| This thesis is out based on the control and monitoring system of a sewage treatment factory of Chongqing.Activated sludge treatment is a kind of biology treatment method, and is used widely all over the world. It's also a kind of typical complex process, whose optimized control research is always focused by a lot of people. At the present time, PID control is still the most popular way in the activated sludge treatment control system. The performance of PID control relies on its own control parameters very much. But because there are a lot of uncertain factors in the activated sludge treatment system, it not only makes the ascertain of PID parameters become very hard but also brings out poor performance of the control system along with the change of the treatment process.On the other hand, because of the particularity of activate sludge treatment, it's hard to measure some essential parameters (especially the BOD, who is the most important parameter to monitor the quality of the waste water) on-line, and can't provide some useful information to the running control system.In order to solve those problems mentioned above, the following jobs are carried out:First, a new model with uncertain parameters of activated sludge wastewater treatment is established in this thesis;Second, based on the analysis of the stability and controllability of the uncertain parameters model, a method of condition calm feedback is given out;Third, Genetic Algorithm is used to optimize the PID parameters to meet the performance requirement of the control system of the system model with changeable parameters;At last, a new soft-measuring approach based on GABP neural network is put forward to make the BOD be measured on-line. In this method, genetic algorithm is used to optimize the net weights of BP neural network.The application of Genetic Algorithm and Neural Network in the process of wastewater treatment is worthful both in academic studies and practical application.
|
PDF Full Text Request