Hydropower station load error compensation method and system

Abstract

A hydropower station load error compensation method and system, based on the deviation between expected power and actual power, obtains a power derivative virtual control quantity through an integral type virtual control law; determines the operating condition according to operating data, calls a water head change rate decomposition model corresponding to the operating condition, and obtains an external source driving term and an opening change sensitivity prediction value; inputs the virtual control quantity and the prediction value into a guide vane opening inertia model to obtain a guide vane opening change rate; controls the unit operation based on the change rate and updates the expected power at the next time, until the compensation period ends; if the repair rate meets the standard, the compensation is completed, otherwise, the cumulative error is taken as the initial error of the next period, and the parameters are adjusted to recompensate. Improve the water energy utilization rate and the power grid scheduling response ability, in the case of existing load error, timely correct the historical power generation energy error, and realize stable power output to accurately meet the power grid scheduling requirements, improve the water energy utilization rate, and avoid water abandonment and energy waste caused by load control error.

Description

Technical Field

[0001] This invention belongs to the field of hydropower control technology and relates to a method and system for compensating for load errors in hydropower stations. Background Technology

[0002] As the core hub in the hydropower dispatching system, the main function of a hydropower station is to efficiently convert water energy into electrical energy through hydro-turbine generator units, and to dynamically adjust the output level according to the real-time load demand of the power grid to meet the electricity needs of industrial production and social life.

[0003] In actual operation, the upstream water level of a hydropower station fluctuates year-round. Furthermore, due to the decreased accuracy of water level measurement devices and the time-sensitive changes in unit operating parameters, significant head measurement errors and nonlinear disturbances exist in the system. Most existing control strategies are based on linear regulation models or experience-based scheduling methods relying on the unit's static characteristic curves. Under conditions of rapid head fluctuations or drastic load changes, these strategies struggle to accurately track the target load, easily leading to continuous deviations in unit output from grid commands. Ultimately, this necessitates extending power generation time or implementing water diversion and flood discharge to meet system energy requirements, reducing the efficiency of hydropower resource utilization and causing significant economic losses. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a method and system for compensating for load errors in hydropower stations. When load errors have occurred, the method can promptly correct historical power generation errors and achieve stable power output to accurately meet the grid dispatch requirements, improve the utilization rate of hydropower, and avoid water wastage and energy waste caused by load control errors.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is: a method for compensating for load errors in hydropower stations, comprising the following steps: S1, in the current compensation cycle, obtain the running data for the historical time period; S2, based on historical operating data, determines the initial power generation energy error for the current compensation cycle; S3. Based on the initial power generation energy error, the preset energy error index decay coefficient and the total compensation time of the current compensation cycle, a multi-stage expected power dynamic trajectory model is constructed, and the expected power at the current moment of the current compensation cycle is determined based on the multi-stage expected power dynamic trajectory model. S4, based on the deviation between the expected power and the actual power of the hydro-generator unit at the current moment, the virtual control quantity of the power derivative of the hydro-generator unit at the current moment is obtained through the integral virtual control law; S5. Based on the operating data at the current moment of the current compensation cycle, determine the operating condition at the current moment, and call the head change rate decomposition model under the corresponding operating condition to obtain the predicted value of the external driving term of the head change rate and the predicted value of the opening change sensitivity at the current moment. S6. Establish the guide vane opening inertial model of the hydro-generator unit, and input the predicted values ​​of the external driving terms of the power derivative virtual control quantity, the head change rate, and the opening change sensitivity into the guide vane opening inertial model of the hydro-generator unit to obtain the guide vane opening change rate of the hydro-generator unit at the current moment. S7, based on the guide vane opening change rate, control the operation of the hydro-generator unit, and redetermine the expected power at the next moment of the current compensation cycle to perform error compensation at the next moment, until the last moment of the current compensation cycle is reached, and calculate the compensation repair rate of the current compensation cycle. S8, construct a real-time error compensation trend monitoring map for the current compensation cycle; the real-time error compensation trend monitoring map includes the head change trend curve, the guide vane opening change trend curve, the tracking comparison curve between expected power generation and actual power generation, and the compensation effect comparison curve between real-time cumulative actual power generation and real-time cumulative expected power generation. S9, Determine if the compensation repair rate of the current compensation cycle is within the preset range, then complete the error compensation repair; if the compensation repair rate of the current compensation cycle is not within the preset range, then use the cumulative energy error of the current compensation cycle as the initial power generation energy error of the next compensation cycle, and adjust the preset energy error index attenuation coefficient and total compensation time, and repeat the error compensation repair.

[0006] The operational data includes hydropower station operational data and turbine generator unit operational data; the historical time period includes the start time of the current compensation cycle and the first preset number of times before the start time; the hydropower station operational data includes load demand data and inflow data; the turbine generator unit operational data includes guide vane opening data, head data, and power generation data; S2-1, Calculate the cumulative power generation capacity to be generated in the previous period based on the load demand data of hydropower stations over a historical period; S2-2, calculate the actual generated energy based on the power generation data of the hydro-turbine generator units over a historical period; S2-3, subtract the accumulated power generation energy from the actual power generation energy to obtain the initial power generation energy error.

[0007] In S3, the multi-stage desired power dynamic trajectory model is as follows: ; in, This represents the expected power generation at time t; This represents the load demand power at time t; This represents the preset energy error index attenuation coefficient; This represents the real-time cumulative energy error at time t; This indicates the time point at which the rapid compensation phase transitions to the regression convergence phase; This indicates the time point at which the regression convergence phase transitions to the steady-state locking phase; Indicates the total compensation time.

[0008] In S4, an integral virtual control law is used, and the power derivative virtual control quantity is calculated using the following formula: ; ; in, This represents the virtual control quantity of the power derivative; Indicates proportional gain; Indicates integral gain; This represents the power generation error at time t; This represents the expected power generation at time t; This represents the actual power generation at time t.

[0009] In S5, specifically including, S5-1, based on the operational data at the current moment of the current compensation cycle, the head change rate and the inflow disturbance index are calculated using the following formula: ; ; in, This represents the rate of change of water head at time t; This represents the head data of the hydro-turbine generator unit at time t; This represents the head data of the hydro-turbine generator unit at time t-1; Indicates the time interval between adjacent moments; This represents the index of sudden changes in water inflow at time t; This represents the inflow rate data of the hydropower station at time t; This represents the inflow data of the hydropower station at time t-1; S5-2, If the time interval between the current moment and the last working condition switching moment is less than the preset minimum dwell time, then the operating condition of the previous moment is taken as the operating condition of the current moment; wherein, the initial moment of the current compensation cycle is taken as the initial working condition switching moment, and the operating condition of the initial moment of the current compensation cycle is initialized as the stable period working condition. S5-3, if the time interval between the current moment and the last time the operating condition switch is greater than or equal to the preset minimum dwell time, and the severe disturbance condition determination criteria are met, then the operating condition at the current moment is determined to be a severe disturbance period condition; the severe disturbance condition determination criteria include the continuous first preset time interval before the current moment. ,or ;in, Indicates the threshold for the rate of change of head; Indicates the threshold for changes in inbound flow; S5-4, if the time interval between the current moment and the last time the operating condition switch is greater than or equal to the preset minimum dwell time, and the stable operating condition determination condition is met, then the operating condition at the current moment is determined to be a stable operating condition; the stable operating condition determination condition includes the second preset time period preceding the current moment. and .

[0010] In S5, the loss function of the head change rate decomposition model is: ; in, The loss function value of the head change rate decomposition model is represented by the trainable parameter θ of the neural network as the independent variable; N represents the total number of samples; k represents the k-th sample; This represents the predicted head data for the k-th sample; This represents the actual head data of the k-th sample; Represents the regularization coefficient; This represents the sum of squares of the trainable parameters θ of the neural network.

