A Multimodal Optimal Load Regulation Method Based on Head-Power Coupling Characteristics

Abstract

This invention discloses a multimodal optimal load regulation method based on head-power coupling characteristics, relating to the field of hydropower generation technology. The method includes: S1 constructing a three-dimensional dynamic coupling model; S2 acquiring unit operating data and AGC commands; S3 checking if the regulation conditions are met, proceeding to S4 if yes, otherwise proceeding to S10; S4 analyzing to obtain the predicted power generation and feedforward compensation opening; S5 determining the optimal guide vane opening; S6 determining the target guide vane opening; S7 determining the input parameters of the guide vane actuator; S8 adjusting the guide vane opening and outputting the power signal to the feedback correction layer to analyze the current controller output; S9 updating the current three-dimensional dynamic coupling model; S10 setting n=n+1, k=k+1, and returning to S2; establishing a three-dimensional dynamic coupling model of head-power-guide vane opening, enabling online real-time parameter correction under head disturbance conditions, and improving model prediction accuracy; the composite control mechanism of feedforward prediction and feedback correction significantly shortens the regulation time to 9.2s, suppresses overshoot to 1.8%, and reduces steady-state error to ±0.12MW.

Description

Technical Field

[0001] This invention relates to the field of hydropower technology, and in particular to a multimodal optimal load regulation method based on head-power coupling characteristics. Background Technology

[0002] Currently, load regulation of hydroelectric generating units mainly relies on two typical schemes: one is traditional PID feedback control, which dynamically adjusts the guide vane opening by detecting the deviation between the actual power of the unit and the set load using a proportional-integral-derivative algorithm. The other is head feedforward compensation control, which introduces an open-loop feedforward channel of head measurement value on the basis of PID, and approximately corrects the guide vane command based on the static head-opening relationship table. Although the traditional PID feedback control scheme is simple in structure, it has significant limitations: head fluctuations are only regarded as external disturbances, and the lag of the power closed-loop feedback, typically 5-15 seconds, leads to slow regulation response. Moreover, fixed PID parameters are difficult to adapt to the strong coupling nonlinearity of head and power, which easily causes power overshoot under large load disturbances, and the regulation performance deteriorates sharply in the low head region. Although head feedforward compensation control partially alleviates the static influence of head, it does not establish a dynamic coupling model between head and power, resulting in the inability to accurately predict the inertial delay of the water flow in the feedforward quantity; especially in the load shedding transient process with significant water hammer effect, the power reverse adjustment phenomenon is prominent. Existing technologies lack an adaptive mechanism for multimodal operation across the entire operating range, and have not achieved coordinated optimization of head disturbance feedforward prediction and power closed-loop feedback, which limits the regulation accuracy and response speed. Summary of the Invention

[0003] The purpose of this invention is to design a multimodal optimal load regulation method based on head-power coupling characteristics to solve the above problems.

[0004] The present invention achieves the above objectives through the following technical solutions:

[0005] Multimodal optimal load regulation methods based on head-power coupling characteristics include:

[0006] S1. Construct a three-dimensional dynamic coupling model of the head parameter H, guide vane opening Y, and active power P, expressed as: Where A, B, and C are the coefficients to be fitted. Y min Y is the minimum guide vane opening. max R represents the maximum guide vane opening. 2 The coefficient of determination is used to verify the model's fitting accuracy and is expressed as: , For the i-th group of measured active power, To predict the power generation of the k-th model, The actual active power is the average value, and n is the number of sampling data points;

[0007] S2. Obtain the unit operation data and AGC commands at time t. The AGC commands include the grid automatic generation control power setpoint. The unit's operating data includes real-time head parameters. Real-time guide vane opening Y t Measured active power P meas Unit speed, head sensitivity coefficient and rated head H r ;

[0008] S3. Based on the unit operating data and power setpoint at time t. Determine whether the adjustment condition is met. If yes, proceed to S4; otherwise, proceed directly to S10.

