimal value function M(t + k) is defined as: M(t + k) = C(t + k, u * ) where u * is the operation selected at the last time point. The boundary conditions ensure that the recursive process can terminate correctly and provide a basis for backward solution. N represents the last time step within the prediction time range;

[0133] Finally, through the backward tracing method, starting from the boundary conditions, trace back step by step to construct the optimal switching strategy at each time point. The optimal strategy Φ(t + k) is defined as the optimal operation selected at each time point. The specific expression is: Φ(t + k) = argmin u∈U {C(t + k, u) + M(t + k + 1)}. Through backward tracing calculation, the optimal strategy Φ(t + k) for the entire time series is obtained, and this strategy is applied in actual operation for tap changer switching to ensure that the power loss at each moment of the system is minimized.

[0134] According to the optimal tap changer switching strategy optimized by dynamic programming, the switching action of the on-load tap changer is controlled in real time, and the optimal strategy is converted into actual operation instructions;

[0135] After the optimal tap changer switching strategy is optimized by dynamic programming, the specific process of controlling the switching action of the on-load tap changer in real time is as follows:

[0136] First, according to the switching time points and corresponding operations determined in the optimal strategy, specific operation instructions are generated;

[0137] These instructions include switching the tap to a specified target position to minimize power loss or optimize system performance.

[0138] Subsequently, the control system executes these instructions through an actuator (such as a motor or a hydraulic drive system) to ensure that the tap is switched at the specified time point.

[0139] The entire process requires real-time monitoring of the system's operating status to ensure the accuracy and timeliness of the switching operation, and to adjust the instructions when necessary to cope with the dynamic changes in the operating environment.

[0140] Adjusting the instructions when necessary means that during the implementation of the optimal tap-changing strategy, the operating status of the converter transformer and the power system is monitored in real time. If unexpected changes occur in the system operating environment, such as abnormal load fluctuations, sharp voltage changes, or system parameters deviating from the expectations of the prediction model, these may all affect the effect of the switching operation. In this case, the control system needs to re-evaluate the current switching strategy based on the latest real-time data and adjust the original operation instructions. For example, delaying or advancing the switching time, or changing the target position of the tap, to ensure that the system can still maintain optimal performance and safety under the new operating conditions. This dynamic adjustment mechanism ensures that the system can flexibly respond to complex operating environments and avoid increased power loss or system instability caused by external changes.

[0141] Through the above-mentioned on-load tap-changer control method for converter transformers, the present invention builds a time series model and uses the dynamic programming optimization method to predict and reasonably arrange the switching timing in advance, reduce unnecessary operations, reduce mechanical wear, extend the service life of the equipment, and improve the long-term reliability of the system.

[0142] By optimizing the switching strategy, introducing a quantitative index of power loss into the cost function, and globally optimizing the switching strategy using dynamic programming, the present invention can screen out the switching scheme with the minimum power loss, enabling each tap change to be carried out with the minimum power loss, improving the economic efficiency and energy efficiency of the system operation, and enhancing the stability of the power system.

[0143] The present invention provides a Figure 2 control system for an on-load tap-changer of a converter transformer as shown, including a data acquisition and model construction module, a state prediction module, an optimal switching timing determination module, a switching cost calculation module, a cost function construction and evaluation module, a dynamic programming optimization module, and a real-time control execution module.

[0144] Through high-precision sensors and data acquisition systems, the key operating parameters of the converter transformer are monitored in real time. Based on the collected operating parameter data, a time series model of the operating state of the converter transformer is constructed to describe the law of change of the operating parameters over time and capture the dynamic characteristics of the operating state.

[0145] Using the constructed time series model, the change trend of the operating state of the converter transformer in a future period of time is predicted.

[0146] According to the predicted change trend of the operating state, the optimal time point for switching the tap of the converter transformer is determined.

[0147] For each determined switching time point, the power loss difference before and after the tap change is obtained and defined as the switching cost.

[0148] A cost function is constructed to evaluate the power loss at each switching time point. By considering the minimization of power loss, the power loss difference before and after the switch is used as the optimization objective, and the power loss is calculated for all switching time points to screen out the switching scheme with the minimum power loss.

