摘 要: 针对影响通信质量的码间干扰问题,提出利用杂草算法的随机性、鲁棒性、自适应性优化神经网络,为 神经网络提供较好的初始权值,再与BP算法的指导性搜索思想结合起来,既能克服寻优中的盲目性进而避免局部收敛 情况的发生,有效地加快收敛速度,减小剩余稳态误差,降低误码率,从而提高信道的盲均衡性能。通过计算机仿真, 证明该算法具有较好的收敛性能。 |
关键词: 入侵杂草算法;初始权值;盲均衡算法;BP神经网络算法 |
中图分类号: TP391.41
文献标识码: A
|
基金项目: 天津市企业科技特派员项目(18JCTPJC66900). |
|
Research on Invasive Weed Optimization Neural Network Blind Equalization Algorithm |
GENG Yanxiang,WANG Guangyan, ZHANG Liyi1,2
|
1.( 1.School of Electric Information Engineering, Tianjin University, Tianjin 300072, China;2. 2.School of Information Engineering, Tianjin University of Commerce, Tianjin 300134, China)
|
Abstract: In view of the inter symbol interference problem,this paper proposes a new method applying the randomness,robustness and adaptability of Invasive Weed Optimization to optimize neural network.This method can provide good initial weights for neural network.By combining the guiding search idea of BP algorithm,it can not only overcome the blindness in seeking optimization to avoid local convergence,but also effectively speed up the convergence speed,reduce the remaining steady-state error,and lower the symbol error rate,thus improving the blind equalization performance of communication channels.The computer simulation results show that this algorithm has better convergence performance. |
Keywords: Invasive Weed Optimization;initial weight;Blind Equalization;BP neural network algorithm |