基于改进量子遗传算法的小波神经网络优化及其软测量应用

Application of Wavelet Neural Networks Based on Improved Quantum Genetic Algorithm in Soft Sensor Modeling

针对简单量子遗传算法在优化高维问题寻优速度慢、收敛率低的缺陷,提出一种改进的量子遗传算法,通过搜索各种群中最优染色体组成当前最优个体,并依此个体来确定量子门的全局最优搜索方向。将改进算法用于优化小波神经网络,藉此建立了4-CBA浓度的软测量模型。仿真结果表明:与简单量子遗传算法相比,改进算法对复杂优化问题具有全局快速寻优性能。

Improvements were made for quantum genetic algorithm to cope with the defect of optimization rate and inefficient convergence. By searching the excellent genome of all colonies, the best unit is built. And the optimal search direction was decided according to the unit. The improved algorithm is used to optimize wavelet neural network, and then is applied in the soft sensor models of 4-CBA concentration. The simulation result of complex optimal problem indicates that the improved algorithm has better performance of optimization rate and efficiency.