CN-121296365-B - Wind turbine variable-pitch working condition characteristic analysis system and method

CN121296365BCN 121296365 BCN121296365 BCN 121296365BCN-121296365-B

Abstract

The invention relates to the technical field of wind power generation equipment control and state monitoring, in particular to a wind turbine variable-pitch working condition characteristic analysis system and method, comprising the following steps: S1, synchronously acquiring at least one parameter of real-time torque, hydraulic circuit pressure fluctuation, blade bearing friction torque and current change rate of a variable-pitch motor of a variable-pitch driving mechanism by using a multi-dimensional sensor through a multi-dimensional transient sensing subsystem. According to the invention, the multi-dimensional transient sensing subsystem is used for collecting multiple parameters such as real-time torque of a variable pitch driving mechanism, pressure fluctuation of a hydraulic circuit and the like, the wind speed and the rotating speed are combined, the transient working condition characteristic vector is obtained through adaptive filtering preprocessing and data fusion mapping, accurate data are provided for transient analysis, and the staged dynamic subsidence modeling subsystem is used for calling a corresponding coupling model according to the stages such as start and stop of variable pitch, so that the subsidence error predicted value is accurately calculated, and the problem that the transient subsidence characteristic is lack of accurate analysis in the prior art is solved.

Inventors

ZHANG HUI
ZHOU JIANXING
YANG TAO
JIANG HONG
LUO JIANQING
ZHOU ZONGJIE
YANG BO
YANG WEI

Assignees

北京亿美博科技有限公司
新疆大学

Dates

Publication Date: 20260512
Application Date: 20250929

Claims (8)

