Nonlinear Model Establishment And Comparative Study Of Control Strategies For Fast Steering Mirror

The fast steering mirror is a small precision optical instrument used to control the direction of beam propagation between the target light source and the receiver.With the development of technology,the performance requirements of various fields are getting higher and higher,so the high-performance FSM system becomes more and more important.This paper focuses on the design and performance improvement of a FSM control system.The specific research ideas are as follows: firstly,a controller is designed to satisfy the expected performance index of the system;then,a controller design method which can easily change the system index is explored;finally,how to further improve the system performance indicators.The main works are the following:The first is modeling,which establishes a simplified non-linear model that can describe the system under all operating conditions,i.e.the essential non-linear model of noise + physical upper and lower bounds + linear model.Specific modeling ideas are as follows: firstly,the linear model of the system is established by combining theoretical modeling with experimental testing,and the parameters are corrected;secondly,the upper and lower bounds of physics are simplified to multiple saturation limits and dead-zone limits,which are determined by experimental testing and theoretical derivation,and the actual noise is measured by experiments.Finally,a simplified non-linear model describing the system under all operating conditions is obtained.The second is the design and performance analysis of the controller.Firstly,classical frequency domain method is used to design lead-lag,integral-notch and other controllers.For the primary resonance and accuracy of the system,the design of classical controller shows the difficulty of debugging and the insufficient control performance.Then,the integral augmented state feedback controller is designed by using modern control theory.The state feedback has the advantages of simple adjustment and good control effect,but it also exposes the problems of difficulty and high cost.At the same time,the performance index and robustness of different control systems are analyzed.The final closed-loop system proves that the linear bandwidth can achieve the desired bandwidth of 500 Hz,but due to the limitation of voltage saturation limit,the actual linear operating range of the system will be reduced.It is concluded that the physical limit of the system itself is the main factor limiting the linear working range of the system.Thirdly,a controller design method which is easy to debug and implement is presented,that is,the modern control theory is used to analyze and design the controller,and the equivalent to the classical controller to achieve.This method is easy to debug and realize.By using this method,the integral augmented state feedback equivalent is reduced to the classical controller of integral-lead-notch,which clarifies the essence of state feedback and simplifies the difficulty and cost of implementation of the controller.At the same time,the realization of the controller is given in two ways: Analog Realization and digital realization.The iteration formula of the digital controller is given for the equivalent reduced order controller.And,the analog circuit is designed by Mulitism software and verified by simulation.Finally,the closed-loop verification of the actual system using the analog controller proves the accuracy of the simulation.Fourthly,the coupling problem is analyzed and discussed to further improve the system performance index.Through identification and analysis of coupling mechanism,a coupling model is established.The classical decoupling controller is designed by classical control theory,and the integral augmented state feedback controller of MIMO is designed by modern control theory.It is proved that the above two controllers can effectively suppress the coupling effect.Finally,the robustness of the two controllers is analyzed.