Simulation Based On Artificial Neural Network For Sbbr Shortcut Nitrification Treatment

In the recent years,nitrogen pollution in the water has become more and more serious, which leads to more and more disaster, so people pay more attention to this problem. It's very important to take actions for nitrogen removal. In this paper,the research findings and up-to-date research progress in here and abroad were introduced. It pointed out the importance of developing biological nitrogen removal in our country,The novel biological treatment technology will significantly improve the level of nitrogen removal. Recently,some new technologies of ammonia removal such as shortcut nitrification has been developed very fast,which provide economic and feasible ways to the treatment of high ammonia wasterwater.The feasibility of dynamic simulation of shortcut nitrification process in the four SBBR reactors based on artificial neural network (ANN) was studied. With Back-Propagation algorithm and the adaptive study rate,a dynamic simulation model was established by Matlab software,reflecting the nonlinear function relationship between NH₄⁺-N,DO,temperature,external carbon source of influent and NH₄⁺-N and NO2--N of effluent. With the optimized operating parameters of BP model: S1=7, mc=0.7,lr=0.25,max_epoch=4000,the results showed that the numerical outputs and the experimental values match well,with a highest error of 13.90% and a lowest error of 0.99%. In the entire sample test,the absolute average error rate of the simulation values for the NH₄⁺-N of effluent was 6.42%,and the absolute average error rate of the simulation values for the NO2--N of effluent was 6.57%.The value contribution relationships between each input factor and output results were studied by weighted average analysis , which indicate that NH₄⁺-N and temperature have tremendous influence on the shortcut nitrification process. The study reveals that the ANN can reflect the nonlinear function between influent and effluent parameters,and is suitable for the dynamic monitoring of the shortcut nitrification bio-process for wastewater.