水质预测的因素状态网络模型研究

Study on factor state artificial neural network model for water quality prediction

通过对因素神经网络理论的研究,提出了因素状态BP网络;通过将信息扩散原理和落影技术结合,形成信息扩散式落影,并与因素状态BP网络有机结合,解决了知识非完备性的问题以及由于训练样本有矛盾样点而使平凡BP因素网络无法训练出结果的问题.以此建立的水质预测的因素状态人工神经网络模型应用于丹江口水库,获得了令人满意的结果.

WT5”BZ]Research on factor neural network theory led to factor state BP artificial neural network. Combination of informative diffusion with falling shadow technique forms information diffusion falling shadow technique together with factor BP artificial neural network for solving the problem of insufficient knowledge, of non countable samples which result in slow or idle BP artificial neural network. Finally, the factor state BP artificial neural network of quality prediction is used on Danjiangkou resevoir and results are very satisfactory.