Hydraulic servo system for water turbine and control device of hydraulic servo system

Abstract

The invention relates to the technical field of hydraulic control, in particular to a hydraulic servo system for a water turbine and a control device.The hydraulic servo system for the water turbine comprises a hydraulic executing mechanism, an adjustable component connected to the water turbine and used for adjusting the angle of a guide vane; the control valve group is used for controlling the hydraulic oil to flow to the hydraulic actuating mechanism; the Internet of Things sensor module monitors operation parameters of the hydraulic system in real time and performs data transmission; the controller is used for receiving the real-time data transmitted by the Internet of Things sensor module, generating a control signal by adopting a self-adaptive neural network model predictive control algorithm and controlling the control valve group so as to reduce hydraulic impact and energy loss and improve the stability and accuracy of the system; according to the hydraulic servo system for the water turbine, by introducing the Internet of Things sensor module, the self-adaptive neural network model predictive control algorithm and the energy recovery module, accurate monitoring and intelligent adjustment of real-time operation parameters of the system and efficient utilization of energy are achieved.

Description

Technical Field

[0001] The invention belongs to the technical field of hydraulic control, and in particular relates to a hydraulic servo system for a water turbine and a control device thereof. Background Art

[0002] As a key component of the turbine speed governor, the performance of the hydraulic servo system directly affects the operating efficiency and stability of the hydropower generation system. Traditional hydraulic servo systems mainly adjust the guide vane angle through mechanical or simple electrical control methods to adjust the water flow. However, with the expansion of the scale of hydropower stations and the improvement of modernization needs, traditional hydraulic servo systems have gradually exposed many problems. For example, the dynamic response speed of the system is insufficient, and hydraulic shock and energy loss are prone to occur during the adjustment process. In addition, due to the lack of intelligent real-time monitoring and control mechanisms, the system is difficult to adapt quickly when facing complex operating conditions, resulting in reduced adjustment accuracy and increased equipment failure rate. In recent years, the rapid development of Internet of Things technology and intelligent control algorithms has provided a new direction for the performance optimization of hydraulic servo systems. By introducing real-time monitoring and intelligent predictive control, the response speed and stability of the hydraulic servo system can be significantly improved, while achieving efficient energy utilization during operation.

[0003] Although some improvement schemes have attempted to improve the performance of hydraulic servo systems by adding sensors or optimizing control strategies, there are still some shortcomings. For example, existing schemes mostly use control models with fixed parameters, which makes it difficult to achieve adaptive adjustment under complex working conditions, resulting in insufficient control accuracy and system stability. In addition, most hydraulic servo systems lack energy recovery devices, and the excess energy caused by hydraulic shock is wasted, resulting in low energy utilization efficiency. In terms of energy management, traditional energy recovery methods are usually simple in design and difficult to operate effectively under multiple working conditions. For data analysis of the system's operating status, simple statistical methods based on historical data are mostly used, which lack real-time and predictive capabilities, limiting the intelligence level of the system. The above problems have seriously restricted the application and promotion of hydraulic servo systems in modern hydropower generation.

[0004] Based on this, the existing turbine hydraulic servo system has significant problems in control accuracy, energy recovery and intelligence level. Summary of the invention

[0005] To achieve the above object, the present invention provides the following technical solution: a hydraulic servo system for a water turbine, comprising:

[0006] A hydraulic actuator connected to the adjustable part of the turbine for adjusting the angle of the guide vanes; a control valve group for controlling the flow of hydraulic oil to the hydraulic actuator;

[0007] IoT sensor modules, including pressure sensors, flow meters, temperature sensors, and vibration sensors, monitor the operating parameters of the hydraulic system in real time and transmit data;

[0008] A controller that receives real-time data transmitted by the IoT sensor module and uses an adaptive neural network model predictive control algorithm to generate control signals and control the control valve group to reduce hydraulic shock and energy loss and improve system stability and accuracy;

[0009] The energy recovery module is arranged in the hydraulic circuit and is used to recover the excess hydraulic energy when pressure shock occurs in the hydraulic system and store it for later use.

[0010] As an optimal technical solution for hydraulic servo system for water turbine,

[0011] The working process of the adaptive neural network model predictive control algorithm includes:

[0012] Step 1: Initialize the system, set initial parameters, including target guide vane angle, initial control input, prediction time domain, control cycle, start the hydraulic pump to establish working pressure, and load the adaptive neural network model predictive control algorithm to the controller;

[0013] Step 2: Collect the pressure, flow, temperature, vibration frequency and amplitude data of the hydraulic system in real time through the IoT sensor module;

[0014] Step 3: Filter and calibrate the collected data, extract characteristic parameters, and use the improved Kalman filter to estimate the system state vector based on the measured pressure and flow data;

[0015] Step 4: Update system parameters online based on the adaptive neural network, where the input of the neural network is the current state and control input, and the output is the predicted state at the next moment;

[0016] Step 5: Use the adaptive neural network after updating the system parameters to construct the objective function of the control model, solve the optimization problem under the constraints of the control input constraint and the state constraint, and obtain the optimal control input;

[0017] Step 6: Send the optimal control input to the control valve group, adjust the valve opening, control the hydraulic actuator, and adjust the angle of the guide vane;

[0018] Step 7: Use the energy recovery device to store the excess energy in the hydraulic system in the supercapacitor through the hydraulic-electric energy conversion device. When the system needs it, the supercapacitor releases the stored electrical energy through the power converter to assist the hydraulic pump in supplying energy and optimize energy utilization.

[0019] As an optimal technical solution for hydraulic servo system for water turbine,

[0020] The controller loads the adaptive neural network model predictive control algorithm, including:

[0021] Loading algorithm: After the controller is started, it loads the adaptive neural network model predictive control algorithm and initializes the network parameters and model parameters.

[0022] Initialize the neural network parameters: The weight matrix W and the bias vector b are randomly initialized to satisfy the Gaussian distribution, and the learning rate η is set.

[0023] As an optimal technical solution for hydraulic servo system for water turbine,

[0024] The IoT sensor module collects data from the hydraulic system in real time, including:

[0025] Using pressure sensors to measure: Install pressure sensors at both ends of the hydraulic cylinder to measure the pressure P in real time. 1 (t) and P 2 (t);

[0026] Install a flow meter at the outlet of the hydraulic pump to measure the hydraulic oil flow Q(t) in real time;

[0027] Install temperature sensors near the oil tank and hydraulic cylinder to measure the hydraulic oil temperature T(t);

[0028] Vibration sensors are installed on the hydraulic cylinder and guide vane mechanism to measure the vibration frequency f(t) and amplitude A(t).