[0011] In S6, specifically including, S6-1, the theoretical value of the rate of change of the guide vane opening at the current moment can be calculated using the following formula: ; in, This represents the theoretical value of the rate of change of the guide vane opening at time t; This represents the virtual control quantity of the power derivative; This represents the power generation constant of the water turbine; This represents the power generation flow rate at time t; Represents time t Predicted value of external driving term for head change rate under operating conditions. When the value is -1, it indicates a period of severe disturbance. =1 indicates a stable operating condition; Represents time t Predicted value of sensitivity to changes in opening degree under operating conditions; This represents the head data of the hydro-turbine generator unit at time t; Indicates the guide vane opening; This represents the partial derivative of the power generation flow rate with respect to the guide vane opening. S6-2, apply a limiting constraint to the theoretical value of the guide vane opening change rate, so that the theoretical value of the guide vane opening change rate is greater than or equal to the preset minimum opening change rate threshold and less than or equal to the preset maximum opening change rate threshold. S6-3, input the theoretical value of the guide vane opening change rate after the amplitude limiting constraint treatment into the guide vane opening inertial model, and obtain the guide vane opening change rate of the hydro-generator unit at the current moment through the following formula: ; ; in, This represents the rate of change of the guide vane opening at time t; Represents the inertial filter coefficients; Indicates the time interval between adjacent moments; This represents the time constant of the guide vane actuator.

[0012] In S7, specifically including, S7-1, based on the real-time error compensation trend monitoring map, determine whether there is a violent oscillation; the judgment conditions for violent oscillation include: in the tracking comparison curve of expected power generation and actual power generation, the power deviation within the third consecutive preset time period before the current time is greater than the preset power deviation threshold, or in the guide vane opening change trend curve, there is a periodic fluctuation with an amplitude exceeding the preset opening change rate oscillation threshold and a frequency exceeding the preset frequency threshold within the third consecutive preset time period before the current time. S7-2, If there is severe oscillation, adjust the proportional gain parameter and integral gain parameter of the proportional definite integral virtual control law, recalculate the power derivative virtual control quantity at the current moment based on the adjusted proportional gain parameter and integral gain parameter, and redetermine the guide vane opening change rate based on the recalculated power derivative virtual control quantity. S7-3 controls the operation of the hydro-generator unit based on the guide vane opening change rate and updates the real-time error compensation trend monitoring map; S7-4, based on the updated real-time error compensation trend monitoring map, redetermines the expected power at the next moment of the current compensation cycle, and determines whether there will be a violent oscillation to carry out error compensation at the next moment, until the last moment of the current compensation cycle is reached. S7-5 calculates the compensation and repair rate for the current compensation cycle based on the real-time error compensation trend monitoring map.

[0013] The calculation method for the compensation and repair rate specifically includes: If the current compensation period is the first compensation period, the compensation repair rate is calculated using the following formula: ; in, This indicates the compensation and repair rate for the first compensation cycle; This indicates the last moment of the first compensation cycle. Real-time cumulative energy error; This represents the initial power generation energy error in the first compensation period; If the current compensation period is longer than the first compensation period, the compensation repair rate is calculated using the following formula: ; in, Indicates the first The compensation and repair rate during the compensation period; Indicates the first The compensation and repair rate during the compensation period; Indicates the first The last moment of the compensation period Real-time cumulative energy error; Indicates the first The initial power generation energy error during the compensation period.

[0014] An error compensation system for a hydropower station load error compensation method includes a data acquisition module for data exchange, an initial error determination module, a desired power determination module, a virtual control quantity calculation module, an operating condition identification and head decomposition module, a guide vane opening change rate determination module, a closed-loop control and monitoring module, a compensation completion determination module, and a compensation iteration module.

[0015] The main beneficial effects of this invention are as follows: By acquiring historical operating data for the current compensation period and determining the initial power generation energy error, the problem of difficulty in quantifying historical power generation energy error when the system has already experienced load errors is solved, thus achieving an accurate assessment of the previous power generation energy gap.

[0016] By constructing a multi-stage dynamic trajectory model of expected power based on the initial power generation energy error, the preset energy error exponential decay coefficient, and the total compensation time, the problem that traditional linear regulation models are difficult to adapt to rapid head fluctuations and drastic load changes is solved, and a staged, adaptive power tracking strategy is realized.

[0017] By using an integral virtual control law based on the deviation between the desired power and the actual power to obtain the virtual control quantity of the power derivative, the problem of insufficient power control accuracy under nonlinear systems is solved, and a fast response and steady-state elimination of power error are achieved.

[0018] By determining the current operating conditions and calling the head change rate decomposition model under the corresponding operating conditions, the predicted values ​​of the external driving term and the opening change sensitivity are obtained. This solves the problem of the difficulty in accurately modeling the dynamic characteristics of head and realizes the accurate prediction of head changes and the decoupled control of opening adjustment.

[0019] By establishing a guide vane opening inertia model and inputting the virtual control quantity of power derivative, the predicted value of external drive term, and the predicted value of opening change sensitivity, the guide vane opening change rate is obtained. This solves the control deviation problem caused by actuator response lag and realizes precise adjustment of guide vane opening and stable power output.

[0020] By iteratively calculating the compensation and repair rate for error compensation, and determining whether the compensation and repair rate is within the preset range to decide whether to iteratively adjust the parameters and repair again, the problem of incomplete single compensation is solved, achieving complete repair of energy errors and precise fulfillment of grid dispatch requirements, thereby improving water energy utilization and avoiding water wastage and energy waste caused by load control errors. Attached Figure Description

[0021] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0022] Figure 1 This is an application diagram of the present invention.

[0023] Figure 2 This is a flowchart of the present invention.

[0024] Figure 3 This is a flowchart of the logic calculation of the present invention.

[0025] Figure 4 This is a flowchart illustrating the compensation process of the present invention.

[0026] Figure 5 This is a curve showing the trend of water head change in this invention.

[0027] Figure 6 This is a graph showing the trend of guide vane opening in this invention.

[0028] Figure 7 This is a curve showing the tracking comparison between the expected power generation and the actual power generation of this invention.

[0029] Figure 8 This is a curve comparing the compensation effect of the real-time cumulative actual power generation energy and the real-time cumulative expected power generation energy of the present invention.

[0030] Figure 9 This is a schematic diagram of the functional modules of the error compensation system of the present invention.

[0031] Figure 10 This is a schematic diagram of the structure of a computer device according to the present invention.

[0032] Figure 11 for Figure 3 Enlarged view of point A.

[0033] Figure 12 for Figure 3 Enlarged view of point B.

[0034] Figure 13 for Figure 3 Enlarged view of point C. Detailed Implementation

[0035] like Figures 1-13 A method for compensating for load errors in hydropower stations. Example 1, Applied to, for example Figure 1 In the application environment shown; Terminal 101 communicates with server 102 via a network.

[0036] The data storage system can store the data that server 102 needs to process.

[0037] The data storage system can be set up separately, integrated into server 102, or placed in the cloud or on other servers.

[0038] Terminal 101 can send the historical time period operation data obtained in the current compensation cycle to server 102. After receiving the historical time period operation data, server 102 determines the initial power generation energy error of the current compensation cycle based on the historical time period operation data.

[0039] Based on the initial power generation energy error, the preset energy error index decay coefficient, and the total compensation time of the current compensation cycle, a multi-stage expected power dynamic trajectory model is constructed, and the expected power at the current moment of the current compensation cycle is determined based on the multi-stage expected power dynamic trajectory model.

