Non-linear state dependent differential Riccati states filter for wastewater treatment process

The most important issues relating to monitoring, quality control and prediction models for environmental protection in the treatment plant waste water are based on the amount of information and measures that are available. The key step in controlling and monitoring the plant is to obtain an accurate and robust estimate of the states model. The paper focuses on estimating non-measurable physical states of wastewater treatment system, which are unavailable because of difficulties techniques or the high cost of physical sensors. The developed filter is dealing with the non-linearity describing the system. The Activated Sludge Process (ASP) as the biological technique most commonly used wastewater treatment, attracts much attention the research community. We developed for this class of processes a robust non-linear estimator known as "state-dependent differential Riccati filter (SDDRF). The sensor software is simple to implement and has a computational cost relatively low. The results are compared with the extended Kalman filter (EKF) to demonstrate the improved performance of the filter SDDRF. The filter allows the online monitoring of process variables, which are not directly measurable. The simulation results prove the advantage of using this approach.