Study On Some Technology Of Intelligent Signal Processing And Its Application

The technology of intelligent signal processing that is the integrated applications including the modern signal processing, artificial neural network obscure systems theory, and evolutionary computing and artificial intelligence theory and methods has become a research hotpot in recent years. And evolutionary computing as an important smart signal processing technology branch has become a dramatic direction.While in research of evolutionary computing, the particle swarm optimization, artificial fish flock optimization algorithm and quantum evolutionary algorithm, have become the research hotpot for its excellent performance. However, because this relatively short time for the three algorithms, there are so many problems to be resolved. Enhancing their ability to solve discrete optimization problems is a worthwhile research direction.To enhance the capacity of PSO and QEA, the paper presented two improvement algorithms. Inspired by the idea of hybrid optimization algorithms, this paper proposes two hybrid quantum evolutionary algorithms (QEA) based on combining QEA with particle swarm optimization (PSO) to improve the performance of QEA called PSEQEA and QBPSO. The experiment results of the knapsack problem, the function optimization problems and multiuser detection problem which is an important signal processing problem show that the both of the proposed methods not only have simpler algorithm structure, but also perform better than conventional QEA and PSO in terms of ability of global optimum.The AFSA is proposed to design the IIR digital filters. And the simulation results proved the effectiveness of the proposed method. Compared with particle swarm algorithm (PSO), AFSA has a good global astringency and rapid convergence ability. Inspired by the idea of adaptive, two adaptive artificial school algorithms called AAFSA_FP and AAFSA_SP is proposed. Then we apply the new algorithms to solve the multiuser detection problems. Simulation results show that the proposed detectors outperform GA detector and PSO detector in terms of BER, near-far resistant and convergence performance.