A method for parallel optimization control of guide vane and blade of axial flow variable pitch water turbine based on frequency disturbance decoupling

Abstract

A parallel optimization control method for guide vanes and propellers of axial-flow propeller turbines based on frequency disturbance decoupling is disclosed, belonging to the field of turbine speed control technology. This method first establishes a mathematical model of the system; then, a frequency disturbance decoupling device is set between the frequency measurement unit and the controller to decompose the grid frequency deviation signal into high-frequency and low-frequency disturbance components, which are input into dedicated controllers for guide vanes and propellers respectively to generate dynamic control signals; simultaneously, based on the current head and power setpoints, the turbine's comprehensive characteristic curve is queried to obtain the optimal opening of the guide vanes and propellers as the reference control signals; the dynamic control signals are superimposed with the corresponding reference signals to form independent final control commands for the guide vanes and propellers; and a particle swarm optimization algorithm is used to collaboratively optimize the controller parameters and the cutoff frequency of the decoupling device, automatically triggering re-optimization when operating conditions change significantly, achieving adaptive parallel control under all operating conditions, and significantly improving regulation response speed and operating efficiency.

Description

Technical Field

[0001] This invention belongs to the field of turbine speed control system technology, and specifically relates to a parallel optimization control method for guide vanes of axial-flow propeller turbines based on frequency disturbance decoupling. Background Technology

[0002] The speed control system of an axial-flow propeller turbine, as the core control equipment of a hydro-generator unit, plays a crucial role in maintaining grid frequency stability and the safe and efficient operation of the unit. Traditional control methods employ a series control architecture based on a fixed coordination relationship. A single PID controller outputs a control signal to drive the guide vanes, and the propeller blades follow the guide vane changes according to a preset head-guide vane opening coordination curve. While this method ensures high operating efficiency under steady-state conditions, it has significant limitations during dynamic regulation: firstly, the control mode where the propeller blades completely follow the guide vane movement results in inherent mechanical lag, causing the propeller blade response to lag significantly behind the guide vane movement, making it difficult to overcome the system's inherent inertia; secondly, the fixed-parameter coordination relationship is difficult to adapt to complex and variable operating conditions, easily leading to problems such as prolonged regulation time and increased overshoot under off-design conditions, severely affecting grid frequency stability.

[0003] To improve the performance of the speed control system, existing improvement schemes mainly focus on two directions: First, using intelligent algorithms to optimize the co-operation curves or PID parameters, such as the co-operation optimization method based on particle swarm optimization. Although this can improve efficiency under specific operating conditions, it does not change the series control architecture and cannot solve the fundamental response lag problem. Second, using advanced control algorithms to replace traditional PID control, such as fuzzy control and neural networks. Although this improves dynamic performance to some extent, the algorithm structure is complex, limiting its engineering practicality and making it difficult to guarantee stability under all operating conditions. These improvement methods have failed to effectively solve the dynamic response lag problem caused by the coupling of guide vane and blade movements, and have not fully utilized the potential of independent control of the dual actuators, thus restricting further improvement of the dynamic performance of the turbine speed control system.

[0004] The existing speed regulation technology for axial-flow propeller turbines has the following main defects and shortcomings: First, traditional series-coordinated control architectures inherently suffer from response lag. The "guide vane-led, blade-following" control mode inevitably causes blade movement to lag behind guide vane changes, and this mechanical lag severely restricts the system's dynamic response speed. When facing grid frequency fluctuations, the system's adjustment time is excessively long, overshoot increases significantly, and may even trigger power oscillations, making it difficult to meet the dynamic response requirements of modern power systems for rapid frequency regulation.

[0005] Secondly, existing control strategies are insufficient in handling the characteristics of strong coupling among multiple variables. Traditional methods treat the guide vanes and propeller blades as a coupled whole, ignoring the fundamental differences in their dynamic response characteristics. The control strategies fail to scientifically decompose and specifically address the spectral characteristics of frequency disturbances, resulting in the actuator's control potential not being fully realized and a significant decline in control performance under complex operating conditions.

[0006] Third, existing parameter optimization methods lack a systematic approach. Most improvement schemes either optimize only the coordination relationship or only tune the PID parameters, failing to coordinate the optimization of actuator characteristics, frequency disturbance characteristics, and control system parameters. This localized optimization approach makes it difficult for the system to maintain optimal performance across the entire operating range, resulting in a disconnect between control performance and operational requirements.

[0007] Finally, existing intelligent control methods lack practical engineering applicability. Although various advanced algorithms have theoretical advantages, they generally suffer from problems such as complex structure, difficulty in parameter tuning, and poor adaptability to extreme operating conditions. They also lack reliable adaptive mechanisms to cope with changes in actual operating conditions such as head fluctuations and sudden load changes. These shortcomings severely restrict the full realization of the frequency regulation potential of hydropower units, and there is an urgent need for a new control method that can fundamentally solve the problem of response lag, fully utilize the control potential of dual actuators, and achieve adaptive optimization under all operating conditions. Summary of the Invention

[0008] The technical problem this invention aims to solve is to provide a parallel optimization control method for guide vanes of axial-flow propeller turbines based on frequency disturbance decoupling. Addressing the inherent limitation of existing series-coordinated control architectures where the blade response inevitably lags behind the guide vane movement, and the inadequacy of fixed-parameter control strategies to adapt to complex and variable operating conditions, this invention proposes a parallel optimization control method based on frequency disturbance decoupling. By establishing independent parallel control channels for the guide vanes and propellers and introducing an intelligent optimization mechanism, the dynamic response speed and control quality of the hydropower unit during grid frequency regulation are effectively improved, meeting the high requirements of modern power systems for rapid frequency regulation.

