基于LSTM和Informer变体的渔业水质预测研究

Research on Fishery Water Quality Prediction Based on LSTM and Informer Variants

现代化渔业养殖需要更快更准确的水质预测,提升养殖水质预测的精度对提升渔业养殖的效率具有重要意义,为此提出一种结合随机森林、SG滤波器、LSTM和Informer变体的渔业水质预测模型。首先通过随机森林算法填补数据中的缺失值,通过SG滤波器减少噪声干扰。其次将LSTM结合静态和瞬时递归网络作为Informer中的内部结构,将数据送入模型中。最后通过全连接层输出得到水质预测结果,对渔场内监测点的水温、pH值、溶解氧进行预测。结果表明所提方法提高了渔业养殖水质预测的预测精度。

Modern aquaculture requires faster and more accurate water quality prediction,so improving the accuracy of aquaculture water quality prediction is of great significance for improving the efficiency of aquaculture.A fishery water quality prediction model combining random forest,SG filter,Informer,and LSTM variants is proposed for this purpose.Firstly,fill in missing values in the data using the random forest algorithm,and then reduce noise interference through the SG filter.Combining LSTM with static real-time recursive network as the internal structure of Informer,data is fed into the model,and finally water quality prediction results are obtained through fully connected layer output.Predict the water temperature,pH value,and dissolved oxygen at monitoring points within the fishing ground.The results indicate that the proposed method improves the prediction accuracy of water quality in aquaculture.