Fuzzy modeling approach to predictions of chemical oxygen demand in activated sludge processes

Abstract

This paper is concerned with predicting the chemical oxygen demand, the common criterion of the outlet water in most wastewater treatment plants. A fuzzy modeling approach is proposed which can efficiently eliminate the complex nonlinear terms existing in Activated Sludge Model No. 1 (ASM1). First, the structure of fuzzy rules in the approach is identified based on ASM1. Further, by combining the well-known fuzzy c-means cluster algorithm and the method of least squares, the fuzzy space of input variables required in the approach is partitioned and the consequent parameters are identified using the data in Benchmark Simulation Model No. 1, which can be replaced by the data in a real wastewater treatment plant. The effectiveness of the proposed approach is illustrated by comparing the predicted values of chemical oxygen demand and the experimental measurements obtained from the Benchmark Simulation Model No. 1.