[0009] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: A parallel optimization control method for guide vanes of an axial-flow propeller turbine based on frequency disturbance decoupling, comprising the following steps: Step 1: Establish mathematical models for each device in the axial-flow propeller turbine speed control system; Step 2: A frequency disturbance decoupling device is set up between the frequency measurement unit and the controller to decompose the grid frequency deviation signal into high-frequency disturbance components and low-frequency disturbance components, which are then input into the dedicated controllers for the guide vanes and blades to generate dynamic control signals. At the same time, the turbine's comprehensive characteristic curve is queried based on the current head and power setpoints to obtain the optimal opening of the guide vanes and blades and convert it into a reference control signal. The dynamic control signal is added to the corresponding reference control signal to form the final control command for the guide vanes and blades, which independently drives their respective actuators. Step 3: The particle swarm optimization algorithm is used to coordinate the optimization of the guide vane controller parameters, the blade controller parameters, and the cutoff frequency of the decoupling device filter. When the head or power setpoint changes significantly and continuously, the re-optimization is automatically triggered to update the control parameters, thereby realizing adaptive optimization control under all operating conditions.

[0010] Preferably, the mathematical model in step 1 includes a guide vane dedicated controller model, a blade dedicated controller model, a guide vane actuator model, a blade actuator model, a water intake system model, a turbine model, and a generator dynamic model.

[0011] Preferably, in step 2, the low-frequency disturbance component is input to the blade-specific controller to generate a blade dynamic control signal; the high-frequency disturbance component is input to the guide vane-specific controller to generate a guide vane dynamic control signal. Based on the current head and power setpoints, the turbine's comprehensive characteristic curve is queried to obtain the optimal guide vane opening and the optimal blade opening, which are then converted into guide vane reference control signals and blade reference control signals, respectively. The guide vane dynamic control signal is added to the guide vane reference control signal to obtain the guide vane final control command; the propeller blade dynamic control signal is added to the propeller blade reference control signal to obtain the propeller blade final control command. The guide vane actuator is driven by the final control command of the guide vane, and the blade actuator is driven by the final control command of the blade.

[0012] Preferably, in step 3, a particle swarm optimization algorithm is used to collaboratively optimize the control parameters of the guide vane controller, the blade controller, and the filter cutoff frequency of the frequency disturbance decoupling device. When the change in head is detected to be greater than a preset head threshold, or the change in power setpoint is greater than a preset power threshold and the duration is greater than a preset duration threshold, a re-optimization procedure is triggered to recalculate and update the control parameters of the guide vane controller, the blade controller, and the filter cutoff frequency of the frequency disturbance decoupling device.

[0013] Preferably, both the guide vane dedicated controller model and the blade dedicated controller model are PID controller models, and the function expression of the PID controller model is: ; in, , and These represent the proportional gain, integral gain, and derivative gain of the PID controller, respectively. The power grid frequency; For a given frequency; Output signals to the controller; The differential time constant; This is the permanent slip coefficient; Δ is the relative deviation; s is a complex variable that represents the frequency response when its real part is zero and its imaginary part is the angular frequency.

[0014] Preferably, in step 1, the guide vane actuator model and the propeller actuator model constitute the actuator model, and the actuator model construction process is as follows: The transfer function for the guide vane actuator model of an axial-flow propeller turbine is: ; The transfer function for the propeller actuator model is: ; in, For guide vane opening; This refers to the blade opening. For the control signal of the guide vane actuator; For the control signal of the blade actuator, and This is the inertial time constant of the actuator.

[0015] Preferably, in step 1, the water diversion system model construction process is as follows: The task of the water diversion system is to guide water from the upstream reservoir to the turbine, where the water flow impacts the blades and drives the turbine to rotate, ultimately flowing to the downstream reservoir; The water intake system includes water intake pipes, a pressure regulating chamber, pressure pipes, and tailrace pipes; Water intake system models are divided into two main categories: rigid water hammer models and elastic water hammer models; In small hydropower stations where the water flow fluctuations within the water diversion pipeline are small and the pipeline length is less than 800 meters, the inertia of the water flow is ignored, and the model is regarded as a rigid water hammer model. However, when the length of the water diversion pipeline exceeds 800 meters, the inertia of the water flow is large, and the elasticity between the water and the pipeline wall needs to be considered, so an elastic water hammer model is adopted. The transfer function for the rigid water hammer model of a pipeline is: ; in, Let be the water time constant; the transfer function of the pipeline elastic water hammer model is: ; in, For the time it takes for water to travel.

[0016] Preferably, in step 2, the turbine model construction process is as follows: A water turbine uses the kinetic and potential energy of water flow to drive an impeller to rotate, thereby driving a generator to produce electricity. Water flows from a higher to a lower elevation; this height difference provides potential energy, while the velocity of the water flow provides kinetic energy. The expression for an axial-flow propeller-type water turbine model is: ; in, For the turbine torque, Indicates flow rate. Indicates the guide vane opening. Indicates the blade rotation angle. Indicates water head. Indicates rotational speed. , , , , and This represents the transmission coefficient of the water turbine. This is the transmission coefficient of turbine torque to blade rotation angle. This is the transfer coefficient of turbine flow rate to blade rotation angle.

[0017] Preferably, in step 1, the generator dynamic model, i.e., the generator-load model construction process, is as follows: The transfer function is: ; in, Indicates the load torque. This represents the total constant that takes into account both the generator's inertial time constant and the load's inertial time constant. This represents the generator's regulation coefficient.

[0018] Preferably, the frequency disturbance decoupling device includes a first-order high-pass filter and a first-order low-pass filter, the first-order high-pass filter and the first-order low-pass filter are connected in parallel, and the cutoff frequency of the first-order high-pass filter is equal to the cutoff frequency of the first-order low-pass filter.

