Pareto-Optimal Allocation of Transactive Energy at Market Equilibrium in Distribution Systems: A Constrained Vector Optimization Approach

Abstract

In a grid constrained transactive distribution system market, distribution locational marginal pricing DLMP is influenced by the distance from the substation to an energy user, thereby causing households that are further away from the substation to be charged more. The Jain index of fairness, which has been recently applied to alleviate this undesirable effect of inefficient energy allocations, is used in this research to quantify fairness. It is shown that the Jain index is strictly quasi-concave. A bilevel distributed mechanism is proposed, where at the lower level, auction mechanisms are invoked simultaneously at each aggregator to obtain energy costs under market equilibrium conditions. A constrained multi gradient ascent algorithm, Augmented Lagrangian Multigradient Approach ALMA, is proposed for implementation at the upper level to attain energy allocations that represent tradeoffs between efficiency and fairness. Theoretical issues pertaining to ALMA as a generic algorithm for constrained vector optimization are considered. It is shown that when the objectives are restricted to be strictly quasi concave functions and if the feasible region is convex, ALMA converges towards global Pareto optimality. The overall effectiveness of the proposed approach is confirmed through a set of MATLAB simulations implemented on a modified IEEE 37-bus system platform.