The Study Of Application Of Nonlinear Adaptive Inverse Control In Automatic Flight Control System Of Unmanned Aeroplane

The traditional control theory is combined with the signal processing theory to form a novel method for the control problems-- adaptive inverse control(AIC), which was first presented by professor Widrow of Stanford University in 1986,. In adaptive inverse control method overcomes the problem that requesting for optimal solution of control plant's modeling and precision of modeling as other traditional control methods.In general, adaptive inverse control may be looked as an embranchment of adaptive control theory, because adaptive inverse control theory imports many control methods and design structures of adaptive control theory, such as modeling reference adaptive control system etc. But we should pay attention to difference of their theory foundation. Comparing with adaptive control, methods of designing and analyzing for controller of adaptive inverse control is quite relatively simple, because of adaptive inverse control substituting controller by inverse model of controlled plant. The modeling method of adaptive inverse control was based on the knowledge of adaptive signal processed and adaptive filter. And adaptive inverse control theory was not only easy to understand but also facility application. The result proved adaptive inverse control was a kind of simple excellent robust adaptive control method.The application of adaptive inverse control in linear system is rather mature, and was succeed in practice application, but the application in nonlinear just was seeds, so we have large numbers of work to do. In general, nonlinear system was not corresponding function connection each other, and was not provided with compensate characteristic of linear SISO system. So they have distinct difference in control methods.Unmanned plane is a complex nonlinear dynamic model, the kind of nonlinearity rooted in coupling six of freedom movement and aerodynamics and moment following flight change. In high performance flight of plane, those nonlinear characteristic is very distinct. If adopting conventional control theories and methods to make plane perform complex flight movement, the control system would be very complex design and uneasily controlled. Along with the developing of neural network technology and electronic computer technology, neural network provided with more better function of capability of big scale parallel and processed nonlinearity and processed uncertainty ,and processed the complex process and system of difficult to using model or rule describe.Radial basis function neural network is an effective feedforward network. It has high convergence rate and high approaching precision, and can avoid local optima. Systematic analysis and research are made to the various learning methods of radial basis function neural network. Its virtue advances the appliance superiority, so the study of adaptive inverse control system based on neural network is provided with more importance to application and theory.For adaptive inverse control system, Professor .Widrow and Professor Plett had provided relatively detailed analysis, and provided to profound discuss and compare both the linear and nonlinear adaptive inverse control systems. In different nonlinear systems, it is different that needing solving issues and adoptive methods.In practical project, most systems can not be analyzed of be nonlinear, its inverse model is difficult to set up by way of tradition modeling. A sort for neural network control structure has been put forward here, because of collateral structure of neural network and its great superiority displaying among the automatic controlled field in real-time.So , has studied the modeling way of the neural network and study algorithm of radial basis function neural network on the basis of analyzing the operation principle and characteristic of the control structure in this paper. The reversibility and the stability of neural network inverse system has been analyzed, which have been ready for following simulation.A nonlinear adaptive inverse control system is proposed for possess the output disturbance cancellation ability of the system. Such systems can cancel the disturbance immediately without affecting other adaptive processing. Another nonlinear adaptive inverse control system is proposed in the paper. A set of adaptive LMS filters are used as identifiers and controllers in the systems. The architecture of the systems is fit for a nonlinear adaptive inverse control. Accordingly realized excellent control unmanned plane nonlinear plant, based on adaptive inverse control.It can satisfy the demands of modeling. Theε-filtering LMS algorithm can deal with nonlinear systems and is insensitive to the imprecision of plant model and its delayed inverse model. So we use theε-filtering LMS algorithm in the unmanned plane controller design. The simulation results prove that control scheme is effective.