Process Modeling and Control of Enhanced CoagulationAmerican Water Works Association, 2000 - 145 páginas Develops and evaluates artificial neural network (ANN) models for clarified water turbidity and total organic carbon. Determines which factors are most important in optimizing enhanced coagulation. Also develops protocols for source data analysis, model architecture selection, model training, and model evaluation. Studies the use of ANN models in process control. |
Índice
INTRODUCTION | 1 |
BACKGROUND INFORMATION | 3 |
METHODOLOGY | 37 |
RESULTS | 53 |
ANALYSIS OF ENHANCED COAGULATION USING MODELS | 105 |
PROCESS CONTROL | 111 |
CONCLUSIONS AND RECOMMENDATIONS | 125 |
Términos y frases comunes
3-layer activation function algorithm alkalinity alum dose Amirtharajah ANN technique AQUALTA Artificial Neural Networks AWWA backpropagation network chemical clarifier effluent color clarifier effluent turbidity coagulant dose Color model Color model results color removal data patterns determination disinfection E.L. Smith WTP enhanced coagulation process evaluated feed-forward flocculation fulvic acid hidden layer neurons historical data humic acid input and output input parameters inverse model learning rate Malcolm Pirnie Malcolm Pirnie Inc mean absolute error model for E.L. model input model residuals natural organic matter North Saskatchewan River number of hidden number of neurons on-line optimal organic matter organic removal output parameters overflow rate PAC dose PID control Polymer dose precursors process parameters production set protocol Randtke range raw water quality real-time data results for E.L. results for Rossdale Rossdale WTP SCADA system testing Trihalomethane Turbidity model results water quality parameter water treatment plant
Referencias a este libro
Artificial Intelligence Systems for Water Treatment Plant Optimization Christopher W. Baxter Vista previa restringida - 2001 |
Guidance Manual for Coagulant Changeover James DeWolfe,AWWA Research Foundation Vista de fragmentos - 2003 |