A Novel Fuzzy Inference System for Predicting Balanced Electricity Demand

Abstract

In developing economies, reducing waste in hard-earned electricity generation has been a major problem. As a case of interest, the issues of unit of power commitment decision-making have befuddled the post-privatised electricity framework in Nigeria. This choice simply determines how many units of power should be provided to a certain location at a given time. To address this issue, a novel model known as the Automated Fuzzy Inference Engine (AFIE) was formulated that mathematically suggests the output in the fuzzy rules in the inference engine instead of relying on intuition and the ranking method of fuzzification to detect the output. Python was utilized in the AFIE model experiment. As a fallout of this, some algorithms were birthed in the process of formulating data sets. According to the model, every 5°C increase in temperature increases electricity consumption by 0.96%. The models support previous findings and the premise that electricity usage is related to temperature, humidity, and standard of living. It also confirms that rainfall has a negligible relationship con electricity use. However, this study discovered that electricity supply is inversely correlated to rainstorms and directly proportional to bill payment history and living standards. Applying the import of this model will put stakeholders in the vantage position to prevent mistakes caused by users setting output values intuitively. Furthermore, this approach may be applied to any fuzzy predictive model as long as the input parameters are correctly categorised and weighted in relation to the output.