基于BFA-ELMAN神经网络的微生物生长预测系统

Microbial Growth Prediction System Based on Improved BFA-ELMAN Neural Network

ELMAN神经网络在处理时序数列问题上具有良好的应用,但是其网络结构有时候并不能合适地处理实际微生物生长问题。通过改进ELMAN网络拓扑结构,以及基于细菌觅食算法对神经网络进行权值优化,有效地提高ELMAN神经网络在处理高维度问题的收敛速率以及微生物生长模型的预测能力。

ELMAN neural network has a good application in dealing with the sequence of time series, but its network structure sometimes cannot properly deal with the problem of actual microbial growth. Through improving the ELMAN network topology, and based on bacterial foraging algorithm for neural network weights optimization, effectively improves the convergence rate of the ELMAN neural network in dealing with high-dimensional problems and the prediction ability of the microbial growth model.