[0029] As an optimal technical solution for hydraulic servo system for water turbine,

[0030] Filtering of the collected data includes:

[0031] A first-order low-pass filter is applied to the acquired raw data to remove high-frequency noise.

[0032] The filtering formula is as follows:

[0033] y(t)=α·x(t)+(1-α)·y(t-1)

[0034] Among them, x(t) is the original data of the current sampling, y(t) is the filtered data, t is the time, and α is the filter coefficient.

[0035] As an optimal technical solution for hydraulic servo system for water turbine,

[0036] Calibration of the filtered data includes,

[0037] The calibration formula is as follows:

[0038] x calibrated(t) = k·y(t) + b

[0039] Where: k is the sensor gain and b is the sensor bias.

[0040] As an optimal technical solution for hydraulic servo system for water turbine,

[0041] Extraction of feature parameters includes:

[0042] Calculate the pressure difference:

[0043] ΔP(t)=P 1 (t)-P 2 (t)

[0044] Calculate the average flow rate:

[0045] Calculate the average flow at the nearest m sampling points:

[0046]

[0047] As an optimal technical solution for hydraulic servo system for water turbine,

[0048] Perform state estimation on the system, including,

[0049] Establish the system state vector, which is defined as

[0050]

[0051] Among them, x p (t) is the piston displacement, is the piston speed.

[0052] As an optimal technical solution for hydraulic servo system for water turbine,

[0053] The state space model is established using the improved Kalman filter:

[0054]

[0055] Where: A, B, C are system matrices, w(t) is process noise, assumed to be zero-mean Gaussian white noise, and the covariance matrix is ​​Q w , v(t) is the measurement noise, which is assumed to be zero-mean Gaussian white noise, and the covariance matrix is ​​R v , y(t) is the measured output.

[0056] The present invention also provides a control device based on the hydraulic servo system for a water turbine, comprising a controller body, an expansion module and a connecting piece, wherein a first card slot is provided at one end of the controller body, a first card buckle is provided in the first card slot, and the first card buckle is connected to a first pressure rod;

[0057] A connecting piece is provided on the side of the controller body, and the connecting piece includes a limiting side strip and a limiting groove between the limiting side strips. A buckle groove is provided on the connecting piece, a buckle edge is provided on one side of the expansion module, and the buckle edge is embedded in the buckle groove. A receiving groove is provided on one side of the expansion module, and a conductive strip is provided in the receiving groove. The limiting side strip is in the receiving groove, and the conductive strip is embedded in the limiting groove. A conductive sheet is provided in the limiting groove.

[0058] Beneficial effects of the present invention: A hydraulic servo system for a water turbine invented by us realizes accurate monitoring of real-time operating parameters of the system, intelligent regulation and efficient utilization of energy by introducing an Internet of Things sensor module, an adaptive neural network model predictive control algorithm and an energy recovery module; the control device of the present invention can facilitate the addition of expansion modules. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor. Among them:

[0060] Figure 1 It is a connection block diagram of the system of the present invention;

[0061] Figure 2 It is a schematic diagram of the overall structure of the control device in the device of the present invention;

[0062] Figure 3 It is a structural schematic diagram of the controller body in the device of the present invention;

[0063] Figure 4 It is a schematic diagram of the installation structure of the expansion module and the connecting piece in the device of the present invention;

[0064] Figure 5 It is a structural schematic diagram of the expansion module in the device of the present invention;

[0065] Figure 6 A schematic diagram showing a comparison of guide vane angle adjustment accuracy in the system of the present invention;

[0066] Figure 7 A schematic diagram showing the comparison of hydraulic pressure stability in the system of the present invention;

[0067] Figure 8 A schematic diagram showing a comparison of energy recovery efficiency in the system of the present invention;

[0068] Fig. 9 Schematic diagram for comparing dynamic responses of control valves in the system of the present invention

[0069] In the figure: 102, first pressure rod; 100, controller body; 201, buckle edge; 303, buckle groove; 301, limit edge strip; 202, accommodating groove; 203, conductive strip; 302, limit groove; 304, conductive sheet; 204, second card slot; 206, second pressure rod; 205, wiring port; 101, first card slot; 400, board frame; 200, expansion module; 207, expansion terminal; 305, third card slot; 306, inclined surface; 300, connecting piece. DETAILED DESCRIPTION

[0070] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0071] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0072] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0073] Secondly, the present invention is described in detail with reference to the schematic diagram. When describing the embodiments of the present invention in detail, for the sake of convenience, the cross-sectional diagrams showing the device structure will not be partially enlarged according to the general scale, and the schematic diagrams are only examples, which should not limit the scope of protection of the present invention. In addition, in actual production, the three-dimensional dimensions of length, width and depth should be included.

[0074] Example 1

[0075] Reference Figure 1 to Figure 5 , this embodiment provides a hydraulic servo system for a water turbine, comprising,

[0076] A hydraulic actuator connected to the adjustable part of the turbine for adjusting the angle of the guide vanes; a control valve group for controlling the flow of hydraulic oil to the hydraulic actuator;

[0077] IoT sensor modules, including pressure sensors, flow meters, temperature sensors, and vibration sensors, monitor the operating parameters of the hydraulic system in real time and transmit data;

[0078] A controller that receives real-time data transmitted by the IoT sensor module and uses an adaptive neural network model predictive control algorithm to generate control signals and control the control valve group to reduce hydraulic shock and energy loss and improve system stability and accuracy;

[0079] Energy recovery module, which is configured in the hydraulic circuit, is used to recover excess hydraulic energy when pressure shock occurs in the hydraulic system and store it for later use

[0080] Step 1: System Initialization

[0081] Specifically, the initial parameters are set, including the target guide vane angle θ ref , initial control input u(0), prediction time domain N, control period T s , start the hydraulic pump to establish working pressure, and the controller loads the adaptive neural network algorithm.

[0082] According to the operation requirements of the turbine, set the target guide vane angle θ ref , its value range is generally from 0° to 90°.

[0083] The initial control input is set to zero, i.e. u(0), indicating that the valve port is initially closed.

