A sewage index prediction method

Abstract

The application comprises the following steps: S1, replacing the hidden layer in the BP neural network with a discrete shear wave function to form an initial BP neural network; S2, optimizing the network structure and network parameters of the initial BP neural network by using an ALO algorithm to obtain a population individual vector with optimal fitness; S3, optimizing the BP-SH neural network structure and parameters by using the obtained individual vector with optimal fitness to obtain an optimal ALO-BP-SH neural network model; and S4, predicting the data of the sewage to be measured by using the optimal BP neural network model and obtaining a prediction result. The application has the beneficial effect that the original neuron hidden layer in the BP neural network is replaced with a shear wave function SH, a multiscale function, compared with the previous invention of replacing the hidden layer with a wavelet function, the BP-SH model proposed in the application increases the shear operation in data processing, so that the hidden layer parameters are more abundant, and in the case of a small training sample, the algorithm can achieve better accuracy.

Images