An Intelligent IoT-Based Predictive Control System for Water Quality and Energy Management in Koi Aquaculture

Area of Science:

• Aquaculture Engineering
• Artificial Intelligence in Environmental Management
• Cyber-Physical Systems

Background:

• Ornamental aquaculture faces challenges in balancing energy consumption and water quality.
• Existing management methods often lack predictive capabilities, leading to inefficient energy use.

Purpose of the Study:

• To develop and evaluate an integrated predictive and energy-aware aquaculture management framework.
• To reduce energy consumption in ornamental aquaculture while ensuring stable water quality.

Main Methods:

• Utilized Internet of Things (IoT) sensors for real-time monitoring of key water parameters (DO, NH3, temperature, pH, turbidity) and energy usage.
• Employed Long Short-Term Memory (LSTM) models for environmental state prediction.
• Integrated a physics-informed Digital Twin (DT) for validating predictions and enhancing control safety.
• Implemented a Cyber-Physical System (CPS) for smart predictive control.

Main Results:

• The Digital Twin accurately simulated pond dynamics, achieving R² values of 0.97 for dissolved oxygen and 0.94 for ammonia.
• The smart predictive control mode reduced total energy consumption by 26.86% compared to manual operation.
• Average daily energy consumption decreased significantly (p < 0.001) from 212 Wh/day to 154.71 Wh/day.

Conclusions:

• The proposed integrated framework effectively reduces energy consumption in aquaculture.
• Combining IoT, LSTM prediction, Digital Twin simulation, and CPS control offers a reliable and energy-efficient solution for aquaculture management.
• This approach enhances operational reliability and control safety in aquaculture systems.