[0084] According to the dynamic characteristics of the system, the prediction time domain N is set to a positive integer, generally between 10 and 20.

[0085] Control cycle T s is the time interval for the controller to perform a control calculation, which is generally 10 milliseconds to 100 milliseconds. In this embodiment, T s =50 milliseconds.

[0086] Further, starting the hydraulic pump to establish the working pressure specifically includes

[0087] Start the hydraulic pump: Press the start button to start the hydraulic pump and establish the system working pressure P s .

[0088] Establish working pressure P s Specifically, after the hydraulic pump is running, the system pressure reaches the set value P by adjusting the overflow valve. s =16MPa.

[0089] Furthermore, the controller loads an adaptive neural network algorithm, including:

[0090] Loading algorithm: After the controller is started, it loads the adaptive neural network model predictive control algorithm and initializes the network parameters and model parameters.

[0091] Initialize the neural network parameters: The weight matrix W and the bias vector b are randomly initialized to satisfy the Gaussian distribution, and the learning rate η is set to 0.01.

[0092] Step 1 is completed, the system has completed initialization and is ready to enter the data collection phase.

[0093] Step 2: Collect the pressure, flow, temperature, vibration frequency and amplitude data of the hydraulic system in real time through the IoT sensor module, including:

[0094] Collect pressure data in real time and measure it using pressure sensors: Install pressure sensors at both ends of the hydraulic cylinder to measure the pressure P in real time 1 (t) and P 2 (t), the sampling frequency is synchronized with the control cycle, that is, every T s = Sampling once every 50 milliseconds.

[0095] Collect flow data in real time and measure it using a flow meter: Install a flow meter at the outlet of the hydraulic pump to measure the hydraulic oil flow Q(t) in real time.

[0096] Collect temperature data in real time and measure it using temperature sensors: Install temperature sensors near the oil tank and hydraulic cylinder to measure the hydraulic oil temperature T(t).

[0097] Collect vibration data in real time and measure with vibration sensors: Install vibration sensors on the hydraulic cylinder and guide vane mechanism to measure the vibration frequency f(t) and amplitude A(t).

[0098] Furthermore, the sensor transmits data to the controller in real time via industrial Ethernet or CAN bus, using security protocols such as TLS / SSL to ensure data integrity and security.

[0099] Step 2 is completed, the real-time data collection is completed, and the data has been transmitted to the controller, ready for data preprocessing and state estimation.

[0100] Step 3: Filter and calibrate the collected data, extract characteristic parameters, and use the improved Kalman filter to estimate the system state vector based on the measured pressure and flow data.

[0101] First, the data is preprocessed, including filtering and data calibration.

[0102] Furthermore, a first-order low-pass filter is applied to the collected raw data to filter out high-frequency noise.

[0103] The filtering formula is as follows:

[0104] y(t)=α·x(t)+(1-Q)·y(t-1)

[0105] Among them, x(t) is the original data currently sampled, y(t) is the filtered data, t is time, and t=0 is the initial startup time.

[0106] α: filter coefficient, the value range is 0<α<1, in this embodiment, α=0.7

[0107] Furthermore, the filtered data is calibrated according to the calibration curve of the sensor to eliminate the system error.

[0108] The calibration formula is as follows: calibrated (t) = k·y(t) + b

[0109] Where: k is the sensor gain and b is the sensor bias.

[0110] Furthermore, feature parameters are extracted.

[0111] Calculate the pressure difference: AP(t) = P 1 (t)-P 2 (t)

[0112] Calculate the average flow rate at the nearest m sampling points, and take m = 5;

[0113]

[0114] Furthermore, the state of the system is estimated, specifically:

[0115] Establish the system state vector, which is defined as

[0116]

[0117] Where: x p (t): piston displacement (measured by sensor or integrated velocity); Piston speed.

[0118] The state space model is established using the improved Kalman filter:

[0119]

[0120] in:

[0121] A, B, C: system matrices; w(t): process noise, assumed to be zero-mean Gaussian white noise, with a covariance matrix of Q w ; v(t): measurement noise, assumed to be zero-mean Gaussian white noise, with a covariance matrix of R v ;y(t): measurement output.

[0122] Prediction steps:

[0123]

[0124] P - (t+1)=AP(t)A T +Q w

[0125] Update the steps:

[0126] K(t+1)=P - (t+1)C T [CP - (t+1)C T +R v ] -1

[0127]

[0128] P(t+1)=[IK(t+1)C]P - (t+1)

[0129] in: State estimate, P(t): estimation error covariance matrix, K(t): Kalman gain.

[0130] The improvement lies in adjusting the process noise covariance Q according to real-time data w and the measurement noise covariance R v , improve filtering accuracy.

[0131] Step 3 ends and the estimated value of the system state at the current moment is obtained Provide input for the next step of adaptive model updating.

[0132] Step 4: Update system parameters online based on the adaptive neural network, where the input of the neural network is the current state and control input, and the output is the predicted state at the next moment.

[0133] The adaptive neural network structure includes an input layer, a hidden layer and an output layer.

[0134] The input layer includes the current state And control input u(t), a total of 3 neurons.

[0135] The hidden layer uses one hidden layer with 10 nodes.

[0136] The output layer outputs the predicted state at the next moment There are 2 neurons in total.

[0137] Online training of the above network structure includes:

[0138] Input vector:

[0139]

[0140] Output vector:

[0141]

[0142] Loss function:

[0143]

[0144] Weight update:

[0145] Using the error back propagation algorithm, update the network weight W and bias b:

[0146]

[0147]

[0148] Where η is the learning rate.

[0149] Next, the model parameters are updated, and the system matrices A and B are updated online according to the prediction results of the neural network for the next step of optimization control.

[0150] Step 4 ends, and the updated system model and the predicted state at the next moment are obtained. Provide a model basis for optimal control.

[0151] Step 5: Use the adaptive neural network after updating the system parameters to construct the objective function of the control model. Under the constraints of the control input constraints and state constraints, solve the optimization problem and obtain the optimal control input to construct the objective function as follows:

[0152]

[0153] Among them, x(k+i|k) is the state predicted at the k+ith moment based on the information at the kth moment;

[0154] x ref (k+i) is the reference state at the k+ith moment;

[0155] Δu(k+i)=u(k+i)-u(k+i-1).

