Optimal Feedback Control for Water Quality in an Activated Sludge Process

Abstract

Two algorithms for control of water quality in an activated sludge process are newly developed by using the optimal regulator theory. One is an optimal water quality control which can regulate the organic matter concentration in effluent discharge. The other is a water quality multivariable control which can regulate sludge age as well as organic matter concentration.
Dynamic behaviors of the activated sludge process are described by an aeration tank model and a final clarifier model. The aeration tank model consists of three nonlinear partial differential equations which express the variations in organic matter concentration, mixed liquor suspended solids and sludge age. Among these models, the model of sludge age is newly developed by the authors.
A linear state space model is obtained by replacing nonlinear partial differential equations by twenty eight linear ordinary differential equations with techniques of linearlization and difference.
Based on this state space model, two optimal control algorithms are derived. New controls have two strong points. First, they can respond quickly and stably by a feedback control of states within the aeration tank. Secondly, they can eliminate the offset errors of output variables by the integral control function.
From the simulation results, it is found that the new control systems are superior to conventional control systems.