[0019] Preferably, the process of obtaining the optimal guide vane opening and the optimal blade opening in step 2 includes: based on the current head and power given values, and combined with the turbine's comprehensive characteristic curve and synergy curve, solving for the guide vane opening and blade rotation angle that satisfy the optimal efficiency condition, which are then used as the optimal guide vane opening and the optimal blade opening.

[0020] Preferably, the power characteristic curve established by the comprehensive characteristic curve can be expressed as: ; in, For guide vane opening; For the blade rotation angle; For water head; The synergy curve can be represented as: .

[0021] Preferably, the final control command for the guide vane is equal to the sum of the guide vane reference control signal and the guide vane dynamic control signal output by the dedicated guide vane controller; the final control command for the propeller blade is equal to the sum of the propeller blade reference control signal and the propeller blade dynamic control signal output by the dedicated propeller blade controller. The blade reference control signal and the guide vane reference control signal constitute the reference control signal. The specific steps for solving the reference control signal are as follows: Step 1) The speed control system receives the current power of the unit. and water head ; Step 2) Substitute Eliminate blade corner ,get . This represents the relationship between the turbine's output power and the guide vane opening at a given water head, assuming the blades are always maintained at their optimal opening. Based on power... Water head and Calculate the optimal guide vane opening under the current operating conditions; Step 3) Based on the head H, optimal guide vane opening, and Calculate the optimal blade rotation angle under the current operating conditions; Step 4) Calculate the reference control signal under the current operating condition based on the relationship between the guide vane opening, the blade rotation angle and the input signal of the actuator.

[0022] Preferably, the objective function of the particle swarm optimization algorithm is a weighted sum of the adjustment time, overshoot, and actuator action, and the constraint condition is that the guide vane opening and blade rotation angle are both within the safe range allowed by the speed control system.

[0023] Preferably, the triggering condition for the re-optimization procedure includes any one of the following two situations: The first scenario: The change in water head is greater than the preset water head threshold; The second scenario is when the change in the power setpoint is greater than the preset power threshold, and the duration of this change is greater than the preset duration threshold.

[0024] Preferably, the high-frequency disturbance component represents a small-amplitude high-frequency deviation caused by random load fluctuations in the power grid, which is quickly adjusted by the guide vane actuator; the low-frequency disturbance component represents a large-amplitude low-frequency deviation caused by systemic power deficit or surplus, which is slowly adjusted by the blade actuator.

[0025] A parallel optimization control system for guide vanes of an axial-flow propeller turbine based on frequency disturbance decoupling is disclosed. The system employs the aforementioned parallel optimization control method for guide vanes of an axial-flow propeller turbine based on frequency disturbance decoupling. The system includes: The system modeling module is used to establish mathematical models of each device in the speed regulation system of the axial flow propeller turbine. The parallel control command generation module is used to set up a frequency disturbance decoupling device between the frequency measurement unit and the controller. It decomposes the grid frequency deviation signal into high-frequency disturbance components and low-frequency disturbance components, and inputs them into the dedicated controllers of the guide vanes and blades to generate dynamic control signals. At the same time, it queries the turbine's comprehensive characteristic curve based on the current head and power setpoints to obtain the optimal opening of the guide vanes and blades and converts it into a reference control signal. The dynamic control signal is added to the corresponding reference control signal to form the final control command for the guide vanes and blades, which independently drives their respective actuators. The adaptive collaborative optimization module is used to collaboratively optimize the parameters of the guide vane controller, the blade controller, and the cutoff frequency of the decoupling device filter using the particle swarm optimization algorithm. It automatically triggers re-optimization and updates the control parameters when the head or power setpoint changes significantly and continuously, thus achieving adaptive optimization control under all operating conditions.

[0026] A computer device includes one or more processors, on which one or more executable programs are stored. When the one or more executable programs are executed by the one or more processors, they are used to implement the aforementioned parallel optimization control method for guide vanes of an axial-flow propeller turbine based on frequency disturbance decoupling.

[0027] A storage medium storing one or more executable programs, which, when executed, implement the aforementioned parallel optimization control method for guide vanes of an axial-flow propeller turbine based on frequency disturbance decoupling.

[0028] The present invention can achieve the following beneficial effects: 1. This invention effectively overcomes the response lag problem of traditional series-coordinated architectures through frequency decoupling and parallel control mechanisms. The guide vanes and blades can operate independently and simultaneously according to the disturbance characteristics, which significantly improves the system's response speed to frequency fluctuations, shortens the adjustment time, reduces overshoot, and helps enhance the frequency stability of the power grid.

[0029] 2. This invention combines optimal efficiency point lookup with dynamic adjustment, enabling the system to maintain high efficiency during steady-state operation and respond quickly during dynamic adjustment. Through a parameter adaptive optimization mechanism, it can effectively adapt to complex operating conditions such as varying heads and loads, thus broadening the system's stable operating range.

[0030] 3. This invention, through the scientific allocation of control tasks, allows the guide vanes to focus on high-frequency fine-tuning and the propeller blades to focus on low-frequency macro-tuning, reducing ineffective movements and wear of the actuators. This method can significantly extend the service life of the equipment and improve the operating economy of the unit.

[0031] 4. This invention optimizes and improves upon the existing system's main structure, employing a modular design for ease of engineering implementation. Its adaptive optimization characteristics enable it to adapt to axial-flow propeller turbines of different capacities and models, demonstrating promising prospects for widespread application. Attached Figure Description

[0032] The present invention will be further described below with reference to the accompanying drawings and embodiments: Figure 1 This is a schematic diagram of the overall framework of the improved axial-flow propeller turbine speed control system of the present invention.

