Employing Consumer Electronics and WSN for Low Carbon Emissions in Agri-Food Practices Through a Hybrid Whale-Wolf Nature-Inspired Model

Climate change is a global disaster, and emissions of greenhouse gases are the main cause. Reduced emissions of greenhouse gases are largely attributable to the agricultural sector. Governments worldwide are implementing laws to encourage the usage of low-carbon energy sources, which are becoming increasingly important due to the worsening impacts of climate change. The food and agriculture industries are utilising sensors and Wireless Sensor Networks (WSNs) to control processes, track quality, and ensure safety. WSNs facilitate real-time data collection and monitoring of environmental parameters, enhancing agricultural systems' capacity to adapt to climate challenges. Because machine learning (ML) is currently being utilized to solve environmental concerns, it is an opportune time to explore ML models for CO2 projections from the agri-food sector. Additionally, the implementation of Carrier Sense Multiple Access with Collision Avoidance (CSMA/CA) in WSNs allows for efficient management of data transmission, ensuring reliable data collection while reducing the risk of data loss. By enhancing our capacity to predict future energy demand and resource availability, the scientific community and businesses stand to gain substantially from artificial intelligence and big data analysis. This research presents a hybrid model named Whale-Wolf, grounded in nature-inspired algorithms. It aims to address carbon dioxide emissions from the food and agriculture industry. By hybridising Whale Optimization and Grey Wolf Optimization, this research efficiently anticipates carbon emissions and facilitates low-carbon economic calculations. A subset of the most relevant features from the agri-food carbon dioxide emissions dataset is selected using the optimised Whale-Wolf algorithm. The Whale-Wolf, GWO, and WOA algorithms select characteristics of 8, 12, and 14 respectively. Using Whale-Wolf, we achieved a 98.61% success rate with the random forest machine learning classifier.