Aquaculture water quality dissolved oxygen prediction method based on data fusion

The invention relates to an aquaculture water quality dissolved oxygen prediction method based on data fusion, mainly comprising the steps of S1, obtaining a plurality of dissolved oxygen values and meteorological data, and performing normalization processing on the dissolved oxygen values; S2, processing normalized data by an RBF network optimized through a K-clustering algorithm to obtain more accurate dissolved oxygen values; S3, re-combining the more accurate dissolved oxygen values with the meteorological data in S1 to form a sample prediction training sample, utilizing the training sample to train a least squares support vector machine model, and obtaining an optimal least squares support vector machine prediction model; and S4, collecting the dissolved oxygen values and meteorological environment data of the aquaculture ecological environment on line in real time, and utilizing the optimal least square support vector machine prediction model to obtain a dissolved oxygen predicted value. The method can