[0040] Based on the deviation between the expected power and the actual power of the hydro-generator unit at the current moment, the virtual control quantity of the power derivative of the hydro-generator unit at the current moment is obtained through an integral virtual control law.

[0041] Based on the operating data at the current moment of the current compensation cycle, the operating condition at the current moment is determined, and the head change rate decomposition model under the corresponding operating condition is called to obtain the predicted value of the external driving term of the head change rate and the predicted value of the opening change sensitivity at the current moment.

[0042] Operating conditions include stable operating conditions and severe disturbance operating conditions.

[0043] The head prediction model is obtained by iteratively training a preset neural network model using a sample dataset. The sample dataset is constructed based on the operating data of the second preset number of time periods before the current compensation cycle under stable or severe disturbance conditions, and the corresponding real head data of the hydro-generator unit.

[0044] An inertial model of the guide vane opening of a hydro-generator unit is established. The predicted values ​​of the power derivative virtual control quantity, the external driving term of the head change rate, and the opening change sensitivity are input into the inertial model of the guide vane opening of the hydro-generator unit to obtain the guide vane opening change rate of the hydro-generator unit at the current moment.

[0045] The operation of the hydro-generator unit is controlled based on the guide vane opening change rate, and the expected power at the next moment of the current compensation cycle is re-determined for error compensation at the next moment, until the last moment of the current compensation cycle is reached, and the compensation repair rate of the current compensation cycle is calculated.

[0046] If the compensation and repair rate of the current compensation cycle is within the preset range, then the error compensation and repair will be completed.

[0047] If the compensation and repair rate of the current compensation cycle is not within the preset range, the cumulative energy error of the current compensation cycle will be used as the initial power generation energy error of the next compensation cycle, and the preset energy error index attenuation coefficient and total compensation time will be adjusted to perform error compensation and repair again.

[0048] Server 102 can feed back the obtained real-time error compensation trend monitoring map and the compensation repair rate of the current compensation period to terminal 101.

[0049] In addition, the hydropower station load error compensation method can also be implemented by the server 102 or the terminal 101. For example, the terminal 101 can directly perform error compensation processing on the historical time period operation data obtained in the current compensation cycle, or the server 102 can obtain the historical time period operation data obtained in the current compensation cycle from the data storage system and perform error compensation processing on the historical time period operation data obtained in the current compensation cycle.

[0050] The terminal 101 can be, but is not limited to, various desktop computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. The server 102 can be implemented using a standalone server or a server cluster consisting of multiple servers, or it can be a cloud server.

[0051] Example 2, like Figure 2As shown, a method for compensating load errors in a hydropower station is provided. This method is executed by computer equipment, specifically by a terminal or server alone, or by both a terminal and a server. In this embodiment, the method is applied to... Figure 1 Taking server 102 as an example, the explanation includes the following steps 201 to 209. Wherein: Step 201: In the current compensation cycle, obtain the operation data of the historical time period; the operation data includes the operation data of the hydropower station and the operation data of the turbine generator unit; the historical time period includes the start time of the current compensation cycle and a first preset number of times before the start time.

[0052] Step 202: Based on the operating data of historical time periods, determine the initial power generation energy error of the current compensation cycle.

[0053] Step 203: Based on the initial power generation energy error, the preset energy error index decay coefficient, and the total compensation time of the current compensation cycle, construct a multi-stage expected power dynamic trajectory model, and determine the expected power at the current moment of the current compensation cycle based on the multi-stage expected power dynamic trajectory model.

[0054] Step 204: Based on the deviation between the expected power and the actual power of the hydro-generator unit at the current moment, the virtual control quantity of the power derivative of the hydro-generator unit at the current moment is obtained through an integral virtual control law.

[0055] Step 205: Based on the operating data at the current moment of the current compensation cycle, determine the operating condition at the current moment, and call the head change rate decomposition model under the corresponding operating condition to obtain the predicted value of the external driving term of the head change rate and the predicted value of the opening change sensitivity at the current moment; the operating conditions include the stable period operating condition and the severe disturbance period operating condition; the head prediction model is obtained by iteratively training a preset neural network model using a sample dataset, and the sample dataset is constructed based on the operating data of the second preset number of time periods before the current compensation cycle under the stable period operating condition or the severe disturbance period operating condition and the corresponding real head data of the hydro-generator unit.

[0056] Step 206: Establish the guide vane opening inertial model of the hydro-generator unit, and input the predicted values ​​of the external driving terms of the power derivative virtual control quantity, the head change rate, and the opening change sensitivity into the guide vane opening inertial model of the hydro-generator unit to obtain the guide vane opening change rate of the hydro-generator unit at the current moment.

[0057] Step 207: Control the operation of the hydro-generator unit based on the guide vane opening change rate, and redetermine the expected power at the next moment of the current compensation cycle to perform error compensation at the next moment, until the last moment of the current compensation cycle is reached, and calculate the compensation repair rate of the current compensation cycle.

[0058] Step 208: If the compensation repair rate of the current compensation cycle is within the preset range, then the error compensation repair is completed.

[0059] Step 209: If the compensation repair rate of the current compensation cycle is not within the preset range, the cumulative energy error of the current compensation cycle is used as the initial power generation energy error of the next compensation cycle, and the preset energy error index attenuation coefficient and total compensation time are adjusted to perform error compensation repair again.

[0060] By implementing steps 201 to 209 above, this application can solve the technical problem that the hydropower station load control method cannot accurately track the grid command under the conditions of head fluctuation and load change, and needs to make up for the energy gap by abandoning water or extending the power generation time. It realizes active, accurate and efficient compensation for historical power generation energy error, and significantly improves the efficiency of water energy utilization and grid dispatch response capability.

[0061] Example 3, Hydropower station operation data includes load demand data and inflow data; turbine generator unit operation data includes guide vane opening data, head data, and power generation data.

[0062] Based on historical operational data, the initial power generation energy error for the current compensation cycle is determined, specifically including: Based on the load demand data of hydropower stations over a historical period, the cumulative power generation capacity to be generated in the previous period was calculated.

[0063] The actual generated energy is calculated based on the power generation data of the hydro-turbine generator units over a historical period.

[0064] The initial power generation error is obtained by subtracting the previously accumulated power generation energy from the actual power generation energy.

[0065] Example 4, The multi-stage expected power dynamic trajectory model is as follows: .

[0066] in, This represents the expected power generation at time t; This represents the load demand power at time t; This represents the preset energy error index attenuation coefficient; This represents the real-time cumulative energy error at time t; This indicates the time point at which the rapid compensation phase transitions to the regression convergence phase; This indicates the time point at which the regression convergence phase transitions to the steady-state locking phase; Indicates the total compensation time.

[0067] Based on the deviation between the expected power and the actual power of the hydro-generator unit at the current moment, the virtual control quantity of the power derivative of the hydro-generator unit at the current moment is obtained through an integral virtual control law, specifically including: Using an integral-type virtual control law, the virtual control quantity of the power derivative is calculated using the following formula: .

[0068] .

[0069] in, This represents the virtual control quantity of the power derivative; Indicates proportional gain; Indicates integral gain; This represents the power generation error at time t; This represents the expected power generation at time t; This represents the actual power generation at time t.

[0070] Example 5, Based on the operating data at the current moment of the current compensation cycle, the operating condition at the current moment is determined, specifically including: Based on the operational data at the current moment of the current compensation cycle, the head change rate and the indices of sudden inflow disturbance are calculated using the following formulas: .

[0071] .

[0072] in, This represents the rate of change of water head at time t; This represents the head data of the hydro-turbine generator unit at time t; This represents the head data of the hydro-turbine generator unit at time t-1; Indicates the time interval between adjacent moments; This represents the index of sudden changes in water inflow at time t; This represents the inflow rate data of the hydropower station at time t; This represents the inflow data of the hydropower station at time t-1.

