THE INFLOW PREDICTION TO SEWAGE PROCESS USING CASE-BASED TECHNOLOGY AND ITS EFFECTIVENESS IN PROCESS OPERATION
Toshiaki Oka, Hiroaki Tsutsui, Tomohiro Konda, Kenichi Yamagata · Chikyuu Kankyou Shimpojiumu kouen rombunshuu · 2002
In a combined sewer system, rainfall causes sudden increase of sewage water. Itsamount often reaches more than disposal capacity of a sewage plant, so the sewage water is poured into rivers and the sea without sewage purification. That causes a problem of an environmental pollution.The prediction of inflow into a sewage plant can provide an optimal operation plan for the sewage plant. So the prediction is useful for preventing from the emission of swage water with heavy pollution load in the beginning of rainfall.However, there are many difficulties to predict the sewage inflow using traditional modeling method because of its complex relationships between process and environment.To overcome the difficulties in this inflow prediction problem, we applied a case-based technology named Topological Case-Based Modeling (TCBM). TCBM has a capable of making a causal model easily from sampled data that include inherent characteristics of the object.In this paper, we introduce the results of inflow predictions by the model using TCBM in a swage plant and show the examples of effective applications of this model.