[0009] S4, Feedforward prediction layer based on real-time head parameters Real-time guide vane opening Y t and power setting value The predicted power generation is obtained by analyzing the current three-dimensional dynamic coupling model. and feedforward compensation opening k is the number of iterations of the current three-dimensional dynamic coupling model;

[0010] S5, Based on head parameters and predicted power generation Query the database to obtain the optimal guide vane opening. ;

[0011] S6. Utilizing the optimal guide vane opening and feedforward compensation opening Determine the target opening Y of the guide vane actuator;

[0012] S7. Based on the guide vane target opening Y and the controller output at the previous moment... Determine the input parameters of the guide vane actuator, including the proportional gain coefficient. Integral gain coefficient and differential gain coefficient ;

[0013] S8. The guide vane actuator adjusts the guide vane opening according to the input parameters and outputs a power signal. The current controller output is analyzed by the feedback correction layer. , represented as: ,in, As an anti-saturation compensation factor, Let be a saturation function, expressed as: ;

[0014] S9, Based on extracted power signal and predicted power generation Update the current three-dimensional dynamic coupling model;

[0015] S10. Let n = n+1, k = k+1, and return to S2.

[0016] The beneficial effects of this invention are as follows: a three-dimensional dynamic coupling model of head-power-guide vane opening is established to realize online real-time correction of parameters under head disturbance conditions, thereby improving the model prediction accuracy; a composite control mechanism of feedforward prediction and feedback correction is designed to significantly shorten the adjustment time to 9.2s, suppress overshoot to 1.8%, and reduce steady-state error to ±0.12MW; the frequency domain characteristics of the system are optimized to increase the phase margin of key frequency bands by more than 15°, effectively compensate for hydraulic lag effects, and ultimately achieve rapid and accurate tracking of load across the entire operating range. Attached Figure Description

[0017] Figure 1 This invention relates to the regulation and control architecture of a multimodal optimal load regulation method based on head-power coupling characteristics.

[0018] Figure 2 It is a comparison of the dynamic response of power point tracking;

[0019] Figure 3 This is the setting for head disturbance conditions;

[0020] Figure 4 It is a comparison of key performance indicators;

[0021] The corresponding figure labels are: Figure 3 The green line in the diagram represents the unadjusted power under head disturbance conditions. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0023] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0024] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0025] In the description of this invention, it should be understood that the terms "upper," "lower," "inner," "outer," "left," "right," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship commonly used when the product of this invention is in use, or the orientation or positional relationship commonly understood by those skilled in the art. They are only used to facilitate the description of this invention and to simplify the description, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0026] Furthermore, the terms "first," "second," etc., are used only to distinguish descriptions and should not be interpreted as indicating or implying relative importance.

[0027] In the description of this invention, it should also be noted that, unless otherwise explicitly specified and limited, terms such as "set" and "connection" should be interpreted broadly. For example, "connection" can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium; it can be a connection within two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0028] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0029] like Figure 1 As shown, the multimodal optimal load regulation method based on head-power coupling characteristics includes:

[0030] S1. Construct a three-dimensional dynamic coupling model of the head parameter H, guide vane opening Y, and active power P, expressed as: Where A, B, and C are the coefficients to be fitted. Y min Y is the minimum guide vane opening. max R represents the maximum guide vane opening. 2 The coefficient of determination is used to verify the model's fitting accuracy and is expressed as: , For the i-th group of measured active power, To predict the power generation of the k-th model, The actual active power is the average value, and n is the number of sampling data points;

[0031] S2. Obtain the unit operation data and AGC commands at time t. The AGC commands include the grid automatic generation control power setpoint. The unit's operating data includes real-time head parameters. Real-time guide vane opening Y t Measured active power P meas Unit speed, head sensitivity coefficient and rated head H r ;

[0032] S3. Based on the unit operating data and power setpoint at time t. Determine if the adjustment condition is met. If yes, proceed to S4; otherwise, proceed directly to S10. Specifically: if the current speed controller mode is power mode and the convergence condition is met, then the adjustment condition is met; if yes... and When the current speed controller is in the opening mode, the convergence condition is satisfied; when the current speed controller is in the opening mode, the convergence condition is satisfied. If so, the adjustment condition is met; among which, the power deviation threshold is... Power change threshold , S represents the rated power of the unit, and S represents the apparent power of the unit.

[0033] S4, Feedforward prediction layer based on real-time head parameters Real-time guide vane opening Y t and power setting value The predicted power generation is obtained by analyzing the current three-dimensional dynamic coupling model. and feedforward compensation opening , represented as: ,in, The head sensitivity coefficient, Here, k represents the rated head parameter, and k is the iteration number of the current three-dimensional dynamic coupling model.

