A Predictive Economic Analysis Under Uncertain Scenario: Evidence Theory for Small-Scale Residential Community Power Network

Residential communities increasingly adopt renewable energy sources (RES) to reduce energy costs. However, these sources are weather-dependent and unpredictable, posing challenges for reliable operations. Therefore, managing uncertainties in power supply has become an important research focus. These uncertainties directly affect the performance of microgrids. Energy storage systems are integral, enabling battery-backed supply for residential loads during unpredictable conditions. This paper presents an evidence-based fusion framework for accurate load and PV power forecasting in uncertain scenarios. Dempster-Shaffer's theory has been used to develop a fusion framework. In addition, electricity network modeling for a day-long electricity cost analysis of a residential community has been done in three scenarios. In scenario 1, the community power supply is connected to the grid; in Scenario 2, users have a PV system on their premises but no storage; and in Scenario 3, users have their own PV-integrated storage system. The electricity price for buying and selling energy from the grid is also assumed to be fixed. The primary objective of this framework is to minimize the operating cost under a set of different constraints. The results showed that the prediction performance of the proposed algorithm is better than others. Thus, under uncertainty, scenario 3 is more economical than the other two, and the total cost with the proposed algorithm is less than $0.705 in scenario 1 and less than $0.339 in scenario 2.