Water treatment approaches in the current wastewater treatment have complicated treatment processes (A2O, AOAO, membrane treatment, ozone treatment and activated carbon treatment besides standard treatment) and measurement parameters.
Our target is to save energy and optimize the entire plant by controlling non-linear behaviors based on multi-variate data (measurement value, control target) and making improvements in rapid response and convergence, while they were infeasible with conventional PID control.
We perform statistical multi-variate analysis that serves as a basis of the modern control theories. By adopting sophisticated control theories such as fuzzy control, neural network control, LQ (Linear Quadratic) control and so forth, we develop mathematical models and control systems suitable for operating an individual plant.
- Graph-editing-type energy-saving simulator
Simulation is conducted by confirming changes of sewage pump units in operation on the graph. In this way, energy saving with a high water level operation and active use of night-time electric power can be achieved.
- Blower control for N2O suppression
By using control logic for N2O generation in a reactor tank, blower operation is controlled to suppress N2O generation through model-based estimation control.
- MLSS control
By watching the MLSS value of a reactor tank, the discharge volume and sludge removal volume from a final sedimentation tank is controlled to maintain MLSS value at a high level. This enhances the treatment efficiency of a water treatment plant.