Font Size: a A A

Adaptive Quantum-behaved Particle Swarm Optimization Algorithm With Mutation Operator And Application

Posted on:2009-02-24Degree:MasterType:Thesis
Country:ChinaCandidate:W H JinFull Text:PDF
GTID:2178360272957293Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
Particle swarm optimization (PSO) is introduced by Kennedy and Eberhart in 1995 [1], it is a population based evolutionary optimization and inspired by the collective behaviors of birds. It has already been shown that PSO has ever since turned out to be a competitor in performance with other evolutionary algorithms. The loss of diversity in the population is inevitable due to the collectiveness, During the latter search period, the particles are investigated to cluster together and its search area is so limited that the whole swarm is very possible to be trapped in a local minima. Quantum-behaved Particle Swarm Optimization (QPSO) is a novel PSO algorithm model in terms of quantum mechanics. The model is based on Delta potential and think the particle had the behavior of quanta. Based on QPSO algorithm ,the mechanism of mutating is proposed to escape local optima. mutating the individuals in adaptive mutating probability, and increasing its diversity of the swarm, At the same time, using firstly the strategy of remaining cadreman and establishing cadreman sequence warehouse in order to avoid the loss of the optimum value, An Adaptive Quantum-behaved Particle Swarm Optimization algorithm with mutation operator (AMQPSO) was proposed.System identification is a theory and method studied how to establish mathematics model in produce process. Although there are many means, they have various limits. The AMQPSO was used to identify the system parameters. The problems on solving nonlinear equations is transformed into that of function optimization. The AMQPSO was used to solve systems of nonlinear equations. How to determine parameters of fuzzy production rules is significant for building a fuzzy Petri net, it is one of research hotspots, also. A hybrid algorithm ABHA that takes full advantages of AMQPSO and BP is originally introduced into the procedure of exploring the parameters of FPN. In the paper, our main work has been done as follows:1 Analyze and tally up QPSO and AMQPSO.2 The AMQPSO Application in system parameter Identification .3 The AMQPSO Application in sloving Nonlinear system of equations.4 The ABHA was introduced into the procedure of exploring the parameters of FPN.In the paper, the AMQPSO was applied to identify system parameter. The identification results show that this method has the capability of overcoming the limits traditional identifying algorithm, the advantages of high parameter identification precision, strong ability of resistance to the noise, good input signal generality and identification of the nonlinear system, so it has important practical values. The AMQPSO algorithm is used for solving problem for nonlinear systems of equations , Numerical results show the feasibility and efficiency of this method.The ABHA was originally introduced into the procedure of exploring the parameters of FPN. Simulated experiment showed that this hybrid algorithm has easier computation, more rapid convergence compared with other traditional learning algorithms, reduce number of training, its global convergence ability is better, the trained parameters gained from above algorithm were highly accurate and the resultant FPN model owned strong generalizing capability and self-adjusting purpose.
Keywords/Search Tags:Adaptive Quantum-behaved Particle Swarm Optimization algorithm with mutation operator, system identification, BP neural networks learning algorithm, nonlinear system, FPN, production rule, fuzzy reasoning, Nonlinear system of equations
PDF Full Text Request
Related items