Support Vector Regression Using Particle Swarm Optimization for Dissolved Oxygen Concentration

To improve the precision and efficiency of dissolved oxygen, a particle swarm optimization (PSO) of support vector regression (SVR) method was proposed to predict the concentration of dissolved oxygen in the soft measurement model because of the problem of the dissolved oxygen concentration cannot be measured online in the wastewater treatment. The particle swarm algorithm was used to optimize the parameters of support vector regression, and the construction of dissolved oxygen and its impact factors in nonlinear soft sensor prediction model were automatically obtained by the best combination of parameters by using the model of dissolved oxygen concentration on the international BSM1 benchmark simulation model. The simulation shows that the prediction model has good prediction effect compared with SVR, RBF neural network model, the PSO-SVR model not only has low computational complexity, but also has the advantages of fast convergence, high forecast precision, strong generalization ability etc.