Energy Efficiency Policy Evaluation Model using Hybrid Evolutionary Programming (EP) - Artificial Neural Network (ANN) for Measurement and Verification Baseline Development

This paper presents the development of Energy Efficiency (EE) Policy evaluation model using Hybrid Evolutionary Programming (EP) - Artificial Neural Network (ANN). There are various of EE policies and programmes imposed by governments to promote EE implementation among the industries and societies. However, very little effort was found to evaluate the effectiveness of EE policy for reducing energy consumption. This has created the need for a model that can be used to assess the performance of EE policy and programmes. In this paper, measurement and verification (M&V) approach is proposed for evaluating the performance of Minimum Energy Performance Standards (MEPS) in Malaysia. Artificial Neural Network (ANN) is used for developing a baseline energy model in the M&V framework. Evolutionary Programming (EP) is then integrated with the ANN baseline energy model to improve the performance of ANN. Results show that hybrid EP-ANN provides the highest regression value followed by multi-regression ANOVA model and ANN. With its highest regression value, the EP-ANN provides more accurate and reliable energy savings of 20.03% for the period between 2013 – 2016 from the implementation of MEPS policy in Malaysia.