[0033] Figure 2 A schematic diagram is generated for frequency decoupling and control commands.

[0034] Figure 3 The flowchart shows how to optimize the coefficient K using the particle swarm optimization algorithm.

[0035] Figure 4 The flowchart for triggering conditions in the re-optimization program. Detailed Implementation

[0036] Preferred solutions include Figures 1 to 4 As shown, a parallel optimization control method for guide vanes of an axial-flow propeller turbine based on frequency disturbance decoupling is presented. This invention constructs a frequency disturbance decoupling and parallel control architecture, combining the dynamic characteristics of the turbine with intelligent optimization algorithms to design an advanced speed control method. The scheme specifically includes the following key components: First, a frequency signal decoupling device is designed, which uses a parallel high-pass filter and a low-pass filter bank to decompose the power grid frequency deviation signal into high-frequency disturbance components and low-frequency disturbance components in real time. Second, establish independent dual PID control channels, design dedicated controllers for guide vanes and propellers respectively for the characteristics of high-frequency and low-frequency components, to achieve precise allocation of control tasks. Third, develop a reference signal generation mechanism based on the comprehensive characteristic curve of the turbine, and calculate the reference opening of the guide vane and blade corresponding to the optimal efficiency point in real time by querying the current head and power setpoint. Fourth, the particle swarm optimization algorithm is used to coordinate the optimization of controller parameters and filter cutoff frequency, and the re-optimization program is automatically triggered according to changes in operating conditions. Fifth, the dynamic control signal and the reference signal are intelligently synthesized to form the final control command, which drives the guide vane and the propeller actuator respectively.

[0037] The specific method is as follows: Step 1: Establish mathematical models for each part of the axial-flow propeller turbine speed control system, including the guide vane controller model, the propeller controller model, the guide vane actuator model, the propeller actuator model, the water intake system model, the turbine model, and the generator dynamic model.

[0038] Step 2: A frequency disturbance decoupling device based on frequency decoupling is added between the frequency measurement unit and the controller. This device receives the real-time detected power grid frequency deviation signal. Frequency deviation signal By using parallel high-pass and low-pass filters, the mixed frequency deviation signal is decoupled into high-frequency disturbance components and low-frequency power components according to the frequency domain decomposition law. The high-frequency component is then sent to a dedicated guide vane controller to generate a guide vane dynamic control signal, and the low-frequency component is sent to a dedicated blade controller to generate a blade dynamic control signal. Finally, these two dynamic control signals are combined with the signal generated by the current head. and power setpoint By synthesizing the optimal efficiency point reference control signal obtained by querying the comprehensive characteristic curve of the turbine, the final parallel control commands for the guide vanes and blades are obtained, and their respective actuators are driven independently.

[0039] Step 3: Utilize optimization algorithms to optimize the parameters of the guide vane controller, the blade controller, and the cutoff frequency of the filter bank in the frequency disturbance decoupling device. The system is optimized and monitors the changes in operating parameters in real time. When the change in head exceeds the preset threshold or the power setpoint changes abruptly, the re-optimization program is automatically triggered. The particle swarm optimization algorithm takes the current operating parameters as input and the weighted comprehensive performance index as the objective function to recalculate and dynamically update the optimal parameters of the guide vane controller, the blade controller, and the cutoff frequency of the filter bank of the frequency disturbance decoupling device, so that the frequency disturbance decoupling device and the controller can perform at their best.

[0040] Furthermore, in step 1, the controller model construction process is as follows: Both the guide vane controller and the blade controller of this invention employ PID controllers. The linear model transfer function of the PID controller can be expressed as: ; in, , and These represent the proportional gain, integral gain, and derivative gain of the PID controller, respectively. For the power grid frequency, For a given frequency, Output signal to the controller The differential time constant is This is the permanent slip coefficient. Δ is the relative deviation, and s is a complex variable that represents the frequency response when its real part is zero and its imaginary part is the angular frequency.

[0041] Furthermore, in step 1, the process of constructing the actuator model is as follows: The actuator has two main functions: first, to convert received electrical signals into mechanical signals, thereby generating corresponding mechanical displacement; and second, to amplify the electrical signals, making them more effectively recognized by the system. Common actuators are simplified to first-order systems. The linear model transfer function of the guide vane actuator for an axial-flow propeller turbine is as follows: ; The linear model transfer function of the propeller actuator is: ; in, For guide vane opening; This refers to the blade opening. For the control signal of the guide vane actuator; For the control signal of the blade actuator, and This is the inertial time constant of the actuator. In practice, actuators typically contain dead zones and saturated nonlinear elements, usually caused by mechanical dead zones, guide vane opening limitations, and speed limitations.

[0042] Furthermore, in step 1, the water diversion system model construction process is as follows: The task of the water diversion system is to guide water from the upstream reservoir to the turbine, where the water flow impacts the blades and drives the turbine to rotate, ultimately flowing to the downstream reservoir. The water diversion system includes the diversion pipeline, surge tank, pressure pipeline, and tailrace. Water diversion system models are divided into two main categories: rigid water hammer models and elastic water hammer models. In small hydropower stations where the water flow fluctuation within the diversion pipeline is small and the pipeline length is less than 800 meters, the water flow inertia can be ignored, and the model can be considered a rigid water hammer model. However, when the diversion pipeline length exceeds 800 meters, the water flow inertia is greater, and the elasticity between the water and the pipeline wall needs to be considered, thus an elastic water hammer model is adopted. The transfer function for the rigid water hammer model of a pipeline is: ; in, Let be the water time constant. The transfer function of the pipeline elastic water hammer model is: ; in, For the time it takes for water to travel.