[0073] If the time interval between the current moment and the last time the operating condition was switched is less than the preset minimum dwell time, then the operating condition of the previous moment will be used as the operating condition of the current moment; wherein, the initial moment of the current compensation cycle will be used as the initial time of the operating condition switch, and the operating condition of the initial moment of the current compensation cycle will be initialized as the stable period operating condition.

[0074] If the time interval between the current moment and the last operating condition switch moment is greater than or equal to the preset minimum dwell time, and the severe disturbance operating condition determination condition is met, then the operating condition at the current moment is determined to be a severe disturbance period operating condition; the severe disturbance operating condition determination condition includes the continuous first preset time period before the current moment. ,or ;in, Indicates the threshold for the rate of change of head; This indicates the threshold for changes in inbound flow.

[0075] If the time interval between the current moment and the last operating condition switch moment is greater than or equal to the preset minimum dwell time, and the stable operating condition determination condition is met, then the operating condition at the current moment is determined to be a stable operating condition; the stable operating condition determination condition includes the period within the second preset time interval preceding the current moment. and .

[0076] Example 6, The loss function of the head change rate decomposition model is: .

[0077] in, The loss function value of the head change rate decomposition model is represented by the trainable parameter θ of the neural network as the independent variable; N represents the total number of samples; k represents the k-th sample; This represents the predicted head data for the k-th sample; This represents the actual head data of the k-th sample; Represents the regularization coefficient; This represents the sum of squares of the trainable parameters θ of the neural network.

[0078] Example 7, An inertial model of the guide vane opening of a hydro-generator unit is established. The predicted values ​​of the power derivative virtual control quantity, the external driving term of the head change rate, and the predicted value of the opening change sensitivity are input into the inertial model to obtain the guide vane opening change rate of the hydro-generator unit at the current moment. Specifically, this includes: The theoretical value of the rate of change of guide vane opening at the current moment can be calculated using the following formula: .

[0079] in, This represents the theoretical value of the rate of change of the guide vane opening at time t; This represents the virtual control quantity of the power derivative; This represents the power generation constant of the water turbine; This represents the power generation flow rate at time t; Represents time t Predicted value of external driving term for head change rate under operating conditions. When the value is -1, it indicates a period of severe disturbance. =1 indicates a stable operating condition; Represents time t Predicted value of sensitivity to changes in opening degree under operating conditions; This represents the head data of the hydro-turbine generator unit at time t; Indicates the guide vane opening; This represents the partial derivative of the power generation flow rate with respect to the guide vane opening.

[0080] The theoretical value of the guide vane opening change rate is subjected to a limiting constraint process, so that the theoretical value of the guide vane opening change rate is greater than or equal to a preset minimum opening change rate threshold and less than or equal to a preset maximum opening change rate threshold.

[0081] The theoretical value of the guide vane opening change rate after amplitude limiting constraint treatment is input into the guide vane opening inertial model, and the guide vane opening change rate of the hydro-generator unit at the current moment is obtained by the following formula: .

[0082] .

[0083] in, This represents the rate of change of the guide vane opening at time t; Represents the inertial filter coefficients; Indicates the time interval between adjacent moments; This represents the time constant of the guide vane actuator.

[0084] Example 8, The hydropower station load error compensation method also includes constructing a real-time error compensation trend monitoring map for the current compensation cycle; the real-time error compensation trend monitoring map includes a head change trend curve, a guide vane opening change trend curve, a tracking comparison curve of expected power generation and actual power generation, and a comparison curve of the compensation effect between real-time cumulative actual power generation and real-time cumulative expected power generation.

[0085] Step 207 specifically includes: Based on the real-time error compensation trend monitoring graph, it is determined whether there is a severe oscillation; the judgment conditions for severe oscillation include: in the tracking comparison curve of expected power generation and actual power generation, the power deviation within the third consecutive preset time period before the current time is greater than the preset power deviation threshold, or in the guide vane opening change trend curve, there is a periodic fluctuation with an amplitude exceeding the preset opening change rate oscillation threshold and a frequency exceeding the preset frequency threshold within the third consecutive preset time period before the current time.

[0086] If there is severe oscillation, the proportional gain parameter and integral gain parameter of the integral virtual control law are adjusted and calibrated. Based on the adjusted proportional gain parameter and integral gain parameter, the power derivative virtual control quantity at the current moment is recalculated, and the guide vane opening change rate is redetermined based on the recalculated power derivative virtual control quantity.

[0087] The operation of the hydro-generator unit is controlled based on the guide vane opening change rate, and the real-time error compensation trend monitoring map is updated.

[0088] Based on the updated real-time error compensation trend monitoring map, the expected power at the next moment of the current compensation cycle is redefined, and it is determined whether there will be a violent oscillation to carry out error compensation at the next moment, until the last moment of the current compensation cycle is reached.

[0089] Based on the real-time error compensation trend monitoring map, the compensation and repair rate for the current compensation cycle is calculated.

[0090] Example 9, The calculation method for the compensation and repair rate specifically includes: If the current compensation period is the first compensation period, the compensation repair rate is calculated using the following formula: .

[0091] in, This indicates the compensation and repair rate for the first compensation cycle; This indicates the last moment of the first compensation cycle. Real-time cumulative energy error; This represents the initial power generation energy error in the first compensation cycle.

[0092] If the current compensation period is longer than the first compensation period, the compensation repair rate is calculated using the following formula: .

[0093] in, Indicates the first The compensation and repair rate during the compensation period; Indicates the first The compensation and repair rate during the compensation period; Indicates the first The last moment of the compensation period Real-time cumulative energy error; Indicates the first The initial power generation energy error during the compensation period.

[0094] Example 10, Taking the specific load error compensation process of a hydropower station as an example, like Figure 3 As shown, a method for compensating for load errors in a hydropower station is provided, comprising the following steps: Step 1: Obtain basic data of the hydropower station, calculate the cumulative power generation capacity to be generated in the early stage, compare it with the actual power generation capacity, and obtain the error of the power generation capacity in the early stage.

[0095] As an optional implementation method, step 1 specifically includes: ① Collect hydropower station operation data (hydropower station operation data) and turbine generator set equipment data (turbine generator set operation data); the hydropower station operation data includes historical operating time, historical load demand data, real-time load demand data, historical inflow data, and real-time inflow data; the turbine generator set equipment data includes historical turbine generator set opening data, real-time turbine generator set opening data, historical head data, real-time head data, historical power data, and real-time power data. The turbine generator set equipment data can be calculated using a power generation calculation model, a turbine flow model, and a turbine power generation efficiency model. The specific calculation formula for the turbine power generation calculation model is as follows: .

[0096] The turbine flow rate model specifically relates flow rate, turbine opening, and turbine head. This relationship is derived through regression calculations based on actual measured data from the unit. The calculation formula is as follows: .

[0097] in, For the number The power generation capacity of the hydro turbine, in units of ; The density of water, in units of ; The acceleration due to gravity is approximately ; For the number Real-time head of the turbine, in units of ; For the number The power generation efficiency of the hydro turbine; For the number The hydroelectric power generation flow rate, in units of ; For the number The turbine constant; Indicates the number is The opening degree of the turbine guide vanes.

[0098] ②Based on the historical load demand of the hydropower station's turbine generator units Compared with historical running time Calculate the cumulative power generation capacity required in the early stage. The specific calculation formula is as follows: .

[0099] in, This refers to the total number of turbines in a hydro-generator unit. For the number The historical power generation capacity of the water turbine; For the number The historical operating time of the water turbine.

