Research On The Active Disturbance Rejection Control Strategy Of Wind Turbine Blade Sub-Component Performance Test System

The development trend of large blade makes the technical difficulties and development cost of full-size test increase sharply.The testing method of sub-components has gradually become a research hotspot in the field of blade design.However,the wafer component performance test system of wind turbine blade is a nonlinear,uncertain and highly interfered complex system,and its precise mathematical model cannot be obtained.Therefore,in the actual field test process,the control performance of the system often deteriorates due to the model imprecision and internal and external disturbances.In order to improve the loading accuracy and anti-interference performance of wind turbine blade blade component performance test system,this paper proposes a wind turbine blade component performance test technology based on active disturbance rejection control algorithm,and uses an improved BP neural network learning algorithm to self-tune the main performance parameters of active disturbance rejection controller online.Specific research contents are as follows:Firstly,the mathematical model of hydraulic servo loading system for wind turbine blade component performance test is established.Through the overall design of the hydraulic servo system loading scheme and the selection of the main components,a complete set of hydraulic station is built.At the same time,starting from the hydraulic servo control principle,wind turbine blade components were equivalent to linear elastomer,and the hydraulic servo system characteristics were analyzed.The mathematical model of the whole hydraulic servo loading system including oil pump,servo valve and hydraulic cylinder was established,which provided necessary conditional support and model reference for the subsequent design of active disturbance rejection control algorithm.Secondly,in order to improve the anti-interference performance and control accuracy of the wind turbine blade component performance test system,the active disturbance rejection control algorithm is designed.The tracking differentiator is used to arrange the transition process of the controlled system to achieve effective acquisition of differential signals and fast tracking of system instruction signals.The total disturbance is extended as a new state variable of the system,and all the states including the original state variable and the total disturbance are observed by the input and output of the system,and then the nonlinear extended state observer is constructed to estimate the system disturbance in real time.According to the control quantity of system disturbance compensation,the state error feedback control rate is designed to realize the disturbance suppression and improve the control effect,so as to form a complete set of active disturbance rejection control algorithm of wind turbine blade component performance test system.Then,in order to solve the problem that the extended state observer has great influence on the system performance and the parameter setting is difficult in active disturbance rejection control algorithm,BP neural network learning algorithm is used to carry out online self-tuning of the main performance parameters in the extended state observer.At the same time,in order to improve the self-tuning energy of the parameters,the momentum term was added to the weight correction rule to suppress the oscillation phenomenon in the training process of the neural network.The adaptive learning efficiency rule is used to automatically adjust the learning efficiency according to the output error of the neural network to further improve the stability and convergence speed of the neural network.Through the simulation analysis of the active disturbance rejection control algorithm based on the improved BP neural network parameter online self-tuning,the control strategy of the wind turbine blade component performance test system designed in this paper has good stability performance,high control precision and good anti-interference.Finally,the man-machine interaction interface of wind turbine blade wafer component performance test system based on Lab VIEW operating environment was developed,and the field test platform was built to test the performance of wind turbine blade multi-seed components in different methods.After analyzing and processing the test results,it is concluded that the wind turbine blade component performance test system designed in this paper has good operation stability,anti-interference performance and robustness.It can realize the exact tracking of actual load and expected load,and significantly improve the control accuracy and following accuracy of the system.