Energy analysis of poultry housing in Ghana using artificial neural networks

Global level improvement of resource use efficiency in agro-ecosystems has always been the target of experts who are keen on reducing the environmental impact emanating from agriculture. To this end, energy audit analysis in agro-ecosystems to determine the energy use at the various subsectors and levels of the agricultural sector, including poultry birds production is crucial. Most of the energy analysis works in Ghana in the past were in the area of commercial buildings using traditional energy analyses approach. Little work has so far been done on energy analysis for poultry houses. EnergyPlus simulation of the ambient conditions and energy inputs were analysed and compared to the Artificial Neural Network (ANN) model to observe the performance of the ANN in predicting energy consumption. The annual energy consumption estimations were found to be 2,044 kWh and 1,452 kWh for lighting and equipment usage respectively. The model robustness checks showed that coefficient of determination values for training, validation, testing and overall were 0.95304, 0.9533, 0.9505 and 0.9527 respectively for each regression plot which shows that the ANN model was suitable for determining the energy consumption in a poultry production facility, and can be replicated for more refined predictions in Ghana. (C) 2022 The Author(s). Published by Elsevier B.V. on behalf of African Institute of Mathematical Sciences / Next Einstein Initiative.

Automated summary

This article highlights the importance of energy audit analysis in agro-ecosystems and its application in predicting energy consumption in poultry production using EnergyPlus simulation and Artificial Neural Network (ANN) model. The results indicate that the ANN model is suitable for determining energy consumption and can be used for more refined predictions in Ghana.