[0100] ③ Obtain the actual generated energy based on the historical power generation data of the hydro-turbine generator units. The error in the initial power generation energy is calculated by subtracting it from the cumulative power generation energy to be generated in the previous period. The specific calculation formula is as follows: .

[0101] Step 2: Based on the real-time cumulative energy error and by introducing the energy error exponential decay coefficient, construct a three-stage dynamic trajectory of expected power, including a rapid compensation stage, a regression convergence stage, and a steady-state locking stage.

[0102] As an optional implementation method, step 2 specifically includes: ①The overall error correction control target should ensure that within the total compensation time... Within this range, the cumulative compensation power generation energy is satisfied, and the specific formula is as follows: .

[0103] .

[0104] in, Let be the real-time power of the hydro-generator unit at time t; hydroelectric power station ; for Time number is The power of the water turbine.

[0105] That is, the real-time cumulative energy error is derived. The time should be satisfied The real-time cumulative energy error is: .

[0106] ② To ensure rapid repair of accumulated energy errors and to guarantee that the corrected power remains stable at the desired power, a segmented compensation strategy is adopted for real-time accumulated energy errors. Specifically, the compensation time is divided into a rapid compensation stage, a regression convergence stage, and a steady-state locking stage. The specific compensation formula is as follows: .

[0107] .

[0108] .

[0109] .

[0110] in, For the rapid compensation phase, To quickly compensate for the stage nodes, The proportion of the correction time allocated to the rapid compensation phase will be determined based on actual demand. To return to the convergence phase, This is the node of the regression convergence phase. The proportion of the correction time during the regression convergence phase is determined based on actual needs. This is the steady-state locking phase. Indicates the total compensation time.

[0111] ③ By introducing an energy error exponential decay coefficient to control the error correction rate, a three-segment dynamic trajectory of the desired power is constructed. The specific construction formula is as follows: .

[0112] in, This represents the expected power generation at time t; This represents the preset energy error index attenuation coefficient, which needs to be calibrated based on the actual operation and compensation of the hydro-generator unit.

[0113] To ensure that the expected power does not exceed the actual operating range of the hydro-generator unit, the following constraints must be met: .

[0114] in, , Set the minimum and maximum power values ​​for the hydro-generator units respectively.

[0115] Step 3: Design a Nonlinear Dynamic Inversion (NDI) controller for adjusting the turbine generator's opening, using the difference between actual and desired power as input. The primary goal is to ensure that the actual power eventually approaches the desired power. The control objective is defined as the difference between the desired and actual power, which serves as the error of the NDI controller. An integral virtual control law is employed, using the control input as the virtual variable. Based on the chain rule, the final output is the power derivative of the turbine generator unit. Central difference estimation is used for the differential term, and the continuous head is dynamically decomposed into external driving terms trained by a neural network and terms directly affecting the opening changes.

[0116] In the above technical solution, step 3 specifically includes: ① To make the actual power eventually approach the desired power, and design the control objective as the difference between the desired power and the actual power, defined as the power generation error at time t of the NDI controller for adjusting the unit opening. The specific formula is as follows: .

[0117] .

[0118] ② An integral-type virtual control law is used, with the control input being a virtual variable. The specific formula is as follows: .

[0119] in, This represents the virtual control quantity of the power derivative; This represents the proportional gain, indicating the direct impact of the current power error on the output of the NDI controller for adjusting the unit's operating degree. Its function is to control the response rate to the power error. Integral gain represents the impact of accumulated real-time power error on the controller, used to eliminate steady-state error and improve control accuracy. This represents the power generation error at time t; This represents the expected power generation at time t; This represents the actual power generation at time t.

[0120] ③ According to the chain rule, the power derivative of the output hydro-generator unit is calculated using the following formula:

[0121] .

[0122] ④Among them, The central difference estimation is used, and the specific formula is as follows: .

[0123] ⑤ Due to real-time water head It is susceptible to changes in turbine opening and inflow rate, therefore its rate of change should be divided into two parts, specifically including the external driving term and the direct impact term of turbine opening changes. Its continuous-time decomposition formula is as follows: .

[0124] In the formula: This represents the rate of change of head of the hydro-generator unit at time t; This represents the rate of change of the guide vane opening of the hydro-generator unit at time t; The predicted value of the external driving term is the rate of change of head caused by changes in upstream water inflow, discharge, reservoir capacity, etc., after the neural network is trained, excluding changes in the opening degree. The predicted value of the sensitivity to changes in opening degree is the local sensitivity function after the neural network is trained, which represents the incremental contribution of the change in opening degree to the rate of change in head. For a moment The inbound flow rate.

[0125] Step 4: Based on the actual inflow scenario, time lag constraints are added to divide the data into stable periods and periods of severe disturbance. Corresponding input features are designed, and a neural network model is trained. The designed neural network model adopts a multi-layer fully connected network. Based on the actual inflow scenario, time lag constraints are added to divide the hydropower station data into stable periods and periods of severe disturbance. The input features are designed as unit head, unit opening degree, inflow, and time period identifiers for stable and severe disturbance periods. The specific outputs are external driving terms and terms directly affecting changes in opening degree. The designed loss function is the error between the predicted and actual values, where the predicted value is the result of the forward calculation of the neural network model, and the input values ​​are the changes in head and opening degree at the current moment.

[0126] As an optional implementation, step 4 specifically includes: ① In actual hydropower station operation, the head change rate fluctuates in a complex manner. The control system should adjust the controller according to different scenarios, which can be divided into stable periods and periods of severe disturbance. The criteria for division are as follows: .

[0127] .

[0128] .

[0129] .

[0130] in, This represents the rate of change of water head at time t; This represents the head data of the hydro-turbine generator unit at time t; This represents the head data of the hydro-turbine generator unit at time t-1; Indicates the time interval between adjacent moments; This represents the index of sudden changes in water inflow at time t; This represents the inflow rate data of the hydropower station at time t; This represents the inflow data of the hydropower station at time t-1; , These are the conditions for transitioning between periods of severe disturbance and periods of stability.

[0131] To prevent frequent fluctuations in the control system during operation, a time hysteresis constraint is added. The specific constraint conditions are as follows: 1. To enter a period of severe disturbance, the following conditions must be met: .

[0132] .

[0133] in, This indicates the total number of time periods used to determine the period of severe disturbance. This indicates the start of a period of severe disturbance. One judgment period; This refers to the entry time (first preset time). This is equivalent to: entering a period of severe disturbance requires... continued .

[0134] 2. To return to a stable period, the following conditions must be met: .

[0135] .

[0136] in, This indicates the total number of time periods used to determine the return to a stable period; Indicates the return to a stable period. One judgment period; This is the exit time (second preset time). Equivalent to: Returning to a stable period requires... continued .

[0137] 3. Minimum stay time: .

[0138] in, This refers to the time when the current operating condition was switched over to the previous time. This is the minimum dwell time.

[0139] ②Based on historical data of unit head, unit operation, and inflow during periods of stable and severe disturbance. It also incorporates markers for periods of stability and periods of severe disturbance, thereby enabling training. , The specific input features are: .

[0140] .

[0141] In the formula: The k-th sample number corresponds to the training discrete time step; The period is divided into a stable period and a period of severe disturbance. =-1 indicates a period of severe disturbance ( Operating conditions When =1, it indicates a stationary period ( Operating conditions.

[0142] ③ A multilayer perceptron (MLP) is used to design the neural network model. The specific output layer is as follows: .

[0143] The forward calculation is defined to predict the water head at the next moment, and the specific formula is as follows: .

[0144] The loss function is designed to account for the error between the predicted and actual values. The specific formula is as follows: .

