abstract |
The invention discloses an intelligent prediction method for total phosphorus in effluent based on heterogeneous migration learning, comprising: acquiring a historical data set; establishing an initial total phosphorus intelligent prediction model based on a fuzzy neural network; The phosphorus intelligent prediction model is trained; the parameter knowledge is extracted from the trained initial total phosphorus intelligent prediction model, and based on the parameter knowledge, the Lagrangian algorithm is used to obtain the mapping knowledge; based on the mapping knowledge and historical data sets, the initial total phosphorus intelligence is optimized Predict the parameters of the model to obtain the trained total phosphorus intelligent prediction model; obtain the current input data set, and apply the trained total phosphorus intelligent prediction model for prediction to obtain the predicted effluent total phosphorus. The invention corrects the parameter knowledge to obtain the mapping knowledge, uses the mapping knowledge and historical data to complete the adjustment of the parameters of the total phosphorus intelligent prediction model, that is, the current model parameters, and realizes the accurate prediction of the total phosphorus in the effluent. |