Adaptive optimal control for a wastewater treatment plant based on a data-driven method.

Area of Science:

• Environmental Engineering
• Control Systems Engineering
• Artificial Intelligence

Background:

• Wastewater treatment plants (WWTPs) require precise control of operating points for efficiency.
• Optimizing dissolved oxygen and nitrate levels is crucial for effluent quality and energy consumption.

Purpose of the Study:

• To propose a data-driven adaptive optimal controller (DDAOC) for WWTPs.
• To optimize operating points for dissolved oxygen and nitrate levels.
• To reduce energy consumption and improve effluent quality.

Main Methods:

• Development of a DDAOC based on adaptive dynamical programming.
• Implementation of an evaluation module to estimate future energy consumption and effluent quality.
• Utilization of an optimization module to adjust operating points iteratively.
• Testing and evaluation on the Benchmark Simulation Model No.1 (BSM1).

Main Results:

• The DDAOC successfully optimized operating points for dissolved oxygen and nitrate levels.
• The controller demonstrated significant reductions in energy consumption compared to a PID controller.
• The DDAOC operates effectively using only input-output plant data, negating the need for a mechanistic model.

Conclusions:

• The proposed DDAOC is an effective data-driven approach for optimizing WWTP operations.
• This method offers substantial energy savings in wastewater treatment.
• The DDAOC provides a viable alternative to traditional control methods, especially when mechanistic models are unavailable.