[0156] Q is the state error weight matrix, Q = diag (q 1 ,q 2 );

[0157] R is the control increment weight matrix, take R = r;

[0158] Get q 1 =1,q 2 =0.1, r=0.01.

[0159] Furthermore, constraints are set.

[0160] Include control input constraints:

[0161] u min ≤u(k+i)≤u max

[0162] in:

[0163] u min =0

[0164] u max =1

[0165] State constraints:

[0166] x min ≤x(k+i|k)≤x max

[0167] in:

[0168]

[0169]

[0170] Pick

[0171] Pick

[0172] Furthermore, the sequential quadratic programming (SQP) algorithm is used to solve the optimization problem.

[0173] The solution process includes initialization, setting the initial control input sequence u(k), u(k+1), ..., u(k+N-1).

[0174] Iterative solution: In each iteration, linearize the objective function and constraints, construct a quadratic programming subproblem, solve the quadratic programming subproblem, and obtain the increment of the control input; update the control input sequence and check the convergence conditions.

[0175] Set the termination condition as follows: when the change in the objective function value is less than the set threshold (such as 10 -6 ), or when the maximum number of iterations is reached, the iteration is terminated.

[0176] Optimal control input u(k): The first value u(k) of the control input sequence obtained by optimization is taken as the optimal control input at the current moment.

[0177] At the end of step 5, the optimal control input u(k) is obtained, which provides input for actuator control.

[0178] Step 6: Send the optimal control input to the control valve group, adjust the valve opening, control the hydraulic actuator, and adjust the angle of the guide vane.

[0179] The optimal control input is converted into valve opening, and the valve opening is calculated as follows:

[0180] A v (u(k))=A max u(k)

[0181] in:

[0182] A max For the maximum opening of the valve port, take A max =10mm 2

[0183] Sending a control signal is beneficial to digital-to-analog conversion, converting the digital control signal u(k) into an analog voltage signal V to control the proportional solenoid valve.

[0184] The conversion relationship is V = V max u(k)

[0185] Where: V max =10V

[0186] Then the proportional solenoid valve acts to adjust the valve opening. After receiving the voltage signal V, the proportional solenoid valve adjusts the valve opening A. v (u(k)), controls the flow rate Q(t) of hydraulic oil.

[0187] Control hydraulic actuator

[0188] Hydraulic cylinder movement:

[0189] The adjusted hydraulic oil flow makes the hydraulic cylinder piston move at a speed and displacement x p (t) is calculated according to the following formula:

[0190]

[0191]

[0192] Among them, A p is the effective area of ​​the piston.

[0193] The guide vane angle is adjusted by calculating and converting the angle:

[0194] Piston displacement x p The relationship between (t) and the guide vane angle θ(t):

[0195] θ(t)=k θ ·x p (t)

[0196] Where: k θ is the displacement-angle conversion coefficient, which depends on the mechanical structure

[0197] At the end of step 6, the guide vane angle has been adjusted according to the optimal control input, achieving precise control of the turbine.

[0198] Step 7: Use the energy recovery device to store the excess energy in the hydraulic system in the supercapacitor through the hydraulic-electric energy conversion device. When the system needs it, the supercapacitor releases the stored electrical energy through the power converter to assist the hydraulic pump in supplying energy and optimize energy utilization.

[0199] Energy recovery is performed through hydraulic-electrical energy conversion. When the pressure in the hydraulic system exceeds the set threshold P th (such as P th =18), the excess hydraulic energy drives the generator through the hydraulic motor to convert mechanical energy into electrical energy, and the generated electrical energy is stored in the supercapacitor through the power converter (DC / DC converter).

[0200] Energy is released to provide auxiliary energy when needed. For example, when the system needs to respond quickly or the hydraulic pump load is too high, the supercapacitor releases the stored electrical energy through the power converter to assist the hydraulic pump in supplying energy.

[0201] For energy management control, charging condition: when P(t)>P th When the battery is charged, the energy recovery device is turned on for charging.

[0202] Discharge condition: When the hydraulic pump motor load exceeds the set value or requires a quick response, turn on the supercapacitor to discharge.

[0203] Energy balance calculation:

[0204] Energy recovery E rec :

[0205]

[0206] in:

[0207] P excess (t) = [P(t) - P th ]·Q rec (t)

[0208] Q rec (t) is the flow rate into the hydraulic motor

[0209] Energy release E rel for

[0210]

[0211] in:

[0212] P cap (t) is the supercapacitor voltage;

[0213] I cap is the supercapacitor discharge current;

[0214] Step 7 ends, completing the energy recovery and utilization, and improving the energy utilization efficiency of the system.

[0215] Through the above detailed steps, this method realizes the precise control and energy optimization of the hydraulic servo system of the turbine. The various steps are closely related, and data and information are effectively transmitted between the steps, ensuring the real-time and stability of the system. The adaptive neural network model predictive control algorithm and energy recovery device adopted are innovative and practical, and can effectively improve the operating efficiency and energy-saving effect of the turbine.

[0216] We have invented a hydraulic servo system for water turbines that introduces Internet of Things sensor modules, adaptive neural network model predictive control algorithms and energy recovery modules to achieve accurate monitoring of real-time operating parameters of the system, intelligent regulation and efficient use of energy.

[0217] Specifically, the role and effect of the hydraulic actuator: By setting up a hydraulic actuator connected to the adjustable component of the turbine, the mechanism can effectively adjust the angle of the guide vane, thereby achieving precise control of the water flow. This design uses hydraulic drive, which has a higher response speed and load capacity than mechanical transmission. At the same time, the high precision and stability of the hydraulic actuator lays a solid foundation for the implementation of subsequent control algorithms, which can significantly improve the overall efficiency of turbine operation.

[0218] Function and effect of control valve group: The control valve group is used to accurately control the flow direction of hydraulic oil to drive the hydraulic actuator to adjust the guide vane angle. By dynamically adjusting the valve group opening, the pressure fluctuation in the hydraulic system can be effectively reduced to avoid energy waste or component wear caused by hydraulic shock. Its function is to provide a stable hydraulic power source for the hydraulic actuator to ensure that the control instructions can be executed quickly and accurately, thereby further improving the regulation performance of the turbine.