[0043] Furthermore, in step 1, the turbine model construction process is as follows: A water turbine uses the kinetic and potential energy of water flow to drive an impeller to rotate, thereby driving a generator to produce electricity. Water flows from a higher to a lower elevation; this height difference provides potential energy, while the velocity of the water flow provides kinetic energy. The expression for an axial-flow propeller-type water turbine model is: ; in, For the turbine torque, Indicates flow rate. Indicates the guide vane opening. Indicates the blade rotation angle. Indicates water head. Indicates rotational speed. , , , , and This represents the transmission coefficient of the water turbine. This is the transmission coefficient of turbine torque to blade rotation angle. This is the transfer coefficient of turbine flow rate to blade rotation angle.

[0044] Furthermore, in step 1, the generator-load model construction process is as follows: The generator is the core component that converts mechanical energy into electrical energy, playing a crucial role in electricity production. Its design and operation must consider the performance of the turbine, the requirements of the power grid, and the overall reliability of the system. The generator-load model is a first-order model, with the transfer function as follows: ; in, Indicates the load torque. This represents the total constant that takes into account both the generator's inertial time constant and the load's inertial time constant. This represents the generator's regulation coefficient.

[0045] Furthermore, the internal structure of the frequency disturbance decoupling device in step 2 is a set of parallel low-pass and high-pass filters.

[0046] The low-pass filter is used to extract low-frequency disturbance components. This filter allows frequencies below the cutoff frequency. The signal components pass through with little or no attenuation, while high-frequency components are significantly attenuated. Its output signal... This mainly includes low-frequency, large-amplitude frequency deviations caused by slow load changes, systemic power deficits or surpluses.

[0047] The high-pass filter is used to extract high-frequency disturbance components. This filter allows frequencies above the cutoff frequency. The high-frequency components pass through, while the low-frequency components are significantly attenuated. Its output signal... This mainly includes high-frequency, low-amplitude frequency jitter caused by random load fluctuations and instantaneous power grid disturbances.

[0048] Considering that the effective response frequency of guide vanes and propellers is usually low due to their mechanical inertia, there is no need for high-order filters to separate the extremely high frequency components that the actuators themselves cannot effectively respond to. Therefore, both the low-pass and high-pass filters are first-order filters.

[0049] The transfer function of a low-pass filter can be expressed as: ; in, The time constant of the low-pass filter, and , This is the cutoff frequency of the pass filter.

[0050] The transfer function of a high-pass filter can be expressed as: ; in, The time constant of the high-pass filter, and , This is the cutoff frequency of the high-pass filter.

[0051] when ,Right now At that time, it will be in the frequency band Signal overlap occurs between the two channels, causing the same frequency component to enter both channels simultaneously, leading to control conflicts.

[0052] when ,Right now At that time, it will be in the frequency band This creates a signal dead zone, where disturbances in that frequency band cannot be effectively handled by any channel.

[0053] In order to convert the frequency deviation signal To decompose the frequencies into high-frequency and low-frequency components without omission or overlap, the following conditions must be met: ; In other words, the low-pass filter and the high-pass filter need to satisfy: ; From the above formula, we can deduce that when At that time, it can satisfy the frequency deviation signal It is decomposed into high-frequency and low-frequency components without omission or overlap. Therefore, Take the cutoff frequency as .

[0054] Furthermore, the process of setting the reference control signal in step 2 is as follows: Based on the current power and head of the unit, the optimal opening of the guide vanes and propellers under the current operating conditions is obtained by combining the turbine's comprehensive characteristic curve and the coordination curve. This optimal opening is then used as the reference control signal. The power characteristic curve established based on the comprehensive characteristic curve can be expressed as follows: ; in, For guide vane opening; This refers to the blade rotation angle.

[0055] The synergy curve can be represented as: ; The specific steps for solving the reference control signal are as follows: Step 1) The speed control system receives the current power of the unit. and water head ; Step 2) Substitute Eliminate blade corner ,get . This represents the relationship between the turbine's output power and the guide vane opening at a given water head, assuming the blades are always maintained at their optimal opening. Based on power... Water head and Calculate the optimal guide vane opening under the current operating conditions; Step 3) Based on the head H, optimal guide vane opening, and Calculate the optimal blade rotation angle under the current operating conditions; Step 4) Calculate the reference control signal under the current operating condition based on the relationship between the guide vane opening, the blade rotation angle and the input signal of the actuator.

[0056] For axial-flow propeller turbines employing modern electro-hydraulic servo systems, within their normal control stroke range, a linear relationship is typically used to describe the relationship between the control signal and the guide vane opening / blade angle.

[0057] The relationship between the guide vane control signal and the guide vane opening can be expressed as: ; in, This is the proportional coefficient for guide vane control; This is the guide vane control signal; is the bias constant, representing the guide vane opening when the control signal is 0.

[0058] The relationship between the blade control signal and the blade rotation angle can be expressed as: ; in, This is the proportional coefficient for blade control; For blade control signals; is the bias constant, representing the blade rotation angle when the control signal is 0.

[0059] Based on the guide vane opening and blade rotation angle obtained in step 3, the reference control signals for guide vane and blade control under the current operating condition can be obtained as follows: ; ; Furthermore, the optimization algorithm in step 3 optimizes the parameters of the guide vane dedicated controller. Parameters of the dedicated blade controller and the cutoff frequency of the filter bank of the frequency disturbance decoupling device The specific optimization process is as follows: Particle swarm optimization (PSO) is a swarm intelligence stochastic optimization algorithm that does not require gradient information. By simulating the behavior of a social group, it performs a global exploration of the solution space, exhibiting strong global optimization capabilities and effectively avoiding getting trapped in local optima, thus reliably approximating the global optimum. Furthermore, this algorithm has a simple structure, requires few parameters to be adjusted, converges quickly, and has high computational efficiency, making it very suitable for solving parameter optimization problems in complex engineering projects. Its ease of handling constraints also facilitates... The search range is limited to the engineering safety range. Therefore, this invention uses a particle swarm optimization algorithm for optimization.

