Embedded Intelligent Controller In The System Of Coagulant Dosage

Coagulant dosing process is the most important part in the domestic routine water treatment procedure. It is a hot topic on how to add coagulation accurately in current water industry. Because of the affect factor of the raw water flow, the turbidity, PH, temperature and other pollutions, the process of coagulant dosage is a very process of time-delay, nonlinear, time-variation, many interference factors. According to these factors, in analysis of the domestic and foreign coagulant dosing systems and the practical situation in China, the paper find out a hybird control structure combinated with Expert System(ES) and BPNN to active dosage control, under the background of a factual project.In the coagulant dosing control algorithm, feedforward compensation ability of ES is used to resolve the interfering factors of flow and turbidity mutation. Turbidity before filter as a feedback value to establish the expert knowledge and reasoning mechanism. According to different rules to choose different contol modes. Through the neural network nonlinear mapping ability to solve the non-linear effects of water treatment, roll learning and feedback to learn to solve the time-varying and time-delay problems. However, ANN has a disadvantage of slow convergence. An improved particle swarm optimization (PSO) algorithm is used to optimize the BPNN and improve the convergence speed.Finally, the intelligent algorithm is applied to embedded system. The embedded system contains PXA270 microprocessor and other hardware, using Windows CE as the operating system. Developping the application program on embedded system platform, the controller communicate with bottom devices and applied to actual water treatment industry. The experiment explains that the intelligent dosing controller can improve the accuracy of coagulant dosage and save the dosage. The Water quality can reach the standard, it has widest application foreground.