[0034] S5, Based on head parameters and predicted power generation Query the database to obtain the optimal guide vane opening. ;

[0035] S6. Utilizing the optimal guide vane opening and feedforward compensation opening The target opening Y of the guide vane actuator is determined as follows: ;

[0036] S7. Based on the guide vane target opening Y and the controller output at the previous moment... Determine the input parameters of the guide vane actuator, including the proportional gain coefficient. Integral gain coefficient and differential gain coefficient ;

[0037] S8. The guide vane actuator adjusts the guide vane opening according to the input parameters and outputs a power signal. The current controller output is analyzed by the feedback correction layer. , represented as: ,in, As an anti-saturation compensation factor, Let be a saturation function, expressed as: ;

[0038] S9, Based on extracted power signal and predicted power generation Update the current 3D dynamic coupling model; update the model parameters of the current 3D dynamic coupling model. and adaptive gain matrix To obtain the updated model parameters of the three-dimensional dynamic coupling model. and the updated adaptive gain matrix , represented as: , , Let be the adaptive matrix for the k-th iteration. Let be the adaptive matrix for the (k+1)th iteration. For learning rate, The power deviation in the k-th iteration , The measured active power in the k-th iteration is... To predict the power generation of the model in the k-th iteration, The model input vector for the k-th iteration , Let the head parameters be those for the k-th iteration. The target guide vane opening for the kth iteration;

[0039] S10. Let n = n+1, k = k+1, and return to S2.

[0040] This method constructs a three-dimensional dynamic coupling model of head-power-guide vane opening that is adaptive to head, enabling online real-time parameter correction under head disturbance conditions and improving model prediction accuracy. A composite control mechanism combining feedforward prediction and feedback correction is designed, significantly shortening the settling time to 9.2s, suppressing overshoot to 1.8%, and reducing steady-state error to ±0.12MW. Based on the Lyapunov stability design mode switching convergence condition, oscillations caused by power noise are eliminated. The system's frequency domain characteristics are optimized, increasing the phase margin of key frequency bands by more than 15°, effectively compensating for hydraulic lag effects, and ultimately achieving rapid and accurate load tracking across the entire operating range.

[0041] Example

[0042] Based on big data analysis of unit operation, a three-dimensional dynamic coupling model of head parameter H, guide vane opening Y, and active power P was established. The optimal fitting equation was obtained by performing nonlinear regression on the operating data of a hydropower station under a 117m head condition using least squares support vector machine (LS-SVM). ; The guide vane opening range;

[0043] The coefficient of determination verifies the strong linearity of the three-dimensional dynamic coupling model under typical working conditions, and the active power is calculated. Then, the coefficient matrix of the three-dimensional dynamic coupling model is updated using an online parameter update algorithm. and adaptive gain matrix This enables dynamic correction of model parameters as the water head changes; In the formula: Target value of active power The power generation value predicted by the three-dimensional dynamic coupling model updated in the kth iteration;

[0044] Obtain the unit operating data and AGC commands at time t. The AGC commands include the grid automatic generation control power setpoint. The unit's operating data includes real-time head parameters. Real-time guide vane opening Y t Measured active power P meas Unit speed, head sensitivity coefficient and rated head H r ;

[0045] Based on the unit operating data and AGC instructions at time t Determine if the adjustment conditions are met. If yes, proceed with load adjustment; otherwise, return and reacquire the unit operating data and AGC commands for the next moment. Specifically: if the current governor mode is power mode and the convergence condition is met, then the adjustment conditions are met; if the conditions are met... and When the current speed controller is in the opening mode, the convergence condition is satisfied; when the current speed controller is in the opening mode, the convergence condition is satisfied. If so, the adjustment condition is met; among which, the power deviation threshold is... Power change threshold , S represents the rated power of the unit, and S represents the apparent power of the unit.

[0046] like Figure 1 As shown, the control system adopts a two-layer optimized structure, and the monitoring system receives AGC commands. Then, based on the head parameters Query the database to obtain the optimal guide vane opening setting. The feedforward compensation amount is input together with the feedforward compensation amount into the guide vane actuator. The feedforward compensation amount is calculated as follows: ;

[0047] The optimal guide vane opening setting value was obtained through querying. The output target guide vane opening of the guide vane actuator is obtained by using the feedforward compensation of the guide vane, and is expressed as: ;

[0048] After the guide vane actuator adjusts the guide vane opening, the corresponding power output will change accordingly. An anti-saturation compensation term is introduced, and the power output is then used to generate a power signal. As input to the feedback correction layer, which is an improved anti-saturation PID controller, the feedback correction layer analyzes the controller output at the current moment. , represented as: Among them, the proportional gain coefficient Integral gain coefficient and differential gain coefficient ; This is an anti-saturation compensation factor to suppress the nonlinear saturation characteristics of the guide vane actuator. Y represents the target guide vane opening. To derive the maximum output value of the actuator. Let be a saturation function, expressed as: ;

[0049] Historical data were processed using this method and tested on the HL-1000 hydraulic turbine simulation platform. The main performance indicators are compared in Table 1. Experimental results are as follows: Figure 2-4 As shown, this method reduces the root mean square error of power tracking to 0.098 MW under the condition of a sudden change in head of 117 m, which is 82.3% higher than the traditional method. Spectrum analysis shows that the phase margin of the system in the 0.1-2 Hz frequency band is improved by more than 15°, verifying the algorithm's compensation effect on hydraulic lag.