[0219] Functions and effects of IoT sensor modules: IoT sensor modules including pressure sensors, flow meters, temperature sensors and vibration sensors can monitor the operating parameters of the hydraulic system in real time and transmit data. This design realizes a comprehensive perception of the system's operating status, can obtain key working data in real time, and provide a reliable input basis for subsequent control algorithms. Through real-time monitoring, abnormal conditions can be quickly discovered to avoid system downtime due to failures, thereby improving the safety and operational stability of the system. In addition, the collection of IoT data also provides support for the analysis and optimization of long-term operating conditions.

[0220] Function and Effect of Step 1 (System Initialization)

[0221] The system is initialized by setting the target guide vane angle, initial control input, prediction time domain and control cycle, starting the hydraulic pump to establish working pressure, and loading the adaptive neural network algorithm. This process ensures that the system is in the best working state before operation, provides a stable control basis for subsequent adjustment operations, and avoids abnormal state interference during the system startup phase.

[0222] Step 2: Collect the operating parameters of the hydraulic system in real time, and accurately monitor the hydraulic status through high-frequency sampling to provide high-quality data for subsequent analysis.

[0223] Step 3: Through filtering and calibration, the measurement noise is eliminated and the system state vector is estimated using the improved Kalman filter, which further improves the reliability of the data and the accuracy of system modeling.

[0224] Step 4: Update the neural network model parameters online so that the system can dynamically adjust the control strategy according to the real-time working conditions, thus realizing intelligent and adaptive control.

[0225] Step 5: By constructing the objective function and solving the optimization problem, the optimal control input is obtained, thereby improving the control accuracy, reducing energy loss, and further improving system efficiency.

[0226] Step 6 (Execution Control) Role and Effect

[0227] The optimal control input is sent to the control valve group to adjust the valve opening to control the hydraulic actuator to achieve precise adjustment of the guide vane angle. This step ensures the consistency between the actual guide vane angle and the target angle through a closed-loop control mechanism, significantly improving the operating accuracy and response speed of the turbine.

[0228] Functions and effects of step 7 (energy recovery)

[0229] The energy recovery device converts the excess energy in the hydraulic system into electrical energy and stores it in the supercapacitor. When the system needs it, the stored energy is released to assist the hydraulic pump in supplying energy. This step not only optimizes the energy utilization efficiency, but also reduces the load on the hydraulic pump, further improving the overall energy saving and operational reliability of the system.

[0230] The control device of the present invention is a PLC control module, which can be conveniently and quickly set with an expandable I / O module to expand the interface use.

[0231] Furthermore, the present invention also discloses a control device for the hydraulic servo system for the above-mentioned turbine, comprising a controller body 100, an expansion module 200 and a connecting piece 300, wherein a first card slot 101 is provided at one end of the controller body 100, a first card buckle is provided in the first card slot 101, and the first card buckle is connected to a first pressure rod 102;

[0232] The sensor is indirectly connected to the connection port 205 and transmits data to the controller body 100 .

[0233] A connecting piece 300 is arranged on the side of the controller body 100, and the connecting piece 300 includes a limiting edge strip 301 and a limiting groove 302 between the limiting edge strip 301, and a buckle groove 303 is arranged on the connecting piece 300. A buckle edge 201 is arranged on one side of the expansion module 200, and the buckle edge 201 is embedded in the buckle groove 303. A receiving groove 202 is arranged on one side of the expansion module 200, and a conductive strip 203 is arranged in the receiving groove 202. The limiting edge strip 301 is in the receiving groove 202, and the conductive strip 203 is embedded in the limiting groove 302. A conductive sheet 304 is arranged in the limiting groove 302.

[0234] A second card slot 204 is provided at one end of the expansion module 200, and a second buckle connected to a second pressure rod 206 is also provided in the second card slot 204. A wiring port 205 is provided at the other end of the expansion module 200. Similarly, the second pressure rod 206 is used to unlock the expansion module 200 installed on the board frame 400.

[0235] It should be noted that the first clip is a prior art, which is arranged in the first slot 101 and is connected to the first pressure rod 102. The first pressure rod 102 is moved away from the first slot 101 by pressing the first pressure rod 102. After being released, the first clip enters the first slot 101. When the first slot 101 is put on the plate frame 400 for fixing, the controller body 100 can be fixed on the plate frame 400 by utilizing the first clip.

[0236] Furthermore, an expansion terminal 207 is provided on the other side of the expansion module 200, and a connecting piece 300 is detachably installed on this side for installation and expansion with the next expansion module 200. A third card slot 305 is provided at one end of the connecting piece 300, and a slope 306 is provided at the other end of the connecting piece 300 for guiding the installation of the expansion module 200 on the connecting piece 300.

[0237] The conductive sheet 304 is connected to the extension terminal 207 .

[0238] When the controller body 100 is fixed on the plate frame 400, the buckle edge 201 can be embedded in the buckle groove 303 from the side, the expansion module 200 can be pushed in, the conductive bar 203 can be embedded in the limiting groove 302 and slide to the conductive sheet 304 for connection, and then the second card slot 204 of the expansion module 200 can be inserted into the plate frame 400 and fixed.

[0239] It should be noted that the traditional method uses a fixed parameter proportional-integral-differential (PID) control method to achieve guide vane angle adjustment, detects the hydraulic system status through a single pressure sensor, has no online optimization function, and has no energy recovery device; the method of the present invention is based on an adaptive neural network predictive control algorithm, combined with the Internet of Things sensor module to collect system data in real time, dynamically adjusts the control strategy through online optimization, and combines the energy recovery module to achieve improved energy utilization of the hydraulic system.

[0240] Preferably, the embodiment of the present invention verifies the superiority of the following performance indicators by comparing the traditional method and the method of the present invention:

[0241] Guide vane angle adjustment accuracy: verify the target angle achievement accuracy and adjustment time;

[0242] Stability of hydraulic system: Analyze the pressure fluctuation of hydraulic system;

[0243] Energy recovery efficiency: compares the efficiency of the energy recovery device with the accumulated energy;

[0244] Dynamic response of the control valve group: test the speed and stability of the valve response time.

[0245]

Experimental environment conditions

[0246] Experimental object: guide vane angle adjustment system of a certain type of turbine;

[0247] Initial guide vane angle: 15°; target angle: 30°; prediction time domain: 0.5 seconds; control cycle: 10 milliseconds; hydraulic pump working pressure: 8MPa; experiment duration: 50 seconds.

