Design Of Digital Filters Based On Particle Swarm Optimization Algorithm

Nowadays, a filter, which is applied extensively in many fields, e.g. communication system, speech processing, image processing, automatic control, radar, military application, aviation, medical treatment and household electrical appliance, is indispensable in the fields of digital signal processing and electronic application technique and is more and more valuable with application.In the design of classic filter, complicated structure and large order are required to satisfy higher precision with several technique demands, and sometimes can fail. As the development of computer technology, digital filter with easy large-scale integration is proposed to achieve high precision and good flexibility compared with analog filter. In the recent 20 years, the fast development of computer technology makes it possible for the design of digital filter. However, as the increase of demand for signal processing, the design of modern filter requires more computation complexity. It is therefore of interest that approaching the several technique demands for digital filter by the help of computer aided design.In 1995, based on the suggestion of birds, fishes and mankind social behavior regulation, particle swarm optimization (PSO) algorithm is proposed by Kennedy and Evberhart, which is a kind of evolvement calculation technique base on Swarm Intelligence. The algorithm imitates the behavior of birds for food, and achieves optimization by collective cooperation of bird The PSO algorithm, being similar with genetic algorithm, is a computational intelligence method base on iterative of community, with no cross and variation arithmetic operator, that the community follows the optimization particle from the solution space, but without cross and variation arithmetic operators. The PSO algorithm has the advantages of fast convergence, little parameter for setting and easy implement, which is suitable for scientific research and engineering application with deep background of intelligence. Therefore, under the case of some optimization rule, the PSO algorithm can be applied to make it optimal for the performance of design for digital filter. It is indicated by simulations that when the optimal design method based on the PSO is effective, when applied to the optimal design of the impulse finite impulse response (IIR) and the infinite impulse response (FIR).This paper discusses the design method of digital filter based on the PSO algorithm. The improved PSO algorithm is proposed to implement the optimal design of FIR filter and IIR filter under the prescribed design demand with a better simulation result. This paper is organized as follows.(1) The technique problem and optimal design method of digital filter are first discussed.(2) The parameter is a key factor of affecting the performance and efficiency of the PSO algorithm. The choice of parameter for the PSO algorithm is then discussed in detail. The principle of parameter selection and a kind of modified algorithm based on inertial weight adaptive particle swarm optimization (PSOAIW) are proposed to improve the search ability and convergence speed for the algorithm.(3) The PSOAIW algorithm is used to design FIR filter and IIR filter in the third part, and the simulation results are then analyzed.(4) Summarization and expectation of this researches are presented finally.Based on the research results of this paper, the design method based on the PSO algorithm achieves a better performance for filter deign. The adaptive particle swarm optimization (APSO), not only has good global search ability, but also has fast convergence speed and high accuracy. The PSO algorithm with good applicability can be used to other optimal problem.