Artificial Neural Networks in Modeling of Dewaterability of Sewage Sludge

Mechanical dewatering is a key process in the management of sewage sludge. However, drainage efficiency depends on number factors, from type and dose conditioning agent to parameters device. The selection appropriate methods sludge task laboratory work. This work can be accelerated through use artificial neural network (ANNs). paper discusses possibilities using ANNs predicting physically conditioned input variables were only four characterizing method by centrifugation. These skeleton builders (cement, gypsum, fly ash, liquid glass), sonication (sonication amplitude time), relative centrifugal force. Dewatering such as hydration separation factor output variables. Due nature research problem, two nonlinear networks selected: multilayer perceptron radial network. Based results prediction networks, it was found that these used forecast effectiveness municipal dewatering. error did not exceed 1.0% real value. ANN therefore useful optimizing process. In case conducted research, optimization function factors. Therefore, possible predict has been tested laboratory, for example, with other doses physical conditioner. condition correct large dataset training