[0248]

Experimental process

[0249] Traditional method: Initialize the system, start the hydraulic pump to establish working pressure, use a fixed parameter PID controller to adjust the valve opening, adjust the hydraulic actuator, collect pressure data through a single sensor, feed it back to the PID controller, and control according to the preset PID parameters. There is no energy recovery device and the hydraulic shock is not processed;

[0250] The method of the present invention comprises the following steps: initializing the system, starting the hydraulic pump, loading an adaptive neural network control algorithm, using an Internet of Things sensor module to collect multi-parameter data of pressure, flow, temperature, and vibration in real time, filtering and calibrating the data through an improved Kalman filter, dynamically adjusting the control strategy based on an adaptive neural network predictive control model, optimizing the valve opening, and using an energy recovery module to recover excess energy in the hydraulic shock and store it as electrical energy for standby use.

[0251]

Experimental results

[0252] Reference Figure 6 It can be seen intuitively that the guide vane angle adjustment speed of the curve where the traditional method is located is slow, and it takes 50 seconds to stabilize to the target angle (30°). The guide vane angle of the curve where the method of the present invention is located is quickly adjusted to the target value within 30 seconds, without overshoot, and the response is fast and accurate; This shows that the method of the present invention improves the response speed and accuracy of the guide vane angle adjustment.

[0253] Reference Figure 7 It can be seen intuitively that the hydraulic pressure fluctuation amplitude of the curve of the traditional method is large (±0.5MPa), which is easy to cause hydraulic shock; the pressure fluctuation amplitude of the curve of the method of the present invention is significantly reduced (±0.2MPa), and the system runs more smoothly; This shows that the method of the present invention effectively reduces the hydraulic shock and improves the stability of the hydraulic system.

[0254] Reference Figure 8 It can be seen intuitively that the traditional method has a low cumulative recovery energy and a slow recovery speed; the energy recovery efficiency of the method of the present invention is significantly improved, and the cumulative recovery energy is much higher than that of the traditional method.

[0255] Reference Fig. 9 It can be seen intuitively that the valve response time of the curve of the traditional method is longer (0.5 seconds), the fluctuation range is large, and the dynamic response performance is poor; the valve response time of the curve of the method of the present invention is shorter (0.4 seconds), the fluctuation range is smaller, and the dynamic response performance is superior; This shows that the method of the present invention improves the dynamic response speed and stability of the control valve group.

[0256] The following is an example of the actual application of the system of this application

[0257] Application scenario: A 100MW turbine generator set was installed in a large hydropower station. In order to improve the operating efficiency and control accuracy of the turbine, it was decided to use a hydraulic servo control system based on adaptive neural network and IoT sensors to accurately adjust the guide vanes of the turbine.

[0258] Objective: To optimize turbine output, reduce energy loss, and improve power generation efficiency by precisely controlling the guide vane angle.

[0259] Specific implementation steps

[0260] Step 1: System Initialization

[0261] 1.1 Setting initial parameters

[0262] Target guide vane angle θ ref : According to the current grid demand, the guide vane angle needs to be adjusted to 30°.

[0263] Initial control input u(0): The initial setting is 0, indicating that the valve port is initially closed.

[0264] Prediction time domain N: Select the prediction time domain as 15 to balance the computational complexity and prediction accuracy.

[0265] Control cycle T s : Set the control period to 50 milliseconds (0.05 seconds) to ensure the real-time performance of the system.

[0266] 1.2 Start the hydraulic pump to establish working pressure

[0267] Start the hydraulic pump: The operator presses the start button in the control room and the hydraulic pump starts running.

[0268] Establish working pressure P s : Adjust the relief valve and set the working pressure of the hydraulic system to 16MPa.

[0269] 1.3 Controller loading adaptive neural network algorithm

[0270] Loading algorithm: The controller (industrial computer) starts the adaptive neural network model predictive control algorithm.

[0271] Initialize the neural network parameters:

[0272] Input layer: 3 nodes (piston displacement, piston velocity, control input).

[0273] Hidden layer: 10 nodes.

[0274] Output layer: 2 nodes (predicted piston displacement and velocity).

[0275] Weights and biases: Randomly initialized, satisfying a normal distribution with mean zero and variance 0.10.10.1.

[0276] Set the learning rate η = 0.01.

[0277] Step 1 is completed, the system is ready to enter the data collection phase.

[0278] Step 2: Data Collection

[0279] 2.1 Real-time collection of pressure data

[0280] Install the pressure sensor: Install the pressure sensor at both ends of the hydraulic cylinder (A and B ends).

[0281] Data collection: Every T s = Collect pressure data once every 50 milliseconds P 1 (t) and P 2 (t).

[0282] 2.2 Real-time collection of traffic data

[0283] Install flow meter: Install a high-precision flow meter at the outlet of the hydraulic pump.

[0284] Data collection: Collect hydraulic oil flow Q(t).

[0285] 2.3 Real-time temperature data collection

[0286] Install temperature sensors: Install temperature sensors near the oil tank and hydraulic cylinder;

[0287] Data collection: Collect hydraulic oil temperature T(t).

[0288] 2.4 Real-time collection of vibration data

[0289] Install vibration sensors: Install vibration sensors on the hydraulic cylinder and guide vane mechanism.

[0290] Data collection: Collect vibration frequency f(t) and amplitude A(t).

[0291] 2.5 Data Transmission

[0292] Communication method: The sensor transmits data to the controller via industrial Ethernet.

[0293] Security measures: Use TLS / SSL protocol to ensure data transmission security.

[0294] Step 2 is completed and the real-time data is successfully transmitted to the controller.

[0295] Step 3: Data preprocessing and state estimation

[0296] 3.1 Data Preprocessing

[0297] 3.1.1 Filtering

[0298] Low-pass filter: Perform first-order low-pass filtering on the collected pressure, flow and other data to filter out high-frequency noise.

[0299] For example, for the pressure data:

[0300] P filtered (t) = α·P(t) + (1-α)·P filtered (tT s )

[0301] Among them, the filter coefficient α=0.7.

[0302] 3.1.2 Data calibration

[0303] Calibration formula: P calibrated (t) = k P ·P filtered (t)+b P

[0304] k P and b P is the calibration coefficient of the sensor (provided by the manufacturer, assuming k P =1.02, b P =-0.1MPa).

[0305] A similar calibration process is performed on other sensor data.