1. The characteristic analysis method for the variable pitch working condition of the wind turbine is characterized by comprising the following steps: s1, synchronously acquiring at least one parameter of real-time torque, hydraulic circuit pressure fluctuation, blade bearing friction torque and current change rate of a variable-pitch motor of a variable-pitch driving mechanism by using a multi-dimensional transient sensing subsystem through a multi-channel data acquisition instrument, and mapping the parameter into a transient working condition feature vector through a data fusion technology by combining wind speed and rotating speed parameters; In the step S1, parameters are obtained through a multidimensional transient sensing subsystem, and the parameters are mapped into transient working condition feature vectors through a data fusion technology by combining wind speed and rotation speed parameters, and the method specifically comprises the following steps: s11, acquiring at least one parameter of wind speed, rotating speed, real-time torque of a variable pitch driving mechanism, pressure fluctuation of a hydraulic circuit, friction torque of a blade bearing and current change rate of a variable pitch motor in real time through a multichannel data acquisition instrument; S12, performing data cleaning pretreatment on the acquired parameters by adopting an adaptive filtering algorithm to eliminate interference signals caused by turbulent wind or power grid fluctuation; s13, mapping the preprocessed multidimensional parameter into a transient working condition feature vector serving as an input variable of a subsequent modeling through a multisource data fusion technology; S2, constructing a local subsystem through staged dynamic subsidence, calling a pre-established dynamic subsidence model according to a working condition stage where a pitching process is located, and calculating a subsidence error predicted value under the current working condition, wherein the pitching process comprises an initial stage and a termination stage, the initial stage is a stage of triggering a pitching instruction to start movement of a blade, and the termination stage is a stage of moving the blade from the beginning to the end; In the step S2, a local subsystem is constructed by staged dynamic subsidence, and a pre-established dynamic subsidence model is called according to the working condition stage where the pitch process is located, so as to calculate a subsidence error prediction value under the current working condition, which specifically comprises: S21, when the pitch process is in an initial stage, at least one of real-time torque, a current change rate of a pitch motor and a friction torque of a blade bearing is taken as an input variable, an inertia overcoming-sinking error coupling model is constructed, and the influence of mechanical inertia on the starting delay of the blade is quantified based on the inertia overcoming-sinking error coupling model so as to establish a dynamic function of the change of the error along with the input variable; S22, when the pitch process is in a termination stage, constructing a damping braking-sinking error coupling model by taking at least one of pressure fluctuation and rotation speed attenuation rate of a hydraulic circuit as input variables, and quantitatively representing the influence of hydraulic damping on stopping overshoot of the blade based on the damping braking-sinking error coupling model so as to form a corresponding relation between errors and the input variables; s23, driving parameter correction of the dynamic subsidence model through real-time data, wherein the real-time data comprises transient working condition feature vectors; S3, generating a composite compensation instruction based on the subsidence error predicted value through a self-adaptive error compensation method, and sending the composite compensation instruction to a pitch drive mechanism to execute compensation action, wherein the composite compensation instruction is a combination of a feedforward compensation instruction and a feedback compensation instruction; and S4, periodically optimizing parameters of the split-stage dynamic subsidence model and compensation coefficients of the self-adaptive error compensation control subsystem through the working condition iteration optimization subsystem.
2. The method for analyzing characteristics of variable-pitch working conditions of a wind turbine according to claim 1, wherein in the step S2, the working condition phases of the variable-pitch process further comprise an acceleration phase and a uniform speed phase, the acceleration phase is a phase of the blade from a static state to a target rotating speed, and the uniform speed phase is a phase of stable adjustment of the blade.
3. The method according to claim 1, wherein in the step S3, a composite compensation command is generated based on the predicted value of the subsidence error by an adaptive error compensation method, and the composite compensation command is sent to the pitch driving mechanism to execute a compensation action, and the method specifically comprises: S31, when the pitch command is triggered, outputting a preset command in advance according to the initial stage model, wherein the preset command is used for counteracting starting delay caused by mechanical inertia so as to generate a feedforward compensation command; S32, before the variable pitch is terminated, dynamically adjusting a hydraulic damping parameter or a motor braking torque according to a termination stage model, wherein the hydraulic damping parameter or the motor braking torque is used for inhibiting a blade overshoot error so as to generate a feedback compensation instruction; S33, combining the characteristic vector of the current transient working condition, automatically matching an optimal compensation coefficient, wherein the optimal compensation coefficient is used for generating a composite compensation instruction.
4. The method for analyzing characteristics of variable pitch working conditions of a wind turbine according to claim 1, wherein in the step S4, parameters of the split-phase dynamic subsidence model and compensation coefficients of the adaptive error compensation control subsystem are periodically optimized through the working condition iterative optimization subsystem, and the method specifically comprises the following steps: S41, establishing a variable-pitch working condition multi-source database, wherein the variable-pitch working condition multi-source database is used for storing multi-dimensional sensing data, subsidence error data and compensation effect data under different turbulence intensities or power grid fault types; S42, adopting a reinforcement learning algorithm, and periodically optimizing parameters of the staged dynamic subsidence model and compensation coefficients of the adaptive error compensation control subsystem by taking at least one of subsidence error minimization, power fluctuation stabilization and structural load reduction as an optimization target.
5. The method for analyzing characteristics of variable-pitch working conditions of a wind turbine according to claim 1, wherein the method further comprises: After the system is started, the multidimensional transient sensing subsystem completes sensor calibration and loads initial model parameters after history working condition optimization.
6. The method for analyzing characteristics of variable-pitch working conditions of a wind turbine according to claim 1, wherein the method further comprises: And periodically extracting working condition data and compensation effect data, and updating parameters of the staged dynamic subsidence model and compensation coefficients of the self-adaptive error compensation control subsystem through the working condition iterative optimization subsystem.
7. A wind turbine variable pitch operating condition characteristic analysis system applied to a wind turbine variable pitch operating condition characteristic analysis method according to any one of claims 1 to 6, the system comprising: The multi-dimensional transient sensing subsystem is used for synchronously acquiring at least one parameter of real-time torque, hydraulic circuit pressure fluctuation, blade bearing friction torque and current change rate of the variable pitch motor of the variable pitch driving mechanism, and mapping the parameter into a transient working condition feature vector through a multi-source data fusion technology by combining wind speed and rotating speed parameters; The staged dynamic subsidence modeling subsystem is used for calling a pre-established dynamic subsidence model according to a working condition stage where a pitching process is located, calculating a subsidence error prediction value under the current working condition, wherein the pitching process comprises an initial stage and a termination stage, the initial stage is a stage from triggering of a pitching instruction to starting of movement of a blade, and the termination stage is a stage from movement to stopping of the blade; The adaptive error compensation control subsystem is used for generating a composite compensation instruction based on the subsidence error predicted value and sending the composite compensation instruction to the pitch drive mechanism to execute a compensation action, wherein the composite compensation instruction is a combination of a feedforward compensation instruction and a feedback compensation instruction; And the working condition iterative optimization subsystem is used for periodically optimizing parameters of the bisection dynamic subsidence model and compensation coefficients of the self-adaptive error compensation control subsystem.
8. The wind turbine variable pitch operating condition feature analysis system of claim 7, wherein the multi-dimensional transient awareness subsystem comprises: The distributed multi-channel sensors are used for synchronously collecting at least one parameter of wind speed, rotating speed, real-time torque of the variable pitch driving mechanism, pressure fluctuation of a hydraulic circuit, friction torque of a blade bearing and current change rate of a variable pitch motor; the data preprocessing module is used for carrying out data cleaning preprocessing on the parameters acquired by the distributed sensor by adopting a self-adaptive filtering algorithm so as to eliminate interference signals caused by environmental noise, equipment vibration noise, turbulent wind or power grid fluctuation; the data fusion module is used for mapping the preprocessed multidimensional parameters into transient working condition feature vectors which are used as input of subsequent modeling.

