A kind of coagulant dosing control model based on ISFLA-SVM

Coagulant dosing is a key link in water treatment. Coagulation process mechanism is complex, the coagulant dosing process is large time delay, nonlinear and difficult to control. Traditional control method effect is poor, effluent turbidity is fluctuations. This Paper proposed an coagulant dosage support vector machine (SVM) control model optimized by improved shuffled frog leaping algorithm (ISFLA). The shuffled frog leaping algorithm was improved with Gauss mutation and chaos perturbation method and the optimal parameters of SVM model was searched by the ISFLA. The proposed model could effectively improved the convergence speed and the precision of SVM. The simulation results show that, the model prediction was accuracy, can accurately determine the coagulant dosage.