Optimal Dispatch of a Standalone PV-Diesel-Battery Microgrid using Non-Commercial MINLP Solvers

The paper presents an approach to optimize the dispatch of a standalone microgrid consisting of diesel generators, photovoltaic modules, and battery storage units. The day-ahead unit commitment and economic dispatch of an example microgrid are formulated in the form of a mixed-integer nonlinear programming (MINLP) problem. The objective function of the optimization problem was defined as a daily sum of the microgrid operational costs. The problem was constrained by the condition of power balance, the unit's minimum up-time and down-time constraints, dispatching availability, minimum and maximum power limits, and the battery storage maximum depth of discharge. Two non-commercial deterministic MINLP solvers were used to solve the formulated problem of finding a global optimal solution for the day-ahead hourly PV-diesel-battery microgrid scheduling.

I. Introduction

A traditional solution for the supply of decentralized consumers in remote areas is the use of diesel power plants. Nevertheless, despite the widespread prevalence of this type of power plant in isolated power supply systems, diesel generators have several significant disadvantages, the main one of which is high operating costs. The integration of power plants based on renewable energy sources into this kind of isolated microgrids along with diesel power plants can significantly reduce the cost of fossil fuels. The operating mode of hybrid power plants, which include power plants based on renewable energy sources, directly depends on the current meteorological situation. This fact significantly complicates the search for the optimal operating mode of such power plants under conditions of uncertainty.