[0149] Using the dynamic programming optimization method to solve the cost function, an optimal tap-changing strategy is obtained. By decomposing the power loss problem into multiple sub-problems and gradually optimizing the switching strategy at each time point, it is ensured that the power loss of the system is minimized at each moment.

[0150] According to the optimal tap-changing strategy obtained by dynamic programming optimization, the switching action of the on-load tap-changer is controlled in real time, and the optimal strategy is converted into actual operation instructions.

[0151] A method for controlling the on-load tap-changer of a converter transformer provided by an embodiment of the present invention is implemented through the above-mentioned on-load tap-changer control system of a converter transformer. The specific methods and processes of the on-load tap-changer control system of a converter transformer are detailed in the embodiments of the above-mentioned method for controlling the on-load tap-changer of a converter transformer, and will not be elaborated here.

[0152] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (such as infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that contains one or more collections of available media. The available media can be magnetic media (such as floppy disks, hard disks, magnetic tapes), optical media (such as DVDs), or semiconductor media. The semiconductor media can be a solid-state drive.

[0153] Those of ordinary skill in the art will appreciate that the units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or in a combination of computer software and electronic hardware. Whether these functions are executed in hardware or software depends on the specific application and design constraints of the technical solution. Skilled artisans can use different methods for each specific application to implement the described functions, but such implementation should not be considered to exceed the scope of the present application.

[0154] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0155] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only one way, and there can be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces. The indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms.

[0156] The unit described as a separation component may or may not be physically separated. The component displayed as a unit may or may not be a physical unit, that is, it may be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0157] In addition, each functional unit in various embodiments of the present application may be integrated into a processing unit, may exist separately as individual physical units, or two or more units may be integrated into one unit.

[0158] If the above functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present application. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0159] The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present application, and all of them should be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claimed rights.

[0160] Finally: The above is only the preferred embodiment of the present invention and is not used to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A control method for an on-load tap-changer of a converter transformer, characterized in that, It includes the following steps: Through high-precision sensors and data acquisition systems, the key operating parameters of the converter transformer are monitored in real time at each moment. Based on the collected operating parameter data, a time series model of the operating state of the converter transformer is constructed. The model describes the law of change of operating parameters over time and captures the dynamic characteristics of the operating state; Using the constructed time series model, predict the change trend of the operating state of the converter transformer in a future period of time; According to the predicted change trend of the operating state, determine the optimal time point for the converter transformer to switch taps; For each determined switching time point, obtain the power loss difference before and after the tap change and define it as the switching cost; Construct a cost function to evaluate the power loss at each switching time point. The cost function, by considering the minimization of power loss, takes the power loss difference before and after the switch as the optimization goal, calculates the power loss for all switching time points, and selects the switching scheme with the minimum power loss; Using the dynamic programming optimization method, solve the cost function to obtain the optimal tap-changing strategy. By decomposing the power loss problem into multiple sub-problems, gradually optimize the switching strategy at each time point to ensure that the power loss of the system is minimized at each moment; According to the optimal tap-changing strategy obtained by dynamic programming optimization, control the switching action of the on-load tap-changer in real time and convert the optimal strategy into actual operation instructions.

2. The on-load tap-changer control method of a commutation transformer according to claim 1, characterized in that, Based on the collected operating parameter data of the converter transformer, the specific steps for constructing a time series model of the operating state of the converter transformer are as follows: Select a high-precision sensor system suitable for monitoring the key parameters of the converter transformer; Construct a data acquisition and transmission system to capture the dynamic changes of the operating state of the converter transformer in real time; After the data is transmitted to the central control system, monitor and preprocess the data in real time; After the preprocessing is completed, construct a time series model based on the cleaned data. The time series model is used to describe the law of change of the key operating parameters of the converter transformer over time and capture the dynamic characteristics therein.

