| Active/semi-active intelligent control technology can significantly suppress the vibration response of structures,so as to meet higher requirements of safety and function.In general,the control idea is to adopt the traditional centralized control idea of "decentralized collection and centralized processing".However,in complex engineering structures of large scale,complex composition and various functions,traditional centralized control is still employed,which is difficult to meet the requirements of stability,robustness,reliability and maintenance of structural controls.Decentralized control theory of large system is a new way to solve the large-scale structure vibration control.Therefore,based on actual demand,this paper studies the decentralized control of structures deeply,thoroughly and comprehensively,and puts forward a variety of application of decentralized control algorithm for vibration control of civil engineering structure,and provides a theoretical basis for the application of decentralized control theory in civil engineering.The main contents of this paper are as follows:1.Combined with the characteristics of decentralized control of civil engineering structures,the concept of fully decentralized control,partially independent and overlapping decentralized control are introduced,and then local optimal,global optimal decentralized control algorithm,inclusion principle decentralized control algorithm,fuzzy sliding mode decentralized control algorithm and adaptive learning rate RBF neural network sliding mode decentralized control algorithm are presented respectively.Based on linear quadratic optimal control theory,local optimal and global optimal decentralized control algorithms are established respectively through different optimization target of subsystem(local and global optimization).At the same time,this paper proposes overlapping optimal control force selection rules for the optimal overlapping decentralized control system,and uses the hybrid swarm algorithm to optimize the sub-controller,getting stable and ideal control effects of decentralized control system.By Applying the system inclusion principle and the extended-decoupling-contraction(EDC)process,the decoupled gain matrices of controllers and filters are derived,and then various kinds of overlapping decentralized control strategies for complex structures,including chain-shaped,ring-shaped,and star-shaped strategy,etc,are considered.In addition,in order to effectively deal with the influence of the interconnected terms of different subsystems and the uncertainty of the external loads,the Lyapunov function and sliding mode control theory are employed to design the decentralized sliding mode control law which only depends on the displacement and the velocity response of relevant subsystem.A decentralized fuzzy sliding mode control algorithm(DFSMC)is proposed here for dealing with the influence of the interconnected terms of different subsystems and the uncertainty of the external loads.On this basis,combining fuzzy control theory and RBF neural network theory,the decentralized fuzzy sliding mode control algorithm(DFSMC)and the decentralized adaptive learning rate RBF neural network sliding mode control(DALRBFSMC)are respectively designed,which can adjust the switching gain of the sliding mode control law in real time.The effectiveness of the proposed algorithms is verified by numerical simulation.2.A new multi-objectiveoptimization design method for decentralized control system is proposed,in which subsystem division,the parameters of controller and the number and allocation of actuator are synchronously optimized.First of all,optimal installation floors of control devices are determined through structural controllability index;Secondly,the amount of actuators and controller gain for each subsystem are optimized through the multi-objective hybrid swarm optimization algorithm,the non-dominated solution set is achieved based on the dealer principle,the simulated annealing algorithm is used for the secondary local search;meanwhile,the leader selection based on boundary point geometry center is adopted,which meets the requirements of the diversity of the population and the convergence speed,and the two indexes that reflect structural vibration control effect and the performance of control strategy have been used as the optimization objective function for each subsystem.The simulation results show that the proposed method has better universality and can solve the optimization problem of the decentralized control system effectively.3.In view of the disadvantages of decentralized control system in which the decentralized controllers work in parallel and are difficult to carry out effective coordination,the problem of coordinated decentralized control and hierarchical decentralized control for the vibration control of civil engineering structures are studied.The stable coordination PD controller and the stable optimal guaranteed cost PID coordination controllers are designed respectively,combined with extremum control principle,the sub-controller is designed with the goal of whole structural control effect.And then the stable PD coordinated decentralized control algorithm and the guaranteed performance PID decentralized control