Process control loop tuning and monitoring using LQG optimality with applications in wastewater treatment plant

The problem of developing a Loop Monitor for control loop performance assessment is considered. Based on some prior modelling experience for a wastewater treatment plant control loop, the approach used a key assumption that the plant could be described using a simple first order stochastic process model. This assumption was shown to lead to explicit formula for optimal LQG controllers and cost function values. Furthermore the resulting controllers were shown to be of PID form; this allowed a valid benchmark comparison to be made for installed industrial PID loop controllers. A generic Loop Monitor algorithm was presented and some aspects investigated using simulation studies. Directions for future work are also given.