Research On Predictive Controlof Wastewater Based On Particle Swarm Optimization

Activated sludge treatment method is used to remove pollutants in wastewater using activated sludge MLSS, the most widely used today is the sequence of activated sludge treatment method (SBR). Compared to the traditional treatment method, SBR method are intermittent in time and space, oxygen is needed in the process of microbial growth and degradation of pollutants, therefore the dissolved oxygen concentration (DO) directly affect the treatment effects, it is generally believed that the sewage treatment works best when DO is 2.00mg / L, so in this paper DO was chosen as the Control volume in the SBR wastewater treatment system and the output DO value has been maintained at 2.00mg / L.Firstly, according to the SBR wastewater treatment process mathematical modeling was set our and proved effective. PID controller and generalized predictive controller were set up by mechanism model and then compare the effect of simulation. Generalized predictive control (GPC) has advantages of predictive control and adaptive control. the advantages is Multi step prediction and online identification function from the effect of two controller, GPC results is better than PID. Output is more stable, and has less overshoots .the results show that the GPC have better control performance and can achieve the purpose of energy-saving to the Wastewater Treatment system which have the feature of nonlinear and Strong oupling.Generally predictive Control performance has reached the standard of basic control. However, in order to improve the accuracy of the output, so that ultra-reduction is small, this rolling optimization in generalized predictive control part joined PSO, And the simulation again regardless of the basic generalized predictive controller, The overshoot decreased from the original 0.35mg / L reduced to 0.15mg / L, the control effect significantly changed for the better, that particles in the whole region to search speed, search accuracy and some greatly improved.The results of MATLAB simulation of predictive control of wastewater based on particle swarm optimization showed that the dissolved oxygen concentration stable at 2.00mg / L, and have small overshoot. Sewage treatment system that accounting for 50% of energy consumption is the most energy-consuming part of the tank, the control performance can reach the energy-saving purpose and laid a good foundation to further research.