[0306] 3.1.3 Feature Parameter Extraction

[0307] Calculate the pressure difference ΔP(t): ΔP(t) = P 1 (t)-P 2 (t)

[0308] Calculate average flow

[0309] 3.2 State Estimation

[0310] 3.2.1 Establishing the system state vector

[0311] Piston displacement x p (t): Directly measured by displacement sensor or obtained by integrating velocity.

[0312] Piston speed Calculated by the following formula:

[0313] Among them, the effective area of ​​the piston is A p Known (assuming A p=0.01m 2 ).

[0314] State vector:

[0315]

[0316] 3.2.2 Application of Improved Kalman Filter

[0317] System model: x(t+1)=Ax(t)+Bu(t)+w(t)

[0318] A and B are system matrices, and the initial settings are:

[0319] Among them, k u is the control gain, assuming k u =0.5.

[0320] The filtering process includes the prediction stage:

[0321] Also includes the update phase:

[0322] K(t+1)=P - (t+1)C T [CP - (t+1)C T +R] -1

[0323]

[0324] Among them, y(t+1) is the measurement value and C is the measurement matrix.

[0325] It also includes the covariance matrix update: P(t+1)=[IK(t+1)C]P - (t+1)

[0326] Step 3 is completed and the estimated value of the system state is obtained

[0327] Step 4: Adaptive model update

[0328] 4.1 Online training of adaptive neural networks

[0329] Input vector:

[0330] The target output is the actual measured state at the next moment:

[0331] The network output is the predicted state at the next moment:

[0332] Among them, NN represents the neural network model.

[0333] 4.2 Calculating the loss function

[0334] Loss function:

[0335]

[0336] 4.3 Update Neural Network Parameters

[0337] Error back propagation: According to the loss function E(t), the gradient of weights and biases is calculated.

[0338] Parameter update:

[0339] 4.4 Update the system model and use the updated neural network to obtain a more accurate system model for the next step of optimization control.

[0340] Step 4 is completed, the system model has been updated, and the prediction accuracy has improved.

[0341] Step 5: Construct and solve the optimization problem

[0342] 5.1 Constructing the objective function

[0343] Reference trajectory x ref (k+i): Set the target guide vane angle θ ref =30° converted to the corresponding piston displacement x ref , assuming that the displacement-angle conversion coefficient k θ =0.003m / °, then:

[0344] x ref =k θ ·θ ref =0.003×30=0.09m

[0345] Objective function:

[0346]

[0347] Among them, the weight coefficient is q=1, r=0.01.

[0348] 5.2 Constraints: Control input constraint: 0≤u(k+i)≤1; State constraint: 0≤x p (k+i|k)≤0.1m.

[0349] 5.3 Optimization Solution

[0350] The sequential quadratic programming (SQP) algorithm is used to solve the optimal control input sequence u(k), u(k+1), ..., u(k+N-1) within the prediction horizon N=15.

[0351] Specific solution steps: Initialize the control input sequence, for example, set all to the control input u(k-1) of the previous moment.

[0352] In each iteration: calculate the predicted state sequence x p (k+i|k), calculate the objective function J and its gradient, update the control input sequence, and iterate until convergence.

[0353] 5.4 Result Output

[0354] Optimal control input u(k): Take the first value of the optimized control input sequence as the control input at the current moment; assume that the calculation result is u(k) = 0.6.

[0355] Step 5 is completed, and the optimal control input u(k)=0.6 is obtained.

[0356] Step 6: Controlling the Actuator

[0357] 6.1 Convert the optimal control input into valve opening, the maximum valve opening A max =10mm 2 .

[0358] Calculate valve opening: A v (u(k))=A max u(k)=10×0.6=6mm 2

[0359] 6.2 Sending control signals

[0360] Digital-to-analog conversion: Convert u(k)=0.6 to a voltage signal V:

[0361] V=V max u(k)=10×0.6=6V

[0362] Send signal: The controller outputs 6V voltage signal to the proportional solenoid valve.

[0363] 6.3 Adjust the valve opening

[0364] Proportional solenoid valve action: valve opening adjusted to 6mm 2 , control the hydraulic oil flow.

[0365] 6.4 Controlling Hydraulic Actuators

[0366] Calculate the hydraulic oil flow Q(t): According to the valve port flow formula:

[0367]

[0368] Where: C d is the flow coefficient, take 0.6.

[0369] ΔP(t)=P s -P load , assuming P load =10MPa, then ΔP(t)=6MPa.

[0370] ρ is the density of hydraulic oil, which is 850kg / m 3 .

[0371] Calculated:

[0372] Calculate piston speed

[0373] Update piston displacement x p (t):

[0374]

[0375] 6.5 Guide vane angle adjustment

[0376] Calculate the guide vane angle θ(t):

[0377] For example, if x p (t) = 0.0495m, then:

[0378] Repeat: Controller every T s = The control calculation is performed once every 50 milliseconds, and the guide vane angle is gradually adjusted to the target angle of 30°.

[0379] Step 6 is completed, the guide vane angle is gradually adjusted, and the system runs smoothly.

[0380] Condition detection: When the hydraulic system pressure P s When the pressure exceeds 18MPa, the energy recovery device is activated.

[0381] Hydraulic-electric energy conversion: Excess hydraulic oil drives the hydraulic motor to rotate, driving the generator to generate electricity.

[0382] Energy storage: The generated energy is stored in supercapacitors via a DC / DC converter.

[0383] Detection demand: When the hydraulic pump load is too high or a quick response is required, the controller sends a command.

[0384] Auxiliary energy supply: The supercapacitor supplies power to the hydraulic pump motor through a DC / DC converter, reducing the load on the main power supply.

[0385] Energy recovery calculation: Assume that at t 0 = 10 seconds to t 1 = 15 seconds, the recovery power is P excess=5kW, then:

[0386] Energy release calculation: During the time period t when auxiliary energy supply is required 2 = 20 seconds to t 3 = 22 seconds, the discharge power is P rel =4kW, then:

[0387]

[0388] Step 7 is completed, the energy recovery and utilization system operates normally, and energy is saved.

[0389] Through the above specific implementation steps, combined with actual application scenarios, the implementation process of the hydraulic servo method for water turbines is described in detail. The data parameters and calculation methods of each step are clear, the process is clear, and it is easy to understand.

[0390] Key results:

[0391] Precisely control the guide vane angle to meet grid requirements and improve power generation efficiency.

[0392] Adaptive neural network control algorithm improves the robustness and response speed of the system.

