2 April 2021 in IEEE Transactions on Power Delivery
This paper presents a comprehensive approach to improve the daily performance of an active distribution network (ADN), which includes renewable resources and responsive load (RL), using distributed network reconguration (DNR). Optimization objectives in this work can be described as (i) reducing active losses, (ii) improving the voltage prole, (iii) improving the network reliability, and (iv) minimizing distribution network operation costs. The suggested approach takes into account the probability of renewable resource failure, given the information collected from their initial state at the beginning of every day, in solving the optimization problem. Furthermore, solar radiation variations are estimated based on past historical data and the impact of the performance of renewable sources such as photovoltaic (PVs) solar generators is determined hourly based on the Markov model. Since the number of reconguration scenarios is very high, stochastic DNR (SDNR) based on the probability distance method is employed to shrink the scenarios set. At the nal stage an improved crow search algorithm (ICSA) is introduced to nd the optimal scenario. The effectiveness of the suggested method is veried for the IEEE 33-bus radial distribution system as a case study.
6 August 2024 in IEEE Transactions on Smart Grid
There is a potential conflict between P2P trade and distribution network operator (DSO) decisions, which slows down the expansion of P2P trade. Expanding the P2P trade requires reducing this conflict and increasing DSO awareness of the actual behavior of prosumers. In other words, mitigating this conflict requires that DSO approximates prosumers’ utility functions (PUFs) based on their actual behavior. On the other hand, PUFs approximated based on the actual behavior of prosumers have various parameters such as freedom in decision-making, collective influence, privacy, and marginal cost/utility. This is a mathematical challenge for DSO because this class of PUFs may not be convex or continuously differentiable. Hence, in this paper, a near-potential game (NPG) framework is proposed to support the design of P2P trade with PUFs belonging to this class. Also, to develop a realistic model of P2P trade, we classify prosumers into residential and non-residential classes and assume that prosumers have limited information about each other’s decisions. Then within an NPG framework, we introduce a learning model, whereby each prosumer obtains an estimate of the prosumers’ decisions in P2P trade.