遗传算法在分类器系统中的应用

The Application of Genetic Algorithum in Classifier System

针对数字式车牌图象识别系统中线性分类器参数辨识这一具体问题,对遗传算法(简称GA)过早收敛的缺点提出了若干改进意见并作出性能分析,说明了自适应参数调整和染色体并行进化相结合的GA方法是有效的.

Based on the issue of Linear Classifier (LC) parameter identification for the image recognition system of digital vehicle ID plate, the author of the article analyzes the shortcoming of too early convergence for canonical genetic algorithum (GA), and makes some madification. Performance analysis inicates that the GA method of adaptive parameter adjustment combined wirh chromosome parallel evolution is available.