3. A method for controlling an on-load tap-changer of a converter transformer according to claim 1, characterized in that The specific steps for determining the optimal time point for the converter transformer to switch taps are as follows: Use the constructed time series model to perform multi-step forward prediction on the operating state in a future period. Let X t represent the key operating parameters, and the predicted values for the next k steps are expressed as: In the formula, is the predicted value at time point t + k, g(·) is the prediction function based on the time series model, θ is the model parameter set, n is the length of the historical data window, and k is the index of the time step; Based on the predicted operating state value, calculate the predicted state change rate at each time point to capture the moment of drastic change in the future operating state, and define the predicted state change rate V t+k as follows: Where p is the non-linear expansion exponent and λ is the time decay factor; Calculate the predicted state change acceleration A t+k , the predicted state change acceleration can capture more subtle dynamic change characteristics, and the calculation expression is: where Δt is the time step, and A t+k represents the predicted state change acceleration at time point t + k; Construct the discriminant function Q at the key time point by combining the predicted state change rate and the predicted state change acceleration t+k , and the expression of the discriminant function is as follows: In the formula, represents the maximum value of the product of the predicted state change rate and the predicted state change acceleration at all future time points t + j, where j represents the time point index used to traverse each future time point, β is the weight coefficient of the predicted state change acceleration, and the function Q t+k normalizes the state change characteristics of each time point; Determine the optimal switching time point by the maximum value of the discrimination function Q t+k , and the calculation expression of the optimal switching time point is: In the formula, t * represents the optimal switching time point.

4. A method for controlling an on-load tap-changer of a converter transformer according to claim 3, characterized in that, Define the power loss difference at each determined switching time point as the switching cost. The specific steps are as follows: Before the determined switching time point, obtain the power loss P at the current tap position pre , and the power loss is calculated from the electrical parameters monitored in real time during the operation of the system. The calculation formula is as follows: In the formula, I t is the current at the current tap position, R pre is the resistance at the current tap position, V t is the voltage at the current tap position, and cos(φ pre ) is the power factor at the current tap position; Prediction of the power loss R after switching to a new tap position post , after switching, the calculation expression of the power loss after switching to the new tap position is as follows: In the formula, R post and cos(φ post ) are the resistance and power factor corresponding to the new tap position after switching, I t+k and V t+k are the current and voltage corresponding to the new tap position after switching; After obtaining the power loss data before and after the switch, calculate the difference between the two to quantify the impact of the switching operation. The power loss difference ΔP is expressed as: ΔP = P post - P pre ; Finally, define the calculated power loss difference ΔP as the switching cost.

5. A method for controlling an on-load tap-changer of a converter transformer according to claim 4, characterized in that, Construct a cost function to evaluate the power loss at each switching time point. The specific steps are: Define the power loss function \(P\) for each switching time point \(t + k\). t+k The power loss function \(P\) t+k represents the power loss of the converter transformer in the future time series. Considering the variations of current, voltage, resistance, and power factor, the expression of the power loss function \(P\) t+k is as follows: where \(I\) t is the current at time \(t\), \(R\) t is the resistance at time \(t\), \(V\) t is the voltage at time \(t\), and \(\cos(\varphi\) t ) is the power factor at time \(t\). Calculate the power loss difference ΔP before and after the switching time point t + k t+k , which directly reflects the impact of the switching operation on the system energy efficiency. The calculation formula is: ΔP t+k = P post，t+k - P pre，t+k , where P pre，t+k is the power loss before switching, and P pre，t+k is the power loss after switching.

6. The on-load tap-changer control method for a commutation transformer according to claim 5, characterized in that Introduce the power loss difference ΔP t+k into the cost function C t+k , and evaluate the power loss at each switching time point. The expression of the cost function is as follows: where, T t+k is the time weight factor for adjusting the importance of the switching time point, and α is the exponential parameter for controlling the nonlinear degree of the function; Using the defined cost function C t+k , the cost is evaluated for all switching time points t + k, specifically by traversing each existing switching time point within the future time period and calculating the corresponding cost value; Establish a data set of the cost values generated at all switching time points, sort the cost values in the data set in order, and select the time point with the minimum cost value from all switching time points as the optimal switching time.

