QPSO算法优化的非线性观测器设计方法研究

Quantum-behaved Particle Swarm Optimization Based Nonlinear Observer

具有量子行为的粒子群优化算法(Quantum-behavedParticleSwarmOptimization,简称QPSO)是继粒子群优化算法(ParticleSwarmOptimization,简称PSO)后,最新提出的一种新型、高效的进化算法。论文在研究基于PSO算法的非线性观测器基础上,提出了一种基于QPSO算法的非线性观测设计方法。以vanderPol系统为例进行了仿真实验,其基本思想是将非线性连续时间系统的状态估计问题转换为非线性函数的在线优化问题,然后利用PSO或QPSO算法获得系统状态的最优估计。仿真结果显示了基于QPSO算法的非观测器比基于PSO算法的非线性观测器的性能更优越。

Quantum-behaved Particle Swarm Optimization（QPSO for short）,is a new type,efficient swarm intelligence algorithm that proposed lately succeed to Particle Swarm Optimization （PSO for short）.Based on investigating PSO based nonlinear observer,in this paper,a QPSO-based nonlinear observer design method is proposed.Take van der Pol system as example to perform the emulational experiment.The basic idea of the method is that the state estimation of nonlinear continuous-time system is converted into an on-line optimization of nonlinear functions,and then particle swarm optimization algorithm or quantum-behaved particle swarm optimization algorithm is employed to find optimal estimation of the system states.Simulation results show the performance of QPSO based nonlinear observer is superior to PSO based nonlinear observer.