基于浮点数编码的自适应遗传算法在模型辨识中的应用

掌握被控对象的数学模型是准确分析该对象特性的前提。由于目前被控对象变得越来越复杂，传统的辨识方法在很多情况下已不能胜任。本文提出将自适应的遗传算法（AGA），在融合、改进了其它一些遗传操作的基础上应用于系统辨识。通过分别对理论模型和实际模型的辨识仿真，结果证明本文的自适应遗传算法有较高的精度和较好的鲁棒性。

Adaptive Genetic Algorithm Based on Float-Encoding and Its Application in Model Identification

Mastery of its mathematic model is the premise of analyzing the character of a controlled device exactly. Now, the controlled device becomes more and more complex and the traditional identification algorithms can not be competent for it in many cases. In the paper, an adaptive genetic algorithm (AGA) based on float-encoding is introduced and applied in system identification, merging and improving some other genetic operations. Its application in simulation of identifying a theoretical model and a real model proves the high degree of accuracy and Robust of AGA.