Optimal Design And Simulation Of FIR Filter Based On Genetic Algorithm

In the digital signal processing, filter is very important, it is useful in voice processing, image processing and spectrum analysis, so researchers and engineers have been paying more attention to its optimal design. Genetic algorithm(GA) is a kind of searching method which simulates the natural evolution, it is simple and easy to implement it, and it does not need the special field knowledge, so it has been used widely. Applying GA to design filter, and using FPGA to implement it are one of the current focus on digital processing research field.Firstly in this dissertation, the principle of GA is introduced, the specific schemes of binary coded genetic operation and real coded genetic operation are elaborated in detail. On this basis, aiming at the optimization design features of FIR digital filter, we apply zooming genetic algorithm (ZGA) to design FIR digital filter for the first time, compared with standard genetic algorithm (SGA), the examples confirm ZGA has higher probability of the optimal solution convergence and better quality solution. According to FIR digital filter optimization design, the scheme using improved real coded genetic algorithm is presented again, through improving adaptive mutation operator to narrow range of algorithm mutation and enhance the ability of local search, then using examples to confirm its efficiency.Finally, basing on FPGA to design the FIR digital filter, with VHDL language and Megawizard as the input ways, a 24 order low-pass FIR filter is designed, compared with the MATLAB software calculation results, the design show correctly.