基于改进量子遗传算法的过程神经元网络训练

Training of process neural networks based on improved quantum genetic algorithm

针对过程神经元网络由于模型参数较多BP算法不易收敛的问题,提出一种基于量子位Bloch坐标的量子遗传算法.将该算法融合于过程神经网络的训练,按权值参数的个数确定量子染色体上的基因数并完成种群编码,通过新的量子旋转门完成个体的更新.算法中的每条染色体携带3条基因链,因此可扩展对解空间的遍历性,加速优化进程.以两组二维三角函数的模式分类问题为例,仿真结果表明该方法不仅收敛速度快,而且寻优能力强.

Aiming at the problem that it is difficult for BP algorithm to converge because of more parameters in training of process neural networks based on orthogonal basis expansion, a solution on the basis of an improved quantum genetic algorithm is proposed in the paper. An improved quantum genetic algorithm based on Bloch coordinates of qubits is proposed, which is integrated into the training of process neural networks. The number of genes on a chromosome is determined by the number of weight parameters and population coding is completed. Individuals in the population are updated by new quantum rotation gate. In this method, each chromosome carries three chains of genes, so can extend ergodicity for solution space and accelerate optimization process. Taking the pattern classification of two groups of two-dimensional trigonometric functions as an example, the simulation results show that the method has not only fast convergence but also good optimization ability.