混合PSO算法在非线性控制系统和电路参数优化中的应用研究-电路与系统专业论文.docxVIP

中南大学硕士学位论文Parameter 中南大学硕士学位论文 Parameter optimization 15 one ofthe most important research objects in many scientific，e咄ecring problems and social economic activities． Currently,scholars have proposed a large number of optimization algorithms and Particle Swarm Optimization(PsO)is a novel,advanced evolutionary optimization algorithm among them．PSO algorithm has been applied to many engineering problems successfully and achieves good results． Auto-adaptive inertia weight PSO with mutation(PSOGA) algorithm and MPSODE algorithm with nonlinear inertia weight which was combined by differential evolutionary algorithm∞E)and PSO were proposed in this paper．Those proposed algorithms aim to increase diversit)，of particle swarm and avoid evolving into local optimal point． PSOGA,M瞳SODE and PSO were encoded in real．which can avoid potential problems．such as initial value sensitive and non-derivative, when using traditional algorithms to solve nonlinear problems．MPSODE, PSOGA and PSO algorithms were used on parameter optimization of semiconductor,typical instances in circuit and control system in order to certify feasibility,effectiveness and advantage of the novel idea and proposed hybrid optimization algorithm in application． Linear mathematic models of semiconductors,circuits and control system’S nonlinear parts were constructed in this paper at first．Then suitable parameters of mathematic models for optimization WeIe extracted and felicitous fitness functions ofthose models were established, which reflected the capability of parameter combination accurately． Under reasonable design flOW,an efficient parallel search in the parameters target space were implemented to obtain the optimal parameter by utilizing the excellent search performance and faSt convergence speed ofPSO and hybrid optimization algodthms proposed． Simulation experiments wore used to certify that,compared with orther algorithms，the proposed hybrid algorithms call receive more satisfied results of semiconducto