Day-ahead battery scheduling in microgrid considering wind power uncertainty using ordinal optimization

Abstract

This paper introduces the Ordinal Optimization (OO) theory into the microgrid operation, considering the wind power uncertainty. An energy balance model is established to obtain a day-ahead battery scheduling. Comparing with the stochastic optimization and the robust optimization, the OO method has the advantages of neither requiring a huge computation burden, nor resulting in an unexpectedly high operating cost. Case studies on different algorithms have been done in the paper. The results show that the OO method outperforms the stochastic and robust solutions, with a more reasonable operating cost while without compromising the microgrid reliability. Index Terms-Microgrid, day-ahead battery scheduling, wind power uncertainty, ordinal optimization. I. NOMENCLATURE Subscript t Value of a variable at the time t Ω B , Ω T , Ω W , Ω U , Ω L Set of batteries, time segments, wind turbines, generating units and system nodes, respectively (•) Unit operation cost function ($) ( ) Electricity price charged by the main grid at time t ($/kWh) ( ) Interchange power between the main grid and the microgrid at time t, when the microgrid is receiving electricity, ( ) is positive (kW) ( ) Cost to charge battery b ($/kWh) ( ) Cost to discharge battery b ($/kWh) ( ) Reward for islanding state ($/kWh) ( , ) Charging power of battery b at time t (kW) ( , ) Discharging power of battery b at time t (kW) ( , ) Scheduled power of unit k at time t (kW) ( , ) Forecasted output of wind turbine w (kW) ( , ) Curtailment of wind turbine w (kW) ( , ) The forecasted load of bus n at time t ( ), ( ) Technical maximum and minimum power output of unit k (kW) ( ), ( ) Maximum and minimum charging power of battery (kW) ( ), ( ) Maximum and minimum discharging power of battery (kW) ( ), ( ) Technical maximum and minimum stored energy of battery (kW) ( ), ( ) Upper and lower bound of interchange power (kW) ( ) Initial stored energy in battery b (kW) ( , )Binary decision variable: "1" if unit k is on at time t; "0" otherwise ( , ) Binary decision variable: "1" if battery is charging at time t; "0" otherwise ( , ) Binary decision variable: "1" if battery is discharging at time t; "0" otherwise ( ) Binary decision variable: "1" if the microgrid is under islanding state at time t; "0" otherwise ( ), ( ) Binary decision variable: "1" if the microgrid start/end islanding state at time t; "0" otherwise up , down Ramp-up/ramp-down rate limit for unit k (kW) up , down Ramp-up/ramp-down rate limit for battery b (kW)

Reliability Evaluation and Selection in Multi-State System

“…Reference [10] explained a viewpoint for the search problem using ordinal optimization. In this paper, we use equation (1) (BP selection rule) to calculates the value of s.…”

“…Therefore, this paper introduces a new model consisted of Monte Carlo simulation and ordinal optimization algorithm [9][10] to shorten the computing time in finding the greatest system among a great number of multi-state series-parallel systems. _______________________ In the proposed model, firstly, we need to classify the systems into several sorts and choose the smallest series-parallel system with least components in each sort, calculate its precise reliability by Monte Carlo simulation.…”

Optimal energy management strategy in microgrids with mixed energy resources and energy storage system

“…The dynamic economic dispatch model for microgrids employing a dynamic programming algorithm is presented in [11]. Xu et al [12] propose a day‐ahead battery scheduling for microgrids based on ordinal optimisation theory. A microgrid optimisation strategy based on linear programming is presented in [13].…”

Effects of correlated photovoltaic power and load uncertainties on grid‐connected microgrid day‐ahead scheduling

“…For instance, in [17] and [18], P‐OPF models are formulated when considering the variation of load uncertainty. In [19], another P‐OPF model is proposed for microgrid day‐ahead battery scheduling when the wind power uncertainty is taken into account. While the above P‐OPF models are very promising in microgrid day‐ahead scheduling, the correlated PV‐DG and load uncertainties have not been considered simultaneously.…”