CAJ | 학술논문

基于遗传算法的原理及性能分析,为解决有源噪声控制中的非线性问题,分别设计了标准遗传算法和多种群遗传算法的自适应有源消声仿真实验系统,并对实验结果进行对比分析.结果表明:这两种遗传算法都能够快速找到自适应滤波器最优权系数组合,且由于其优化过程中不依赖于梯度,具有很强的鲁棒性和全局搜索能力,但多种群遗传算法能够更加有效的维持种群的多样性,兼顾算法的局部搜索能力和降低计算结果对遗传控制参数的敏感性,从而有效解决了采用标准遗传算法控制自适应滤波器系数更新过程中出现早熟收敛的问题,为遗传算法在有源消声系统中的实际应用奠定了一定的理论基础.
기우유전산법적원리급성능분석,위해결유원조성공제중적비선성문제,분별설계료표준유전산법화다충군유전산법적자괄응유원소성방진실험계통,병대실험결과진행대비분석.결과표명:저량충유전산법도능구쾌속조도자괄응려파기최우권계수조합,차유우기우화과정중불의뢰우제도,구유흔강적로봉성화전국수색능력,단다충군유전산법능구경가유효적유지충군적다양성,겸고산법적국부수색능력화강저계산결과대유전공제삼수적민감성,종이유효해결료채용표준유전산법공제자괄응려파기계수경신과정중출현조숙수렴적문제,위유전산법재유원소성계통중적실제응용전정료일정적이론기출.
In order to solve the nonlinear problem of active noise control, it designed the adaptive noise reduction system of standard genetic algorithm and multiple population genetic algorithm based on the principle and performance analysis of genetic algorithm. The simulation experiment results show that the two kinds of genetic algorithm can quickly find the optimal combination of weight function in adaptive filter and have a strong robustness and global search ability in the optimization process because of not dependent on the gradient. But the multiple population genetic algorithm can maintain the diversity of population effectively, both the local search ability of the algorithm and reduce the sensitivity of genetic control parameters to the results, and solved the premature problem effectively in the process of updating the adaptive filter coefficient based on the standard genetic algorithm. This conclusion provides a theoretical basis for the genetic algorithm used in the active noise control system.