Active Vibration Control Of Intelligent Euler-bernoulli Beam Based On Fuzzy RBF Neural Network

As the aerospace industry develops,the vibration of spacecraft in orbit is inevitable.In terms of spacecraft body structure,the space structures such as solar panels,space robotic arms and spacecraft infrastructure have characteristics of low stiffness and small damping.These factors have caused spacecraft to be easily interfered inside and outside and so to vibrate and the normal operation of spacecraft is affected.Therefore,real-time vibration control of aerospace equipment and devices has become a key technical issue to be solved urgently in the field of aerospace engineering.This paper mainly studies the vibration and control of the flexible structure under external force,and the flexible cantilever beam is used as a control object and the Polyvinylidene Fluoride(PDVF)is used as a sensor and actuators to analyze the control problem of vibration.First,the finite element analysis software ANSYS APDL is used to do modeling,static,modal,buckling and transient analysis of the cantilever beam structure.The response curve of the structure and the natural frequency and vibration mode of the piezoelectric beam before active control are obtained,and then the best paste position of the PDVF is determined by vibration mode curve.Second,the basic performance of PVDF and the characteristics of the positive and inverse piezoelectric effect are analyzed.On this basis,the sensor equation of the piezoelectric sensor and the actuating equation of the piezoelectric actuator are derived.The coupling relationship between the piezoelectric sheet and the cantilever beam is analyzed,and the electromechanical coupling kinetics model of the cantilever beam and the state-space equation based on the closed-loop control system are established.The mass matrix and the stiffness matrix obtained by the ANSYS are introduced into the state-space equation,so the mathematical model of the output feedback of the piezoelectric beam is derived.Third,eight fuzzy controllers with different membership degrees are designed and analyzed according to fuzzy controller design method and fuzzy control basic theory.The active control simulation model of the cantilever beam is established by MATLAB/Simulink,and these eight fuzzy controllers are simulated and the control effects of fuzzy control method and Proportion Integration Differentiation(PID)are compared.Based on that fuzzy controller Gaussian membership function controller has the smallest steady-state error,but does not have self-learning ability and cannot automatically adapt to the change of the external environment,the fuzzy controller and Radial Basis Function(RBF)neural network is integrated.Aimed to the shortages of traditional fuzzy RBF neural network processing multidimensional data and excessive depends on the selection of implicit layer data centers,this paper proposes an improved Salp Swarm Algorithm(SSA)for the optimization of fuzzy radial basis neural network to design a cantilever beam active controller.The SSA that is improved from population initialization strategy of chaotic mapping,leader position update strategy of crazy operator,and follower position update strategy of elite retention and dynamic inertia weights is used to optimize the fuzzy radial basis neural network weights.At last,MATLAB is used to test the vibration control effect of the piezoelectric cantilever beam by the fuzzy RBF neural network control system before and after improving SSA,and the control effect of each algorithm in vibration suppression can be obtained by simulation.At the same time,the simulation results show that the application of improved fuzzy radial nerve network controller can effectively improve the vibration effect of active control,and compared to the original control method of SSA,the system vibration suppression has increased by 55%on average and the control performance is more effective.