基于MEC-BP神经网络在水产养殖水质预测中的应用

摘要

溶解氧作为水产养殖中最为重要且最容易控制的水质参数,其关系到养殖的成败,如果能精确掌握溶解氧的变化规律,可大大降低养殖风险,增加养殖成功率。本文综合考虑相关水质参数,建立BP神经网络水质预测模型,并在此基础上,构建基于思维进化算法( Mind Evolutionary Computation,MEC)的BP神经网络水质预测模型,通过对广西茅尾海海域的水质历史数据进行仿真实验,结果表明,思维进化BP神经网络预测值的精确度和准确度要高于BP神经网络。因此,将该算法应用于水产养殖水质预测是可行的。%Dissolved oxygen as the most important and the most easily control water quality parameter of aquaculture relates to the success or failure of aquaculture. If we can accurately grasp the change rule of dissolved oxygen, the risk of breeding will greatly reduce and the breeding success rate will increase. In this paper, considering the related water quality parameters, we established the BP neural network model of water quality prediction. And on this basis, the BP neural network model based on mind evolu-tionary algorithm ( MEC) of water quality prediction was established. Through the simulation of historical data of water quality of Maowei Sea, Guangxi, the results show that the precision and accuracy of predicted value of MEC-BP neural network is higher than that of the BP neural network. Therefore, it is feasible to apply this algorithm to predict water quality of aquaculture.