[0050] Table 1 Parameter Comparison Table

[0051]

[0052] The technical solutions of the present invention are not limited to the specific embodiments described above. Any technical modifications made in accordance with the technical solutions of the present invention fall within the protection scope of the present invention.

Claims

1. A multimodal optimal load regulation method based on head-power coupling characteristics, characterized in that, include: S1. Construct a three-dimensional dynamic coupling model of the head parameter H, guide vane opening Y, and active power P, expressed as: Where A, B, and C are the coefficients to be fitted. Y min Y is the minimum guide vane opening. max R represents the maximum guide vane opening. 2 The coefficient of determination is used to verify the model's fitting accuracy. S2. Obtain the unit operation data and AGC commands at time t. The AGC commands include the grid automatic generation control power setpoint. The unit's operating data includes real-time head parameters. Real-time guide vane opening Y t Measured active power P meas Unit speed, head sensitivity coefficient and rated head H r ; S3. Based on the unit operating data and power setpoint at time t. Determine whether the adjustment condition is met. If yes, proceed to S4; otherwise, proceed directly to S10. S4, Feedforward prediction layer based on real-time head parameters Real-time guide vane opening Y t and power setting value The predicted power generation is obtained by analyzing the current three-dimensional dynamic coupling model. and feedforward compensation opening k is the number of iterations of the current three-dimensional dynamic coupling model; S5, Based on head parameters and predicted power generation Query the database to obtain the optimal guide vane opening. ; S6. Utilizing the optimal guide vane opening and feedforward compensation opening Determine the target opening Y of the guide vane actuator; S7. Based on the target opening Y of the guide vane and the controller output at the previous moment... Determine the input parameters of the guide vane actuator, including the proportional gain coefficient. Integral gain coefficient and differential gain coefficient ; S8. The guide vane actuator adjusts the guide vane opening according to the input parameters and outputs a power signal. The current controller output is analyzed by the feedback correction layer. ; S9, Based on extracted power signal and predicted power generation Update the current three-dimensional dynamic coupling model; S10. Let n = n+1, k = k+1, and return to S2.

2. The multimodal optimal load regulation method based on head-power coupling characteristics according to claim 1, characterized in that, In S3, if the current speed controller is in power mode and the convergence condition is met, then the regulation condition is met; and When the current speed controller is in the opening mode, the convergence condition is satisfied; when the current speed controller is in the opening mode, the convergence condition is satisfied. If so, the adjustment condition is met; among which, the power deviation threshold is... Power change threshold , S represents the rated power of the unit, and S represents the apparent power of the unit.

3. The multimodal optimal load regulation method based on head-power coupling characteristics according to claim 1, characterized in that, In S4, the feedforward compensation opening Represented as: ,in, The head sensitivity coefficient, Here, k represents the rated head parameter, and k is the iteration number of the current three-dimensional dynamic coupling model.

4. The multimodal optimal load regulation method based on head-power coupling characteristics according to claim 1, characterized in that, The target aperture Y of the guide vane is expressed as: .

5. The multimodal optimal load regulation method based on head-power coupling characteristics according to claim 1, characterized in that, In S8, the controller output at the current moment. Represented as: ,in, As an anti-saturation compensation factor, Let be a saturation function, expressed as: .

6. The multimodal optimal load regulation method based on head-power coupling characteristics according to claim 1, characterized in that, In S9, update the model parameters of the current 3D dynamically coupled model. and adaptive gain matrix To obtain the updated model parameters of the three-dimensional dynamic coupling model. and the updated adaptive gain matrix , is represented as: , , Let be the adaptive matrix for the k-th iteration. Let be the adaptive matrix for the (k+1)th iteration. For learning rate, The power deviation in the k-th iteration , The measured active power in the k-th iteration is... To predict the power generation of the model in the k-th iteration, The model input vector for the k-th iteration , Let the head parameters be those for the k-th iteration. The target guide vane opening for the kth iteration.

Images