Optimal performance of microgrid in the presence of demand response exchange: A stochastic multi-objective model

Abstract

This paper investigates optimal scheduling of the microgrids, including renewable energy sources and conventional generators. Due to the stochastic nature of the wind speed and the sun irradiation, generated power of the wind turbines and photovoltaic are highly uncertain and in the proposed model the uncertainty of the load, price, and renewable energy sources generation are considered by utilizing the normal distribution function. Incentive-based demand response program is implemented in the operating process. The optimal economic status is achieved by maximizing the microgrid's demand response program profit and minimizing the generators cost, and the trading cost. This multi-objective model is solved by the weighted sum technique in order to produce the optimal Pareto solutions and the trade-off solution is selected by applying the fuzzy satisfying method. It is performed on two different microgrids and sensitivity analysis is executed. Results demonstrate that demand response program reduces unused energy in both scenarios.

Introduction

The division of the grid into productive sub-systems, so-called microgrids (MGs) which integrate distributed generation (DG) for local demand, is proposed to increase manageability and reduce transportation losses [1]. The MG can be connected either to other MGs or the main grid for energy exchange or can be run in island mode [2] as circumstances or economics dictate. Both conventional generators and renewable energy sources (RESs) such as wind turbine (WT) and photovoltaic (PV) systems can supply the power demand of the MGs.

The MGs are capable of controlling and efficiently using distributed generators, energy storage systems [3], electrical vehicles, and loads in a more coordinated way and according to the energy management principles [4]. Purchasing or selling energy to the power system, as well as providing ancillary services, improving continuity of supply and power quality delivered to local loads, and reducing the environmental impact are the most important aims of the MGs [5]. The optimized operation of the MG is significant to achieve efficient energy management at minimum cost and maximum benefit [6]. The modeling of the RES output is required to coordinate the daily utilization of the RES intermittent character for reducing the operating costs and preserving accounting for the electrical energy price of selling/purchasing energy as well as satisfying the demanded power [7]. An optimization model is proposed in [8] for optimizing the MG operations, accounting for the MG operation constraints as well as distributed energy resources with stochastic generation and time-varying demand. A probabilistic technique is surveyed for planning active distribution grids with the presence of the demand response program (DRP) by applying uncertainties [9]. Moreover, a scheduling program is investigated for the PV generators in the electricity market [10]. Grid-connected MGs with variable electricity prices are analyzed in [11]. Reference [12] proposed a strategy for controlling the MG, consisting of a conventional generator, storage, and the PV. A collective energy dispatch solution is presented in [13] to optimally coordinate the DGs, distributed storage units and critical demands across multiple autonomous MGs based on a “tree stem-leaves” approach. A predictive strategy model for a system consisting of a battery, the PV, and a conventional generator is proposed in [14]. The energy management problem of the MGs is investigated in [15]. In [16], a multi-objective optimization matter is presented for reducing the emission and cost of a hybrid system by applying the DRP. Utilization of the DRPs helps the MG's operator to achieve a reliable state [17]. The DRPs decrease the MG operational cost and improve MG operation [18]. In another study, a stochastic optimization model is proposed for determining the optimal forward loads and selling prices to consumers by a single retailer [19]. Despite the above studies, more researches are required for utilization of the DRP in the power system problems to participate customers and minimize the MG fuel costs.

To the best of our knowledge, no similar research has been so far performed. In this work, the uncertainties of load, market price and output power of the PV and WT in grid-connected mode are considered. Simultaneously, two conflicting objective functions are optimized, which were neglected in [18].

Novelty and contributions of this research are briefly presented as follows:

• A stochastic energy cost function is simulated by implementing the DRP, which was not applied in [20].

• The probabilistic nature of the market price, load and generated power of the PV and WT is considered, and a robust stochastic model is applied in the proposed method.

• The best solution among the Pareto optimal solutions is selected by the fuzzy satisfying method and solved by the weighted sum technique. Moreover, in comparison to [21], instead of using an imperfect method for selecting the proper answer, optimal results are achieved by generating the Pareto optimal solutions and implementing the fuzzy satisfying method.

The other parts of this paper are as follows: Section 2 describes the model of the MG including the DRP. Section 3 concentrates on the method implemented on the case studies, and results are shown in Section 4. In Section 5, the study is concluded.