Stochastic Optimization Techniques as Effective Tools to Load Forecasting and Scheduling Using Distributed Energy Resources (DERs)

Curbing incessant and erratic power supply to halls of residence within the University of Ibadan Campus has been an impetus that has led to an upsurge in the number of proposals all geared towards providing solution to this critical problem. One of such proposals opines the design of a virtual power plant (VPP). In proposing such, the author seeks to address the problem on a double approach – tackle the erratic power supply and reduce carbon footprints. The proposal in achieving these aims takes advantage of the flexibility of Distributed Energy Resources (DERs), advancements in Information and Communications Technology (ICTs) and the stochastic nature of evolutionary algorithms (EAs) and artificial intelligence (AI) in creating a frame work for interaction between these components, the end users of electricity and the generation/distribution end. The crucial property of electricity being toyed with in the proposal is the ability of electricity to move in both directions depending on existing potential difference. A problem arising from this brilliant proposal though is the fact that loads have not been grouped or biased. The intermittent and stochastic nature of renewables limits their application to certain loads within the halls as such critical loads have to be connected to the school grid for uninterrupted supply. These loads could range from medical to cooking points. This paper seeks to address this issue of load biasing while taking advantage of stochastic optimization techniques in scheduling loads for supply and forecasting demand. The author in attempting to do this hopes to improve quality of supply and optimize demand among students within Independence Hall by suggesting creation of incentives, data mining to observe if a pattern exists which to a great extent mirrors students behavior and other EA tools which would prove useful.