[0060] The velocity and position update formulas for the i-th particle in the particle swarm optimization algorithm can be expressed as: ; in, For the number of iterations, The particle dimension number. For particles In the In the nth iteration dimensional velocity vector For particles In the In the nth iteration dimensional position vector, For particles In the In the nth iteration The historical optimal position of dimension, that is, the position in the dimensional... After the nth iteration, the th The optimal solution obtained by searching for individual particles (particles) For the group in the first In the nth iteration The historical optimal position of dimension, that is, the position in the dimensional... The optimal solution in the entire particle swarm after the [number] iterations. For inertial weights, For individual learning factors, As a group learning factor, , For interval Random numbers within.

[0061] exist In the optimization, the objective function should be the system settling time. Overshoot and the amount of action of the executing agency Weighted composite value The constraint is that the opening of the guide vanes and blades must be within the acceptable range of the speed control system, and the opening of the guide vanes and blades must not exceed the safety value specified by the system.

[0062] The objective function can be expressed as: ; Adjustment time Overshoot and the amount of action of the executing agency The weights, and satisfying ; for The scope of normalization; for The scope of normalization; for The scope of normalization.

[0063] When the operating conditions of an axial-flow propeller turbine change, the original The rationality of signal allocation at this time can no longer be guaranteed, and adjustments need to be made according to changes in operating conditions.

[0064] This invention uses a preset threshold for water head changes. Or a preset threshold for power variation To determine whether to trigger the re-optimization process to recalculate Specifically, this can manifest in the following three ways: (1) The change in water head exceeds the preset threshold. When this happens, the re-optimization process will be automatically triggered; (2) Power change exceeds preset threshold And the duration exceeds the preset duration threshold. If it is further confirmed that this is a "continuous" change in operating conditions rather than a momentary disturbance, the re-optimization program will be automatically triggered to prevent false triggering caused by scheduling command jitter or misoperation.

[0065] (3) Otherwise, the re-optimization program will not respond.

[0066] In the parameter tuning process of the Particle Swarm Optimization (PSO) algorithm described in this invention, the setting of its key parameters follows the principles described below: Particle swarm size N Choosing a size of 40 ensures both population diversity and global search capability while maintaining computational efficiency. A size that is too small is prone to getting trapped in local optima, while a size that is too large significantly increases the computational burden. (Particle dimension) D The value is 7, corresponding to the seven optimization variables of this invention: guide vane proportional coefficient. Integral coefficient Differential coefficients Blade ratio coefficient Integral coefficient Differential coefficients and filter cutoff frequency Number of iterations k A value of 100 is recommended to strike a balance between solution quality and computational time cost. Individual learning factor. c1 Social learning factors c2 A value of 1.8 is recommended to balance the particles' self-awareness and group cooperation capabilities. A value that is too low will slow down the convergence speed, while a value that is too high will easily lead to search oscillations. Inertia Weight ω An adaptive strategy is adopted, linearly decreasing from 0.9 to 0.4, which enables the algorithm to have strong global exploration capabilities in the early stages of iteration, while focusing on fine-grained local search in the later stages, thereby comprehensively improving optimization efficiency and accuracy. The particle flight speed limit is set according to the actual physical meaning of each parameter to prevent it from exceeding the effective search space and missing the optimal solution.

[0067] Furthermore, the final parallel control command in step 2 is synthesized by the reference control signal and the PID controller, as shown in the following formula: The final control signal for the guide vane can be expressed as: ; in, This is the guide vane control signal output by the dedicated guide vane controller.

[0068] The final control signal for the propeller blades can be expressed as: ; in, This is the blade control signal output by the dedicated blade controller.

[0069] Furthermore, the parallel optimization control method for guide vanes of axial-flow propeller turbines based on frequency disturbance decoupling provided by this invention has the following significant advantages compared to traditional coordinated control: (1) Decoupling control: By using a frequency signal decoupling distribution device, wideband frequency disturbances are decomposed into high-frequency and low-frequency components, which are then processed by the guide vanes and propeller blades with the best matching response characteristics, thus avoiding motion coupling at the source and improving regulation efficiency. (2) Adaptive optimization: The intelligent optimization algorithm is used to tune the parameters of the controller and filter, and can automatically trigger re-optimization according to the changes in operating conditions, so that the control system can maintain the best performance under different heads and power, and has strong robustness.

[0070] (3) Stable and efficient: The combination of dynamic control signals and reference control signals based on the optimal efficiency point not only ensures the unit's rapid response to frequency disturbances, but also ensures that it always maintains the high efficiency zone during steady-state operation, taking into account both the stability of the power grid frequency and the economic operation of the power plant.