[0145] in, This represents the output of the k-th sample; This represents the actual head data of the k-th sample; This is a regularization term used to reduce the risk of poor fitting. For all trainable parameters of the neural network; Sum the squares of all parameters; This is a hyperparameter used to control the regularization strength, typically 1e. -5 ~1e -3 .

[0146] Step 5: Dynamically invert and derive the formula for calculating the guide vane opening change rate, construct the guide vane opening inertial model, and couple the NDI controller output.

[0147] As an optional implementation, step 5 specifically includes: ① This allows for the linear transformation of the nonlinear system, and the rate of change of the guide vane opening of the hydro-generator unit can be dynamically inverted and derived. The specific formula is as follows:

[0148] .

[0149] ②Based on the actual changes in the hydro-generator units, The specific formula for applying amplitude limiting constraints is as follows: .

[0150] in, This indicates the preset minimum opening change rate threshold; This indicates the preset maximum opening change rate threshold.

[0151] ③ Considering the actuator response lag, an inertial model of the guide vane opening of the hydro-generator unit is established. The specific model should be constructed according to the characteristics of the hydro-generator unit. For ease of explanation, it is modeled as a first-order inertial model here as an example. The inertial model is coupled to the NDI controller output, and the results are subject to saturation constraints, thereby eliminating the guide vane opening response error. The specific formula is as follows: .

[0152] .

[0153] .

[0154] .

[0155] in, This represents the rate of change of the guide vane opening at time t; Represents the inertial filter coefficients; Indicates the time interval between adjacent moments; This represents the time constant of the guide vane actuator; This represents the guide vane opening at time t+1; This represents the guide vane opening at time t.

[0156] Step 6: Based on head changes, guide vane opening changes, generator power tracking, and energy compensation effectiveness, a real-time error compensation trend monitoring graph is constructed to evaluate the error compensation effect. The real-time energy generated by the turbine generator unit is summed with the target energy generated. The energy compensation repair rate is calculated, and the error compensation effect is evaluated. If the graph curve shows severe oscillations, the proportional gain and integral gain parameters are recalibrated, and the operation is repeated. If the energy compensation repair is not within the specified range, the energy error exponential decay coefficient and compensation time are recalibrated, and the operation is repeated until the energy compensation repair is complete. Step 6 solves the problem of difficulty in real-time evaluation of error compensation effectiveness, achieving full visualization and quantification of the compensation process.

[0157] As an optional implementation, step 6 specifically includes: ① The real-time energy generated by the hydro-generator unit and the energy demanded by the load are calculated separately, and the accumulated energy demanded by the load is added to the energy generated in the previous period, which is recorded as the target energy generated. The specific formula is as follows: .

[0158] .

[0159] .

[0160] in, for Real-time cumulative power generation at any given moment; for Real-time cumulative load energy demand at any given moment; for Real-time cumulative expected power generation energy at any given moment.

[0161] ② When performing compensation, it is necessary to first construct a real-time error compensation trend monitoring map, which specifically includes the following relationships of the hydro-generator unit: a. real-time head change; b. guide vane opening change; c. real-time system power, expected system power, and user target power; d. real-time cumulative actual power generation and real-time cumulative expected power generation.

[0162] ③ During actual operation, staff can monitor the real-time error compensation of the hydro-generator unit throughout the entire process using the real-time error compensation trend monitoring graph. If strong oscillations are detected in the curve, the proportional gain parameters can be recalibrated. and integral gain parameter Update the current cumulative error and repeat the above steps to ensure complete energy compensation.

[0163] ④ After the compensation is completed, calculate the energy compensation and restoration rate. The specific formula is as follows: .

[0164] in, This indicates the compensation and repair rate for the first compensation cycle; This indicates the last moment of the first compensation period. Real-time cumulative energy error; This represents the initial power generation energy error in the first compensation cycle.

[0165] If in the end If the energy compensation and repair are satisfactory, the operation can be terminated. Minor energy errors can be readjusted quickly using conventional methods or by repeating the above steps. Conversely, if the energy error is not within this range, the energy error exponential decay coefficient can be recalibrated. Compensation time Repeat the above steps until the energy compensation is complete.

[0166] Each time the operation is repeated, the current cumulative error should be used as the energy error for the next period of power generation, and the energy compensation and repair rate should be proportionally added to the previous energy compensation and repair rate, as shown in the following formula: .

[0167] .

[0168] in, Indicates the first The initial power generation energy error during the compensation period; Indicates the first Real-time cumulative expected power generation energy during the compensation period; Indicates the first Real-time cumulative power generation during the compensation period; Indicates the first The compensation and repair rate during the compensation period; Indicates the first The compensation and repair rate during the compensation period; Indicates the first The last moment of the first compensation cycle Real-time cumulative energy error; Indicates the first The initial power generation energy error during the compensation period.

[0169] like Figure 4 As shown, a schematic diagram of the user's actual operation process is provided.

[0170] The present application will now be described in detail using the inflow rate, head of Unit 5, simulated power generation, simulated power generation flow, and simulated opening data of the A Town hydropower station as examples.

[0171] Step 1: Obtain basic data of the hydropower station, calculate the cumulative power generation capacity to be generated in the early stage, compare it with the actual power generation capacity, and obtain the error of the power generation capacity in the early stage.

[0172] This embodiment uses data such as the inflow rate of the A Town hydropower station and the head of Unit 5 for four consecutive days, and employs a simulation operation model to simulate the actual power generation, flow rate, and operating status of Unit 5. Meanwhile, to highlight the corrective function of this application, it is directly assumed here that the previous power generation energy error is -300MWh.

[0173] Step 2: Based on the real-time cumulative energy error and by introducing the energy error exponential decay coefficient, construct a three-stage dynamic trajectory of expected power, including a rapid compensation stage, a regression convergence stage, and a steady-state locking stage.

[0174] According to the default settings, the current introduced energy error exponential decay coefficient is 0.3. In this embodiment, the total adjustment time selected by the user is 1440 min (24 h). The constructed three-segment expected power dynamic trajectory time range T1 is 0.6 times the total adjustment time, T2 is 0.8 times the total adjustment time, and the specific fast compensation stage time range is [0, 864 min), the regression convergence stage time range is [864 min, 1152 min), and the steady-state locking stage time range is [1152 min, 1440 min].

[0175] Step 3: Design a NDI controller for adjusting the unit opening with the difference between the actual power and the expected power as input. Use central difference estimation for the differential term and dynamically decompose the continuous head into external driving terms trained by the neural network and terms directly affected by the opening change.

[0176] The design of the NDI controller for adjusting the unit opening, as well as the calculation methods for the differential term, external drive term, and direct impact term of the opening change, are consistent with those described in this application. Specifically, the differential term calculation uses simulated power generation flow data from Unit 5 of the A Town hydropower station for central difference estimation; the calculation of the external drive term and the direct impact term of the opening change employs a neural network model, with specific training data including the inflow of the A Town hydropower station over the previous three days, the head of Unit 5, and simulated opening data of Unit 5; the initial proportional gain parameter is determined. The default value is 1.2, and the integral gain parameter is... The default value is 0.2.

[0177] Step 4: Based on the actual water inflow scenario, time lag constraints are added to divide the period into a stable period and a period of severe disturbance. Corresponding input features are designed and the neural network model is trained.

[0178] Specifically, the threshold for the rate of change of water head during the stable period and the period of severe disturbance is 0.01 m / min, and the threshold for sudden change in inflow is 0.01 m. 3 The minimum time interval is 10 minutes, with entry time being 5 minutes, exit time being 5 minutes, and minimum dwell time being 10 minutes. Specifically, 1 and -1 are used as identifiers to represent the stable period and the period of severe disturbance, respectively. The 4 days of data used are divided, and the labeling of the first 3 days is also used as training data input for neural network model training.

