Research On Swarm Intelligence Optimization Algorithm And Its Application In System Identification Of Complex Nonlinear Systems

The swarm intelligent optimization algorithm has the advantages of solving complex nonlinear system identification problems,which simulates the behavior of foraging characteristic such as self-organization,self-learning ability and simple structure and easy implementation in nature.In this dissertation,the swarm intelligent optimization algorithm has been deeply studied.A series of work of this dissertation is launched from algorithm improvement,performance evaluation to comparative analysis,and its related application for the parameter identification and controller parameter tuning in complex nonlinear systems is studied in this dissertations.The main contributions are summarized as follows:(1)An improved particle swarm optimization(PSO)algorithm is proposed and the parameter identification of Cole impedance model with fractional-order characteristics is realized.The particle swarm optimization algorithm is improved from three aspects: Firstly,the velocity update strategy is improved,and the information interaction between individuals is fully utilized to improve the algorithm search ability.Then,in order to avoid the occurrence of extreme values and prevent the algorithm from falling into local convergence,an improved Tent mapping is used to process the particle position information.Besides,inspired by the variation idea of genetic algorithm,a piecewise mutation probability operator is designed to maintain the diversity of the population.The performance of the proposed algorithm is verified by the standard functions.The experimental results based on the double-dispersion Cole model parameter identification show that the proposed algorithm is an effective and high-precision identification method.(2)Based on a multi-strategy mechanism,an improved quantum bacterial foraging optimization algorithm is proposed to solve the system identification problem in the case of fractional-order known system structure and unknown system structure.Three strategies are proposed in the improved algorithm.Firstly,in the initialization process,in order to prevent the solution remaining unchanged,the concept of optimal rotation angle correction probability is proposed.Secondly,the probability amplitude correction operator is introduced to enrich the population diversity.Thirdly,an exponentially varying nonlinear quantum rotation angle is designed to make the rotation angle adaptively and continuously updated,which improves the global search ability of the algorithm.The performance of the proposed algorithm is tested by standard functions.The test results show that the improved algorithm can achieve better optimization accuracy with relatively fast convergence speed.In order to verify the effectiveness of the algorithm,the proposed algorithm is used for parameter identification of fractional-order systems.Simulation results of both known and unknown system structures show that the improved quantum bacterial foraging optimization algorithm can effectively identify the structural parameters of the system while maintaining a fast convergence rate.(3)An improved bacterial foraging optimization algorithm is proposed,and the parameter tuning of the integer-order PID controller in the servo system is realized.In order to improve the ability of classical bacterial foraging optimization algorithm in dealing with complex problems,a new bacterial foraging optimization algorithm is proposed based on three improvement strategies.In the chemotaxis operation,a nonlinear adaptive chemotaxis step is designed,which effectively balances the global search ability and local search ability.The reproduction operation is modified to improve the global search ability.Then,the mutation operation is applied to compensate for the lack of population diversity caused by the reproduction operation.The modified elimination-dispersal probability improves the search efficiency of the algorithm.Test results on the benchmark function verify the effectiveness of the presented algorithm.The three-closed-loop servo system is modeled in Matlab/Simulink,and the parameters of the position loop integer-order PID controller in the servo system are tuned using the proposed algorithm.The tuning results show that the proposed algorithm has obvious advantages in solving the parameters tuning problem of the integer-order PID controller in servo system.(4)A quantum bacterial foraging optimization algorithm with nonlinear adaptive rotation angle is proposed,and the parameter tuning of fractional-order PID controller in servo system is realized.In the chemotaxis operation,a nonlinear dynamic adaptive rotation angle is designed to accelerate the convergence speed and improve the optimization efficiency of the algorithm.The convergence of the algorithm is analyzed.The performance test based on the benchmark function verifies the effectiveness of the algorithm.The proposed algorithm is applied to tune the parameters of the position loop controller of servo control system.The tuning results show that the algorithm can effectively tune the parameters of the fractional-order PID controller.(5)To solve the problem of stator winding short-circuit fault of PMSM,a motor fault diagnosis method is proposed based on parameter identification.The application of parameter identification in motor fault diagnosis is studied.Based on an improved quantum bacterial foraging optimization algorithm for parameter identification,an effective motor fault diagnosis method is proposed to solve the short-circuit fault of the stator winding in the PMSM system.In the proposed parameter identification strategy,the coding mode is changed from binary to decimal,which improves the efficiency of the algorithm.In order to ensure the feasibility of the algorithm,the concept of optimal rotation angle correction probability is proposed in the update phase.The optimization design of quantum rotation angle also improves the global searching ability.The classical test functions verify the effectiveness of the proposed algorithm.The results of PMSM stator winding short-circuit fault diagnosis based on parameter identification,as well as the short circuit fault analysis of motor stator winding based on finite element method show that not only the relevant stator winding fault can be detected by the relative change of PMSM electrical parameters,but also the severity of the fault can be seen from the variation of the parameter amplitude.