[0071] The above embodiments are merely preferred technical solutions of the present invention and should not be considered as limitations on the present invention. 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 parallel optimization control method for guide vanes of an axial-flow propeller turbine based on frequency disturbance decoupling, characterized in that, Includes the following steps: Step 1: Establish mathematical models for each device in the axial-flow propeller turbine speed control system; Step 2: A frequency disturbance decoupling device is set up between the frequency measurement unit and the controller to decompose the grid frequency deviation signal into high-frequency disturbance components and low-frequency disturbance components, which are then input into the dedicated controllers for the guide vanes and blades to generate dynamic control signals. At the same time, the turbine's comprehensive characteristic curve is queried based on the current head and power setpoints to obtain the optimal opening of the guide vanes and blades and convert it into a reference control signal. The dynamic control signal is added to the corresponding reference control signal to form the final control command for the guide vanes and blades, which independently drives their respective actuators. Step 3: The particle swarm optimization algorithm is used to coordinate the optimization of the guide vane controller parameters, the blade controller parameters, and the cutoff frequency of the decoupling device filter. When the head or power setpoint changes significantly and continuously, the re-optimization is automatically triggered to update the control parameters, thereby realizing adaptive optimization control under all operating conditions.

2. The parallel optimization control method for guide vanes of an axial-flow propeller turbine based on frequency disturbance decoupling as described in claim 1, characterized in that, The mathematical models in step 1 include the guide vane dedicated controller model, the blade dedicated controller model, the guide vane actuator model, the blade actuator model, the water diversion system model, the turbine model, and the generator dynamic model.

3. The parallel optimization control method for guide vanes of an axial-flow propeller turbine based on frequency disturbance decoupling as described in claim 1, characterized in that, In step 2, the low-frequency disturbance component is input into the blade-specific controller to generate a blade dynamic control signal; the high-frequency disturbance component is input into the guide vane-specific controller to generate a guide vane dynamic control signal. Based on the current head and power setpoints, the turbine's comprehensive characteristic curve is queried to obtain the optimal guide vane opening and the optimal blade opening, which are then converted into guide vane reference control signals and blade reference control signals, respectively. The guide vane dynamic control signal is added to the guide vane reference control signal to obtain the guide vane final control command; the propeller blade dynamic control signal is added to the propeller blade reference control signal to obtain the propeller blade final control command. The guide vane actuator is driven by the final control command of the guide vane, and the blade actuator is driven by the final control command of the blade.

4. The parallel optimization control method for guide vanes of an axial-flow propeller turbine based on frequency disturbance decoupling as described in claim 1, characterized in that, In step 3, the particle swarm optimization algorithm is used to collaboratively optimize the control parameters of the guide vane controller, the blade controller, and the filter cutoff frequency of the frequency disturbance decoupling device. When the change in head is detected to be greater than the preset head threshold, or the change in power setpoint is greater than the preset power threshold and the duration is greater than the preset duration threshold, the re-optimization program is triggered to recalculate and update the control parameters of the guide vane controller, the blade controller, and the filter cutoff frequency of the frequency disturbance decoupling device.

5. The parallel optimization control method for guide vanes of an axial-flow propeller turbine based on frequency disturbance decoupling as described in claim 2, characterized in that, Both the guide vane dedicated controller model and the blade dedicated controller model are PID controller models, and the function expression of the PID controller model is: ； in, , and These represent the proportional gain, integral gain, and derivative gain of the PID controller, respectively. The power grid frequency; For a given frequency; Output signals to the controller; The differential time constant; This is the permanent slip coefficient; Δ is the relative deviation; s is a complex variable that represents the frequency response when its real part is zero and its imaginary part is the angular frequency.

6. The parallel optimization control method for guide vanes of an axial-flow propeller turbine based on frequency disturbance decoupling as described in claim 2, characterized in that, In step 1, the guide vane actuator model and the propeller actuator model constitute the actuator model. The actuator model construction process is as follows: The transfer function for the guide vane actuator model of an axial-flow propeller turbine is: ； The transfer function for the propeller actuator model is: ； in, For guide vane opening; This refers to the blade opening. For the control signal of the guide vane actuator; For the control signal of the blade actuator, and This is the inertial time constant of the actuator.

7. The parallel optimization control method for guide vanes of an axial-flow propeller turbine based on frequency disturbance decoupling as described in claim 2, characterized in that, In step 1, the water diversion system model construction process is as follows: The task of the water diversion system is to guide water from the upstream reservoir to the turbine, where the water flow impacts the blades and drives the turbine to rotate, ultimately flowing to the downstream reservoir; The water intake system includes water intake pipes, a pressure regulating chamber, pressure pipes, and tailrace pipes; Water intake system models are divided into two main categories: rigid water hammer models and elastic water hammer models; In small hydropower stations where the water flow fluctuations within the water diversion pipeline are small and the pipeline length is less than 800 meters, the inertia of the water flow is ignored, and the model is regarded as a rigid water hammer model. However, when the length of the water diversion pipeline exceeds 800 meters, the inertia of the water flow is large, and the elasticity between the water and the pipeline wall needs to be considered, so an elastic water hammer model is adopted. The transfer function for the rigid water hammer model of a pipeline is: ； in, Let be the water time constant; the transfer function of the pipeline elastic water hammer model is: ； in, For the time it takes for water to travel.

8. The parallel optimization control method for guide vanes of an axial-flow propeller turbine based on frequency disturbance decoupling as described in claim 2, characterized in that, In step 2, the turbine model construction process is as follows: A water turbine uses the kinetic and potential energy of water flow to drive an impeller to rotate, thereby driving a generator to produce electricity. Water flows from a higher to a lower elevation; this height difference provides potential energy, while the velocity of the water flow provides kinetic energy. The expression for an axial-flow propeller-type water turbine model is: ； in, For the turbine torque, Indicates flow rate. Indicates the guide vane opening. Indicates the blade rotation angle. Indicates water head. Indicates rotational speed. , , , , and This represents the transmission coefficient of the water turbine. This is the transmission coefficient of turbine torque to blade rotation angle. This is the transfer coefficient of turbine flow rate to blade rotation angle.