[0179] Step 5: Dynamically invert and derive the formula for calculating the guide vane opening change rate, construct the guide vane opening inertial model, and couple the NDI controller output.

[0180] The inverse derivation of the formula for calculating the rate of change of guide vane opening is consistent with that described in this application. The constructed guide vane opening inertial model uses a first-order inertial model as a reference. The actual guide vane opening inertial model should be adjusted according to the unit conditions. It is only used here for the purpose of illustrating the example.

[0181] Step 6: Based on changes in head, guide vane opening, unit power generation tracking, and energy compensation effectiveness, construct a real-time error compensation trend monitoring map to evaluate the error compensation effect.

[0182] Based on actual power generation dispatch scenarios, users typically input their desired power output in 15-minute intervals. This simulation uses a time series of desired power outputs to mimic actual user operations. Specific data used includes the inflow rate of Unit 5 in Town A and the head data for Unit 5. The actual changes in the unit's performance are simulated using a simulation operation model. This leads to the construction of a real-time error compensation trend monitoring graph, such as... Figure 5 This is a graph showing the trend of water head changes. Figure 6 This is a graph showing the trend of guide vane opening. Figure 7 This is a curve showing the tracking and comparison between expected and actual power generation. Figure 8 This is a curve comparing the compensation effect between the real-time accumulated actual power generation and the real-time accumulated expected power generation. The final calculated accumulated actual power generation was 3173.6390 MWh, and the sum of the deviations between the accumulated target power generation and the initial energy was 3178.0000 MWh. The energy compensation and repair rate was 99.55%, which met the system shutdown requirements, and the energy compensation was complete.

[0183] Based on the same inventive concept, this application also provides a hydropower station load error compensation system for implementing the aforementioned hydropower station load error compensation method. The solution provided by this system is similar to the implementation scheme described in the above method; therefore, the specific limitations in one or more embodiments of the hydropower station load error compensation system provided below can be found in the limitations of the hydropower station load error compensation method described above, and will not be repeated here.

[0184] In one exemplary embodiment, such as Figure 9 As shown, a hydropower station load error compensation system is provided, comprising: The data acquisition module 301 is used to acquire historical time period operation data during the current compensation cycle; the operation data includes hydropower station operation data and turbine generator unit operation data; the historical time period includes the start time of the current compensation cycle and a first preset number of times before the start time.

[0185] The initial error determination module 302 is used to determine the initial power generation energy error of the current compensation cycle based on the operating data of historical time periods.

[0186] The expected power determination module 303 is used to construct a multi-stage expected power dynamic trajectory model based on the initial power generation energy error, the preset energy error index decay coefficient and the total compensation time of the current compensation cycle, and to determine the expected power at the current moment of the current compensation cycle based on the multi-stage expected power dynamic trajectory model.

[0187] The virtual control quantity calculation module 304 is used to obtain the virtual control quantity of the power derivative of the hydro-generator unit at the current moment based on the deviation between the expected power and the actual power of the hydro-generator unit at the current moment through an integral virtual control law.

[0188] The operating condition identification and head decomposition module 305 is used to determine the operating condition at the current moment based on the operating data at the current moment of the current compensation cycle, and call the head change rate decomposition model under the corresponding operating condition to obtain the predicted value of the external driving term of the head change rate and the predicted value of the opening change sensitivity at the current moment; the operating conditions include the stable period operating condition and the severe disturbance period operating condition; the head prediction model is obtained by iteratively training a preset neural network model using a sample dataset, and the sample dataset is constructed based on the operating data of the second preset number of time periods before the current compensation cycle under the stable period operating condition or the severe disturbance period operating condition and the corresponding real head data of the turbine generator unit.

[0189] The guide vane opening change rate determination module 306 is used to establish the guide vane opening inertial model of the hydro-generator unit, and input the power derivative virtual control quantity, the predicted value of the external driving term of the head change rate, and the predicted value of the opening change sensitivity into the guide vane opening inertial model of the hydro-generator unit to obtain the guide vane opening change rate of the hydro-generator unit at the current moment.

[0190] The closed-loop control and monitoring module 307 is used to control the operation of the hydro-generator unit based on the guide vane opening change rate, and redetermine the expected power at the next moment of the current compensation cycle to perform error compensation at the next moment, until the last moment of the current compensation cycle is reached, and calculate the compensation and repair rate of the current compensation cycle.

[0191] The compensation completion determination module 308 is used to complete the error compensation and repair if the compensation repair rate of the current compensation cycle is within a preset range.

[0192] The compensation iteration module 309 is used to take the cumulative energy error of the current compensation cycle as the initial power generation energy error of the next compensation cycle if the compensation repair rate of the current compensation cycle is not within the preset range, and adjust the preset energy error index attenuation coefficient and total compensation time to perform error compensation repair again.

[0193] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 10 As shown, the computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores data related to hydropower station load error compensation processing. The I / O interfaces are used for information exchange between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a hydropower station load error compensation method.

[0194] Those skilled in the art will understand that Figure 10The structures shown are merely block diagrams of some structures related to the present application and do not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than shown in the figures, or combine certain components, or have different component arrangements. In an exemplary embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0195] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0196] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0197] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0198] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0199] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0200] The above embodiments are merely preferred technical solutions of the present invention and should not be considered as limitations on the present invention. The embodiments and features described in these embodiments can be arbitrarily combined without conflict. The scope of protection of the present invention should be limited to the technical solutions described in the claims, including equivalent substitutions of the technical features described in the claims. That is, equivalent substitutions and improvements within this scope are also within the scope of protection of the present invention.

Claims

1. A method for compensating for load errors in a hydropower station, characterized in that, Includes the following steps: S1, in the current compensation cycle, obtain the running data for the historical time period; S2, based on historical operating data, determines the initial power generation energy error for the current compensation cycle; S3. Based on the initial power generation energy error, the preset energy error index decay coefficient and the total compensation time of the current compensation cycle, a multi-stage expected power dynamic trajectory model is constructed, and the expected power at the current moment of the current compensation cycle is determined based on the multi-stage expected power dynamic trajectory model. S4, based on the deviation between the expected power and the actual power of the hydro-generator unit at the current moment, the virtual control quantity of the power derivative of the hydro-generator unit at the current moment is obtained through the integral virtual control law; S5. Based on the operating data at the current moment of the current compensation cycle, determine the operating condition at the current moment, and call the head change rate decomposition model under the corresponding operating condition to obtain the predicted value of the external driving term of the head change rate and the predicted value of the opening change sensitivity at the current moment. S6. Establish the guide vane opening inertial model of the hydro-generator unit, and input the predicted values ​​of the external driving terms of the power derivative virtual control quantity, the head change rate, and the opening change sensitivity into the guide vane opening inertial model of the hydro-generator unit to obtain the guide vane opening change rate of the hydro-generator unit at the current moment. S7, based on the guide vane opening change rate, control the operation of the hydro-generator unit, and redetermine the expected power at the next moment of the current compensation cycle to perform error compensation at the next moment, until the last moment of the current compensation cycle is reached, and calculate the compensation repair rate of the current compensation cycle. S8, construct a real-time error compensation trend monitoring map for the current compensation cycle; The real-time error compensation trend monitoring graph includes the head change trend curve, the guide vane opening change trend curve, the tracking comparison curve between expected power generation and actual power generation, and the compensation effect comparison curve between real-time cumulative actual power generation and real-time cumulative expected power generation. S9, Determine if the compensation repair rate of the current compensation cycle is within the preset range, then complete the error compensation repair; if the compensation repair rate of the current compensation cycle is not within the preset range, then use the cumulative energy error of the current compensation cycle as the initial power generation energy error of the next compensation cycle, and adjust the preset energy error index attenuation coefficient and total compensation time, and repeat the error compensation repair.