7. A method for controlling an on-load tap changer of a converter transformer according to claim 6, characterized in that, Using the dynamic programming optimization method, solve the cost function to obtain the optimal tap-changing strategy. The specific steps are: First, define the state transition function S(t + k, u), where u represents the tap - changing operation at time point t + k, and S(t + k, u) describes the transfer of the current state to the next time point by choosing u. The state transition function is expressed as: S(t + k, u)=P t+k +f(S(t + k - 1, u - 1)), where P t+k is the power loss at time point t + k, and the function f(S(t + k - 1, u - 1)) represents the influence of the state at the previous time point on the current state; In dynamic programming, the construction-phase cost function \(C(t + k, u)\) is used to represent the power loss cost when operation \(u\) is selected at time point \(t + k\). The construction expression of the phase cost function \(C(t + k, u)\) is: \(C(t + k, u)=r(S(t + k, u), u)\), where \(r(S, u)\) is the functional relationship between power loss and operation, representing the total power loss after selecting \(u\).

8. A method for controlling an on-load tap-changer of a converter transformer according to claim 7, characterized in that, Using the idea of dynamic programming, recursively solve the optimal value problem at each time point. Define the optimal value function M(t + k) as the minimum cumulative power loss after time point t + k, which is obtained through recursive calculation. The calculation expression is: M(t + k) = min u∈U {C(t + k, u) + M(t + k + 1)}, where U represents the set of all existing operations; In dynamic programming, boundary conditions are used to determine the starting or ending points of recursive calculations. The optimal value function M(t + k) at the last time point t + N is set as the known terminal value. Here, the optimal value function M(t + k) is defined as: M(t + k) = C(t + k, u * ), where u * is the operation selected at the last time point, and N represents the last time step within the prediction time range; By the reverse backtracking method, starting from the boundary conditions and gradually backtracking, an optimal switching strategy at each time point is constructed. The optimal strategy Φ(t + k) is defined as the optimal operation selected at each time point, and the specific expression is: Φ(t + k) = argmin u∈U {C(t + k, u) + M(t + k + 1)}. The optimal strategy Φ(t + k) over the entire time series is obtained through backtracking calculations, and this strategy is applied in actual operations for tap switching to ensure that the power loss at each moment of the system is minimized.

9. A method for controlling an on-load tap changer of a converter transformer according to claim 1, characterized in that, After the optimal tap-changing strategy is obtained through dynamic programming optimization, the specific process of real-time controlling the switching action of the on-load tap-changer is as follows: First, specific operation instructions are generated according to the switching time points and corresponding operations determined in the optimal strategy; Subsequently, the control system executes these instructions through the actuator to ensure that the tap is switched at the specified time point.

10. A on-load tap-changer control system for a converter transformer is used to implement the on-load tap-changer control method for a converter transformer described in any one of the above claims 1-9, and is characterized in that, It includes a data acquisition and model construction module, a state prediction module, an optimal switching timing determination module, a switching cost calculation module, a cost function construction and evaluation module, a dynamic programming optimization module, and a real-time control execution module; Through high-precision sensors and a data acquisition system, the key operating parameters of the converter transformer are monitored in real time. Based on the collected operating parameter data, a time series model of the operating state of the converter transformer is constructed to describe the law of change of the operating parameters over time and capture the dynamic characteristics of the operating state; Using the constructed time series model, the change trend of the operating state of the converter transformer in the future for a period of time is predicted; According to the predicted change trend of the operating state, the optimal time point for switching the tap of the converter transformer is determined; For each determined switching time point, the power loss difference before and after the tap change is obtained and defined as the switching cost; A cost function is constructed to evaluate the power loss at each switching time point. By considering the minimization of power loss, the power loss difference before and after the switch is used as the optimization target, and the power loss is calculated for all switching time points to screen out the switching scheme with the minimum power loss; Using the dynamic programming optimization method, the cost function is solved to obtain the optimal tap-changing strategy. By decomposing the power loss problem into multiple sub-problems, the switching strategy at each time point is gradually optimized to ensure that the power loss of the system is minimized at each moment; According to the optimal tap-changing strategy obtained through dynamic programming optimization, the switching action of the on-load tap-changer is controlled in real time, and the optimal strategy is converted into actual operation instructions.

Images