9. The parallel optimization control method for guide vanes of an axial-flow propeller turbine based on frequency disturbance decoupling as described in claim 1, characterized in that: In step 1, the generator dynamic model, i.e., the generator-load model construction process, is as follows: The transfer function is: ； in, Indicates the load torque. This represents the total constant that takes into account both the generator's inertial time constant and the load's inertial time constant. This represents the generator's regulation coefficient.

10. The parallel optimization control method for guide vanes of an axial-flow propeller turbine based on frequency disturbance decoupling as described in claim 1, characterized in that, The frequency disturbance decoupling device includes a first-order high-pass filter and a first-order low-pass filter, the first-order high-pass filter and the first-order low-pass filter are connected in parallel, and the cutoff frequency of the first-order high-pass filter is equal to the cutoff frequency of the first-order low-pass filter.

11. The parallel optimization control method for guide vanes of an axial-flow propeller turbine based on frequency disturbance decoupling as described in claim 3, characterized in that, The process of obtaining the optimal guide vane opening and the optimal blade opening in step 2 includes: based on the current head and power given values, combined with the turbine's comprehensive characteristic curve and synergy curve, solving for the guide vane opening and blade angle that satisfy the optimal efficiency condition, which are then used as the optimal guide vane opening and the optimal blade opening.

12. The parallel optimization control method for guide vanes of an axial-flow propeller turbine based on frequency disturbance decoupling as described in claim 11, characterized in that, The power characteristic curve established by the comprehensive characteristic curve can be expressed as: ； in, For guide vane opening; For the blade rotation angle; For water head; The synergy curve can be represented as: 。 13. The parallel optimization control method for guide vanes of an axial-flow propeller turbine based on frequency disturbance decoupling as described in claim 12, characterized in that, The final control command for the guide vane is equal to the sum of the guide vane reference control signal and the guide vane dynamic control signal output by the guide vane dedicated controller; the final control command for the blade is equal to the sum of the blade reference control signal and the blade dynamic control signal output by the blade dedicated controller. The blade reference control signal and the guide vane reference control signal constitute the reference control signal. The specific steps for solving the reference control signal are as follows: Step 1) The speed control system receives the current power of the unit. and water head ; Step 2) Substitute Eliminate blade corner ,get ; This represents the relationship between the turbine output power and the guide vane opening at a given water head, assuming the blades are always maintained at their optimal opening. Based on the power... Water head and Calculate the optimal guide vane opening under the current operating conditions; Step 3) Based on the head H, optimal guide vane opening, and Calculate the optimal blade rotation angle under the current operating conditions; Step 4) Calculate the reference control signal under the current operating condition based on the relationship between the guide vane opening, the blade rotation angle and the input signal of the actuator.

14. The parallel optimization control method for guide vanes of an axial-flow propeller turbine based on frequency disturbance decoupling as described in claim 1, characterized in that, The objective function of the particle swarm optimization algorithm is a weighted sum of the adjustment time, overshoot, and actuator action, with the constraint that the guide vane opening and blade rotation angle are both within the safe range allowed by the speed control system.

15. The parallel optimization control method for guide vanes of an axial-flow propeller turbine based on frequency disturbance decoupling as described in claim 4, characterized in that, The triggering conditions for the re-optimization process include either of the following two scenarios: The first scenario: The change in water head is greater than the preset water head threshold; The second scenario is when the change in the power setpoint is greater than the preset power threshold, and the duration of this change is greater than the preset duration threshold.

16. The parallel optimization control method for guide vanes of an axial-flow propeller turbine based on frequency disturbance decoupling as described in claim 1, characterized in that, The high-frequency disturbance component represents a small-amplitude high-frequency deviation caused by random load fluctuations in the power grid, which is quickly adjusted by the guide vane actuator; the low-frequency disturbance component represents a large-amplitude low-frequency deviation caused by systemic power deficit or surplus, which is slowly adjusted by the blade actuator.

17. A parallel optimization control system for guide vanes of an axial-flow propeller turbine based on frequency disturbance decoupling, characterized in that, The system employs a parallel optimization control method for guide vanes of an axial-flow propeller turbine based on frequency disturbance decoupling, as described in any one of claims 1-16. The system includes: The system modeling module is used to establish mathematical models of each device in the speed regulation system of the axial flow propeller turbine. The parallel control command generation module is used to set up a frequency disturbance decoupling device between the frequency measurement unit and the controller. It decomposes the grid frequency deviation signal into high-frequency disturbance components and low-frequency disturbance components, and inputs them into the dedicated controllers of the guide vanes and blades to generate dynamic control signals. At the same time, it queries the turbine's comprehensive characteristic curve based on the current head and power setpoints to obtain the optimal opening of the guide vanes and blades and converts it into a reference control signal. The dynamic control signal is added to the corresponding reference control signal to form the final control command for the guide vanes and blades, which independently drives their respective actuators. The adaptive collaborative optimization module is used to collaboratively optimize the parameters of the guide vane controller, the blade controller, and the cutoff frequency of the decoupling device filter using the particle swarm optimization algorithm. It automatically triggers re-optimization and updates the control parameters when the head or power setpoint changes significantly and continuously, thus achieving adaptive optimization control under all operating conditions.

18. A computer device, characterized in that, The device includes one or more processors, on which one or more executable programs are stored. When the one or more executable programs are executed by the one or more processors, they are used to implement the parallel optimization control method for guide vanes of an axial-flow propeller turbine based on frequency disturbance decoupling, as described in any one of claims 1-16.

19. A storage medium, characterized in that, It stores one or more executable programs, which, when executed, are used to implement the parallel optimization control method for guide vanes of axial-flow propeller turbines based on frequency disturbance decoupling, as described in any one of claims 1-16.

Images