2. The hydropower station load error compensation method according to claim 1, characterized in that: The operational data includes hydropower station operational data and turbine generator unit operational data; the historical time period includes the start time of the current compensation cycle and the first preset number of times before the start time; the hydropower station operational data includes load demand data and inflow data; the turbine generator unit operational data includes guide vane opening data, head data, and power generation data; S2-1, Calculate the cumulative power generation capacity to be generated in the previous period based on the load demand data of hydropower stations over a historical period; S2-2, calculate the actual generated energy based on the power generation data of the hydro-turbine generator units over a historical period; S2-3, subtract the accumulated power generation energy from the actual power generation energy to obtain the initial power generation energy error.

3. The hydropower station load error compensation method according to claim 1, characterized in that: In S3, the multi-stage desired power dynamic trajectory model is as follows: ； in, This represents the expected power generation at time t; This represents the load demand power at time t; This represents the preset energy error index attenuation coefficient; This represents the real-time cumulative energy error at time t; This indicates the time point at which the rapid compensation phase transitions to the regression convergence phase; This indicates the time point at which the regression convergence phase transitions to the steady-state locking phase; Indicates the total compensation time.

4. The hydropower station load error compensation method according to claim 1, characterized in that: in In S4, an integral virtual control law is used, and the virtual control quantity of the power derivative is calculated using the following formula: ； ； in, This represents the virtual control quantity of the power derivative; Indicates proportional gain; Indicates integral gain; This represents the power generation error at time t; This represents the expected power generation at time t; This represents the actual power generation at time t.

5. The hydropower station load error compensation method according to claim 1, characterized in that: In S5, specifically including, S5-1, based on the operational data at the current moment of the current compensation cycle, the head change rate and the inflow disturbance index are calculated using the following formula: ； ； in, This represents the rate of change of water head at time t; This represents the head data of the hydro-turbine generator unit at time t; This represents the head data of the hydro-turbine generator unit at time t-1; Indicates the time interval between adjacent moments; This represents the index of sudden changes in water inflow at time t; This represents the inflow rate data of the hydropower station at time t; This represents the inflow data of the hydropower station at time t-1; S5-2, If the time interval between the current moment and the last working condition switching moment is less than the preset minimum dwell time, then the operating condition of the previous moment is taken as the operating condition of the current moment; wherein, the initial moment of the current compensation cycle is taken as the initial working condition switching moment, and the operating condition of the initial moment of the current compensation cycle is initialized as the stable period working condition. S5-3, if the time interval between the current moment and the last time the operating condition switch is greater than or equal to the preset minimum dwell time, and the severe disturbance condition determination criteria are met, then the operating condition at the current moment is determined to be a severe disturbance period condition; the severe disturbance condition determination criteria include the continuous first preset time interval before the current moment. ,or ;in, Indicates the threshold for the rate of change of head; Indicates the threshold for changes in inbound flow; S5-4, if the time interval between the current moment and the last time the operating condition switch is greater than or equal to the preset minimum dwell time, and the stable operating condition determination condition is met, then the operating condition at the current moment is determined to be a stable operating condition; the stable operating condition determination condition includes the second preset time period preceding the current moment. and .

6. The hydropower station load error compensation method according to claim 1, characterized in that: in In S5, the loss function of the head change rate decomposition model is: ； in, The loss function value of the head change rate decomposition model is represented by the trainable parameter θ of the neural network as the independent variable; N represents the total number of samples; k represents the k-th sample; This represents the predicted head data for the k-th sample; This represents the actual head data of the k-th sample; Represents the regularization coefficient; This represents the sum of squares of the trainable parameters θ of the neural network.

7. The hydropower station load error compensation method according to claim 1, characterized in that: in S6 specifically includes, S6-1, the theoretical value of the rate of change of the guide vane opening at the current moment can be calculated using the following formula: ； in, This represents the theoretical value of the rate of change of the guide vane opening at time t; This represents the virtual control quantity of the power derivative; This represents the power generation constant of the water turbine; This represents the power generation flow rate at time t; Represents time t Predicted value of external driving term for head change rate under operating conditions. When the value is -1, it indicates a period of severe disturbance. =1 indicates a stable operating condition; Represents time t Predicted value of sensitivity to changes in opening degree under operating conditions; This represents the head data of the hydro-turbine generator unit at time t; Indicates the guide vane opening; This represents the partial derivative of the power generation flow rate with respect to the guide vane opening. S6-2, apply a limiting constraint to the theoretical value of the guide vane opening change rate, so that the theoretical value of the guide vane opening change rate is greater than or equal to the preset minimum opening change rate threshold and less than or equal to the preset maximum opening change rate threshold. S6-3, input the theoretical value of the guide vane opening change rate after the amplitude limiting constraint treatment into the guide vane opening inertial model, and obtain the guide vane opening change rate of the hydro-generator unit at the current moment through the following formula: ； ； in, This represents the rate of change of the guide vane opening at time t; Represents the inertial filter coefficients; Indicates the time interval between adjacent moments; This represents the time constant of the guide vane actuator.

8. The hydropower station load error compensation method according to claim 1, characterized in that: in S7 specifically includes, S7-1, based on the real-time error compensation trend monitoring map, determines whether there is a violent oscillation; The criteria for judging severe oscillations include: in the tracking comparison curve of expected power generation and actual power generation, the power deviation is greater than the preset power deviation threshold for the third consecutive preset time period before the current moment; or in the guide vane opening change trend curve, there are periodic fluctuations with amplitude exceeding the preset opening change rate oscillation threshold and frequency exceeding the preset frequency threshold for the third consecutive preset time period before the current moment. S7-2, If there is severe oscillation, adjust the proportional gain parameter and integral gain parameter of the proportional definite integral virtual control law, recalculate the power derivative virtual control quantity at the current moment based on the adjusted proportional gain parameter and integral gain parameter, and redetermine the guide vane opening change rate based on the recalculated power derivative virtual control quantity. S7-3 controls the operation of the hydro-generator unit based on the guide vane opening change rate and updates the real-time error compensation trend monitoring map; S7-4, based on the updated real-time error compensation trend monitoring map, redetermines the expected power at the next moment of the current compensation cycle, and determines whether there will be a violent oscillation to carry out error compensation at the next moment, until the last moment of the current compensation cycle is reached. S7-5 calculates the compensation and repair rate for the current compensation cycle based on the real-time error compensation trend monitoring map.

9. The hydropower station load error compensation method according to claim 8, characterized in that: The calculation method for the compensation and repair rate specifically includes: If the current compensation period is the first compensation period, the compensation repair rate is calculated using the following formula: ； in, This indicates the compensation and repair rate for the first compensation cycle; This indicates the last moment of the first compensation cycle. Real-time cumulative energy error; This represents the initial power generation energy error in the first compensation period; If the current compensation period is longer than the first compensation period, the compensation repair rate is calculated using the following formula: ； in, Indicates the first The compensation and repair rate during the compensation period; Indicates the first The compensation and repair rate during the compensation period; Indicates the first The last moment of the compensation period Real-time cumulative energy error; Indicates the first The initial power generation energy error during the compensation period.

10. The error compensation system of the hydropower station load error compensation method according to any one of claims 1 to 9, characterized in that: The error compensation system includes a data acquisition module for data exchange, an initial error determination module, a desired power determination module, a virtual control quantity calculation module, an operating condition identification and head decomposition module, a guide vane opening change rate determination module, a closed-loop control and monitoring module, a compensation completion judgment module, and a compensation iteration module.

Images