Uncertain Cascade Systems Control Based On Backstepping

With the development of software and hardware technology,more and more industrial equipment is being upgraded intelligently.In most practical industrial scenarios,this feature is common in actual scenarios,such as uncertainty,cascaded,nonlinearity.Therefore,this type of system is classified as an uncertain cascaded nonlinear system,and an uncertain cascaded nonlinear system Control research has always been a major field so far.Taking these requirements of the actual scene into account,it is an important research direction to use advanced control algorithms to improve the control quality of the uncertain cascade system.This paper proposes to use the RBF neural network approximation theory to solve the problem of unmeasurable parameters in the mathematical model of the actual scene,so as to obtain an accurate mathematical model.Furthermore,based on the established precise mathematical model,the controller is designed using Backstepping adaptive control method to improve the anti-interference ability and control accuracy of this type of system.This paper selects two uncertain cascade systems in actual scenarios to carry out relevant research,namely the power plant superheated steam temperature system and the four-rotor unmanned aerial vehicle under wind disturbance conditions,and focuses on the uncertain parameter prediction,mechanism model and controller design of the two actual systems.The main research contents and conclusions of this paper are as follows:1)Due to the uncertain or unmeasurable signal in the uncertain cascade system,this signal has an extremely important influence on the stability of the system.Since most of the uncertain parameters belong to the output parameters of the coupled system,it is difficult to model the coupled system.Therefore,the uncertain parameters of the two actual systems are analyzed to determine the key meteorological factors that affect the wind speed,such as daily average temperature,average air pressure,average humidity,and date;key parameters that directly affect the heat release on the flue gas side are determined,such as Power generation load,main steam flow,main steam temperature,etc.Using the RBF neural network method,the neural network models of the heat release on the flue gas side and the wind speed of the wind field were established respectively.The results of the identification experiment and the actual results were compared and analyzed.The two models obtained can well reflect the characteristics of the system and solve the key problem of unmeasurable and uncertain parameters.2)In the actual system,the mechanism model method of each actual scenarios is different.Therefore,according to the characteristics of the two actual systems,the actual needs of controller design and the requirements of algorithm verification,this paper adopts detailed modeling methods to establish the superheated steam temperature system mechanism model base on export fluid parameter as the lumped parameter and the flight dynamics mechanism model of the UAV under the disturbance of the wind field,The mechanism models of two practical application scenarios are summarized as uncertain cascade systems.The above work provides a strong theoretical basis and simulation verification platform for subsequent controller design and simulation research.3)Although the mechanism model of the actual scene can be attributed to the uncertain cascade system,due to the needs of the controller design,the mathematical model is converted into a strictly feedback affine form of parameters.From the analysis of the two system models,it can be seen that the mechanism model of the superheated steam temperature system in this paper is a non-affine form,while the rotorcraft belongs to a strict parameter feedback system.Therefore,in order to facilitate the design of the controller,the output of the desuperheater,that is,the superheated steam temperature,is converted into the superheated steam temperature enthalpy.At this time,the mathematical model of the superheated steam temperature system is converted into a strict feedback form.Considering that the uncertain parameters estimated by the RBF neural network will still have errors in the actual scene,according to the mathematical model of the affine form,the backstepping adaptive controllers and the uncertain parameter adaptive update rate of the two systems are designed respectively.The simulation experiment proves that the control effect of this method is more suitable for the application in the actual scene than the backstepping control,and it has better robustness.