The Research Of Design For FIR Digital Filter Based On Partical Swarm Optimization Algorithm

Digital Filter play a very important role in the applications of digital signal processing.In a certain symmetry conditions,the linear phase can be achieved,and good stability,and thus FIR digital filter access to a wide range of its application.Its optimized design method has been paid close attention by most of scholars and engineers.PSO algorithm simulated the foraging behavior of flying birds.Groups search following the best particles according to their speed in the solution space.PSO algorithm has the advantages of fast convergence,less set of parameters and easy to achieve,and has outstanding performance in addressing non-linear optimization problem,which lead pso algorithm to become an important tool for optimization.Based on particle swarm optimization algorithm theory,PSO algorithm is used to optimize design the FIR digital filter in this paper,that is,combining with the advantages of PSO algorithm and the design requirements of FIR digital filter, we can find better filter coefficients with the help of PSO algorithm,in order to approach the frequency response of engineering requirements.Firstly,The paper introduces the principle of PSO algorithm,mathematical description,the meaning of the parameters and various models.Secondly,the design theory and methods of the FIR digital filter have been analyzed.Under the specified technical indicators,a high-pass filter has been designed in the application of PSO algorithm,which is simulated in the environment of Matlab6.5. The results show that the high-pass filter has better filtering effect such as smaller pass-band ripple and bigger stop-band attenuation than which designed by conditional optimization algorithm.Finally,the paper studies the influence on PSO algorithm parameters on performance.To the basic ideas of easy understand and simple achieve to the algorithm,for the phenomenon of nonlinear and highly complex in the actual search process for particles and from the change in the parameters of PSO,we want that the inertia weight changes with the Running posture of the particles.So a particle swarm optimization algorithm which dynamic change in the inertia weight is given. The designed high-pass filter using this algorithm has faster convergence and more robust from the performance indicators of on-line performance,off-line performance and the optimal fitness each generation of the group.