Description

Wind turbine variable-pitch working condition characteristic analysis system and method Technical Field The invention relates to the technical field of wind power generation equipment control and state monitoring, in particular to a wind turbine variable-pitch working condition characteristic analysis system and method. Background In a wind generating set, a pitch system is a core component for realizing power adjustment and set protection, and controls the rotating speed and output power of a wind wheel by adjusting the pitch angle of blades. Conventional pitch systems typically rely on motor drive and gear drive arrangements, requiring real-time monitoring of wind speed, rotational speed, etc. parameters and dynamic adjustment of pitch angle. Along with the large-scale development of the wind turbine generator, the dynamic response precision and reliability requirements of the pitch system under the complex working condition are obviously improved, and particularly under the condition of extreme wind speed fluctuation or grid fault, the problems of overspeed of the wind turbine generator, rapid increase of mechanical load and the like can be caused by delay or error of pitch action. In the prior art, performance evaluation of a pitch system depends on parameter calibration under steady-state working conditions, and the dynamic subsidence characteristics of the pitch system in a transient process are lack of accurate analysis. The method is characterized in that in the rapid adjustment process of the pitch system, due to superposition of factors such as hydraulic damping, mechanical inertia and control delay, nonlinear deviation exists between the actual position and the target position of the blade, and the deviation is negligible in a steady state, but power fluctuation and structural fatigue are caused when the pitch system is frequently started and stopped or suddenly changed. The existing solution focuses on improving the control algorithm or hardware response speed, but does not build a dynamic characteristic model aiming at the subsidence amount of the initial stage and the final stage of the pitching action, so that the system cannot adaptively compensate the subsidence errors under different working conditions, and the running stability of the large-sized unit in a turbulent wind field is seriously affected. Disclosure of Invention Aiming at the defects of the prior art, the invention provides a wind turbine variable-pitch working condition characteristic analysis system and a wind turbine variable-pitch working condition characteristic analysis method, which solve the problems of power fluctuation and structural fatigue caused by lack of accurate analysis of transient subsidence characteristics and incapability of self-adaptive compensation of subsidence errors when compared with the prior art. In order to achieve the purpose, the invention is realized by the following technical scheme that the method for analyzing the characteristics of the variable pitch working condition of the wind turbine comprises the following steps: s1, synchronously acquiring at least one parameter of real-time torque, hydraulic circuit pressure fluctuation, blade bearing friction torque and current change rate of a variable-pitch motor of a variable-pitch driving mechanism by using a multi-dimensional transient sensing subsystem through a multi-channel data acquisition instrument, and mapping the parameter into a transient working condition feature vector through a data fusion technology by combining wind speed and rotating speed parameters; S2, constructing a local subsystem through staged dynamic subsidence, calling a pre-established dynamic subsidence model according to a working condition stage where a pitching process is located, and calculating a subsidence error predicted value under the current working condition, wherein the pitching process comprises an initial stage and a termination stage, the initial stage is a stage of triggering a pitching instruction to start movement of a blade, and the termination stage is a stage of moving the blade from the beginning to the end; S3, generating a composite compensation instruction based on the subsidence error predicted value through a self-adaptive error compensation method, and sending the composite compensation instruction to a pitch drive mechanism to execute compensation action, wherein the composite compensation instruction is a combination of a feedforward compensation instruction and a feedback compensation instruction; and S4, periodically optimizing parameters of the split-stage dynamic subsidence model and compensation coefficients of the self-adaptive error compensation control subsystem through the working condition iteration optimization subsystem. Further, in the step S1, parameters are obtained through a multidimensional transient sensing subsystem, and the parameters are mapped into transient working condition feature vectors through a multi-source dat