[0393] Energy recovery and utilization effectively save energy and reduce operating costs.

[0394] Further additional explanation:

[0395] Hydraulic servo system: controls the movement of mechanical parts through the pressure and flow of hydraulic oil to achieve precise adjustment.

[0396] Adaptive neural network: An artificial intelligence algorithm that automatically adjusts control strategies based on real-time data to improve system performance.

[0397] Kalman filter: A data processing algorithm that can estimate the true state of a system from noisy measurement data.

[0398] Energy recovery device: Recovers and stores excess energy in the system and reuses it when needed to achieve energy saving.

[0399] [Table data example]

[0400] Table 1: Initial parameter settings

[0401]

[0402] Table 2: Optimized control parameters

[0403]

[0404] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A hydraulic servo system for a water turbine, characterized in that: include, A hydraulic actuator connected to the adjustable part of the turbine to adjust the angle of the guide vanes; A control valve group is used to control the flow of hydraulic oil to the hydraulic actuator; IoT sensor modules, including pressure sensors, flow meters, temperature sensors, and vibration sensors, monitor the operating parameters of the hydraulic system in real time and transmit data; A controller that receives real-time data transmitted by the IoT sensor module and uses an adaptive neural network model predictive control algorithm to generate control signals and control the control valve group to reduce hydraulic shock and energy loss and improve system stability and accuracy; The energy recovery module is arranged in the hydraulic circuit and is used to recover the excess hydraulic energy when pressure shock occurs in the hydraulic system and store it for later use.

2. The hydraulic servo system for a water turbine according to claim 1, characterized in that: The working process of the adaptive neural network model predictive control algorithm includes: Step 1: Initialize the system, set initial parameters, including target guide vane angle, initial control input, prediction time domain, control cycle, start the hydraulic pump to establish working pressure, and load the adaptive neural network model predictive control algorithm to the controller; Step 2: Collect the pressure, flow, temperature, vibration frequency and amplitude data of the hydraulic system in real time through the IoT sensor module; Step 3: Filter and calibrate the collected data, extract characteristic parameters, and use the improved Kalman filter to estimate the system state vector based on the measured pressure and flow data; Step 4: Update system parameters online based on the adaptive neural network, where the input of the neural network is the current state and control input, and the output is the predicted state at the next moment; Step 5: Use the adaptive neural network after updating the system parameters to construct the objective function of the control model, solve the optimization problem under the constraints of the control input constraint and the state constraint, and obtain the optimal control input; Step 6: Send the optimal control input to the control valve group, adjust the valve opening, control the hydraulic actuator, and adjust the angle of the guide vane; Step 7: Use the energy recovery device to store the excess energy in the hydraulic system in the supercapacitor through the hydraulic-electric energy conversion device. When the system needs it, the supercapacitor releases the stored electrical energy through the power converter to assist the hydraulic pump in supplying energy and optimize energy utilization.

3. The hydraulic servo system for a water turbine according to claim 2, characterized in that: The controller loads the adaptive neural network model predictive control algorithm, including: Loading algorithm: After the controller is started, it loads the adaptive neural network model predictive control algorithm and initializes the network parameters and model parameters; Initialize the neural network parameters: The weight matrix W and the bias vector b are randomly initialized to satisfy the Gaussian distribution, and the learning rate η is set.

4. The hydraulic servo system for a water turbine according to claim 3, characterized in that: The IoT sensor module collects data from the hydraulic system in real time, including: Measurement using pressure sensors: Install pressure sensors at both ends of the hydraulic cylinder to measure the pressures P1(t) and P2(t) in real time; Install a flow meter at the outlet of the hydraulic pump to measure the hydraulic oil flow Q(t) in real time; Install temperature sensors near the oil tank and hydraulic cylinder to measure the hydraulic oil temperature T(t); Vibration sensors are installed on the hydraulic cylinder and guide vane mechanism to measure the vibration frequency f(t) and amplitude A(t).

5. The hydraulic servo system for a water turbine according to claim 4, characterized in that: Filtering of the collected data includes: Apply a first-order low-pass filter to the acquired raw data to remove high-frequency noise; The filtering formula is as follows: y(t)=α·x(t)+(1-α)·y(t-1) Among them, x(t) is the original data of the current sampling, y(t) is the filtered data, t is the time, and α is the filter coefficient.

6. The hydraulic servo system for a water turbine according to claim 5, characterized in that: Calibration of the filtered data includes, The calibration formula is as follows: x calionraed (t)=k·y(t)+b Where: k is the sensor gain and b is the sensor bias.

7. The hydraulic servo system for a water turbine according to claim 6, characterized in that: Extraction of feature parameters includes: Calculate the pressure difference: ΔP(t)=P1(t)-P2(t) Calculate the average flow rate: Calculate the average flow rate at the most recent m sampling points:

8. The hydraulic servo system for a water turbine according to claim 7, characterized in that: Perform state estimation on the system, including, Establish the system state vector, which is defined as Among them, x p (t) is the piston displacement, is the piston speed.

9. The hydraulic servo system for a water turbine according to claim 8, characterized in that: The state space model is established using the improved Kalman filter: Where: A, B, C are system matrices, w(t) is process noise, assumed to be zero-mean Gaussian white noise, and the covariance matrix is ​​Q w , v(t) is the measurement noise, which is assumed to be zero-mean Gaussian white noise, and the covariance matrix is ​​R v , y(t) is the measurement output.

10. A control device for the hydraulic servo system for the above-mentioned water turbine, characterized in that: The controller comprises a controller body (100), an expansion module (200) and a connecting piece (300); a first card slot (101) is provided at one end of the controller body (100); a first card buckle is provided in the first card slot (101); and the first card buckle is connected to a first pressure rod (102); A connecting piece (300) is provided on the side of the controller body (100), the connecting piece (300) comprises a limiting edge strip (301) and a limiting groove (302) between the limiting edge strip (301), a buckle groove (303) is provided on the connecting piece (300), a buckle edge (201) is provided on one side of the expansion module (200), the buckle edge (201) is embedded in the buckle groove (303), a receiving groove (202) is provided on one side of the expansion module (200), a conductive strip (203) is provided in the receiving groove (202), the limiting edge strip (301) is in the receiving groove (202), the conductive strip (203) is embedded in the limiting groove (302), and a conductive sheet (304) is provided in the limiting groove (302).

Images