Dynamic Setpoint Optimization Using Metaheuristic Algorithms for Wastewater Treatment Plants

Project developed with the objective of using metaheuristic algorithms to update, online, the dissolved oxygen reference values for the PI control of the biological reactor of a wastewater treatment plant.

Table of Contents

• General Information
• Technologies Used
• Optimization framework
• Setup
• Results
• Project Status
• Room for Improvement
• Acknowledgements
• Contact

General Information

• This work originated a paper that was presented at the 48th Annual Conference of the IEEE Industrial Electronics Society - IECON 2022, 17-23 October 2022 - Brussels - Belgium.
• Program developed to provide, in real time, the reference value for the PI control of the biological reactor of a wastewater treatment plant (WWTP).
• The program developed in Python optimizes the aeration system of a WWTP represented by the Benchmark Simulation Model no. 2 (BSM2).
• Metaheuristic algorithms used:
  • Genetic Algorithm (GA)
  • Particle Swarm Optimization (PSO)
  • Simulated Annealing (SA)

Technologies Used

• Python - version 3.8.8
• Keras - version 2.6.0
• TensorFlow - version 2.6.0
• Benchmark Simulation Model no. 2 (BSM2)
• MATLAB/Simulink - version 2021a

Optimization framework

• Proposed optimization framework

Setup

• To run the optimization model proposed in the article, the BSM2 simulator must be installed. More information about the simulator, its installation and operation can be found here https://github.com/wwtmodels and here https://wwtmodels.pubpub.org.
• To obtain the BSM2 surrogate function, run the notebook data_AE, present in the “general” folder. By running the notebook cells in sequence an artificial neural network will be trained. You can use the model model_ann_AE3.h5 already trained, present in the same folder. This model must be added to the directory that will be opened by MATLAB/Simulink.
• With the BSM2 installed, open the Simulink model OTZ_bsm2_cl, present in the General folder.
• For each analyzed algorithm, Genetic Algorithm (GA), Particle Swarm Optimization (PSO) and Simulated Annealing (SA), there is a folder with their respective programs. All these programs must be in the same directory.
• Start the Simulink model OTZ_bsm2_cl, and then run the following programs:
  • To implement GA optimization, run: teste_GA_09_02
  • To implement SA optimization, run: annealing
  • To implement PSO optimization, run: PSO_main

Results

Variation of aeration energy consumption in a three-day sample:

Project Status

The project was completed, and the results were presented in the aforementioned paper.

Room for Improvement

Include areas you believe need improvement / could be improved. Also add TODOs for future development.

Room for improvement:

• The results indicate a reduction in the consumption of aeration energy, but this reduction can lead to the worsening of other factors related to the treatment of wastewater. For applications in real situations other factors must be considered, in addition to the aeration energy, in a multiobjective optimization.
• A different control could be used for the aeration system. Possibly the association of different control techniques and metaheuristic algorithms would provide better results.
• It would be interesting to use other metaheuristic algorithms for this problem.

Acknowledgements

• This research was supported by the ERDF and national funds through the project "PLATAFORMA DE ESTIMAÇÃO, CONTROLO E OTIMIZAÇÃO DE ESTAÇÕES DE TRATAMENTO DE ÁGUAS RESIDUAIS" - SYNAPPS (CENTRO-01-0247-FEDER-046978), and by Fundação para a Ciência e a Tecnologia (FCT) under the project UIDB/00048/2020.
• Institute of Systems and Robotic - ISR
• DEEC - University of Coimbra - Portugal.

Contact

Rodrigo Salles. Email: engenharia.salles@gmail.com

...feel free to contact me!