algorithm are obtained.In addition,according to the hierarchical decentralized control idea,the associated coupling between subsystems is eliminated by setting the global controller.Lyapunov stability theory and RBF neural network theory are employed to design the adaptive control law which only depends on the displacement and the velocity response of relevant subsystem,and the parameters of adaptive RBF neural network local sub-controller is optimized through using the differential evolution(DE)algorithm.And then,the adaptive RBF neural network hierarchical decentralized control(ARBFHDC)algorithm is established for structure vibration control.The applicability of the proposed algorithms is verified by numerical simulation.4.Output feedback guaranteed cost robust decentralized control algorithm with pole constraints and output feedback H? guaranteed cost robust decentralized control algorithm are proposed to deal with the influence of the parameter uncertainties of structures.Firstly,matrix inequalities for the quadratic stability and pole constraints of decentralized control system are given based on LMI method,and the output feedback guaranteed cost robust decentralized control algorithm with pole constraints is established by using variable substitution method.Secondly,on the basis of ensuring the stability of the decentralized control system,sufficient conditions for the existence of the robust H? guaranteed cost decentralized controller is given and proved.Then the output feedback H? guaranteed cost robust decentralized control algorithm is established by using variable substitution methods.Furthermore,by introducing the constraint conditions,the H? guaranteed cost robust controller design is transformed into a convex optimization problem with linear matrix inequality constraints.Numerical simulation results indicate that for the structures with large uncertain parameters,the proposed two output feedback robust decentralized control algorithms have better control effects and robust performance than the traditional LQG centralized control algorithm.5.In view of the disadvantages of structural complex nonlinear and uncertainty,guaranteed cost adaptive RBF neural network nonlinear robust decentralized control algorithm(GCARBF)and adaptive H2/H? nonlinear robust decentralized control algorithm(A-H2/H?)are put forward respectively,combined with linear matrix inequality method,RBF neural network and H2/H? control theory.The corresponding MR semi-active decentralized control system are designed by combining these two algorithms with Clipped-optimal semi-active control algorithm.Firstly,the degenerated Bouc-Wen hysteretic model is utilized to simulate the restoring forces,and the error state equation of the sub-control system is established by considering the uncertainty of the model parameters(mass,stiffness and damping)and the coupling between subsystems.Secondly,sub-controller is designed which composes of guaranteed cost control term and adaptive approximation control term.The guaranteed cost control term is obtained by solving the guaranteed cost control problem which is transformed into a linear matrix inequality.The approximation control term is determined by the adaptive control law of RBF neural network.And then the GCARBF algorithm is established.In addition,based on the linear matrix inequality method,a H2/H? robust decentralized control law is obtained for the H2 control target under a given H? control objective by establishing and solving a convex optimization problem.The adaptive functions are set up to adjust the proportional gain and differential gain,and then the A-H2/H? nonlinear robust decentralized control algorithm is established to realize time varying regulation control.The simulation results show that the proposed nonlinear robust decentralized control algorithms have obvious advantages to nonlinear structures with large uncertainties,and can further reduce the damage of the structure,and improve the safety and stability of the control system.6.In order to solve the problem of fault tolerant control of actuator and sensor faults in decentralized control system,.small gain decentralized stabilization fault tolerant control algorithm and adaptive decentralized intelligent fault tolerant control algorithm are proposed.First of all,considering the correlation of the sub-control system,the sensor failures and the actuator failures,a sufficient condition for decentralized stabilizability is proposed and proved for the class system of discrete-time systems with delay interconnections by using the Lyapunov stability theory and the linear matrix inequality(LMI)approach.Furthermore,the small gain decentralized stabilization fault tolerant control algorithm is proposed based on the small gain principle.Secondly,a direct adaptive decentralized control algorithm with compensation function is designed to deal with the actuator faults and the coupling between subsystems.The adaptive neural fuzzy inference system(ANFIS)is used to diagnose sensor fault and repair fault signal,and then the adaptive decentralized intelligent fault tolerant control algorithm is obtained.Simulation results show the effectiveness and superiority of the proposed fault tolerant control algorithms. |
PDF Full Text Request