Optimal MV/LV transformer allocation in distribution network for power losses reduction and cost minimization: A new multi‐objective framework

Abstract

Among distribution network planning (DNP) problems, medium voltage (MV) to low voltage (LV) transformer allocation is one of the most important and challenging ones. This article presents a multi‐objective optimization framework for solving MV/LV transformer allocation problem considering two goals: power losses and investment cost. Transformer allocation problem has been solved in three scenarios: (1) base case, (2) mid‐term planning, and (3) long‐term planning. In order to effectively and efficiently solve this nonlinear problem, multi‐objective crow search algorithm (MOCSA) and multi‐objective particle swarm optimization (MOPSO) have been applied and compared in terms of Pareto front. Simulation results show that solving transformer allocation problem in multi‐objective framework results in a set non‐dominated solutions which provide a trade‐off between losses and cost. Moreover, MOCSA produces more promising results than MOPSO.

An efficient optimization framework for distribution network planning by simultaneous allocation of photovoltaic distributed generations and transformers

“…It should be noted that all the load blocks should be serviced via a transformer. Table 2 indicates the details of the transformers used in this research (including rated capacity, rated voltage, no-load losses and short circuit losses) [4]. This information has been adopted from the transformer data sheet.…”

“…where x t and y t coordinates are related to the transformer, x b and y b coordinates are related to the load block and k is a constant coefficient which is used to convert the direct distance (between load block and transformer) to the real one (here 1.55) [4].…”

“…Equation () can be used to calculate the distance between the load block and the transformer as follows: $$$$\begin{equation}d = k \times \sqrt {{{\left( {{x}_t - {x}_b} \right)}}^2 + {{\left( {{y}_t - {y}_b} \right)}}^2} \end{equation}$$$$where $${x}_t$$ and $${y}_t$$ coordinates are related to the transformer, $${x}_b$$ and $${y}_b$$ coordinates are related to the load block and k is a constant coefficient which is used to convert the direct distance (between load block and transformer) to the real one (here 1.55) [4].…”

“…Results indicate that the load balancing significantly decreases the operation cost. In [4], a multi‐objective crow search algorithm (MOCSA) has been used to allocate MV/LV transformers in the distribution network with the aim of minimizing power losses and overall investment costs. The final result of the multi‐objective framework is a series of non‐dominated solutions, named Pareto Front, which provide a trade‐off between power losses and investment costs.…”

A risk‐averse strategy based on information gap decision theory for optimal placement of service transformers in distribution networks

“…Load uncertainty and the cost of LV lines have been ignored in the models. In [2], a multi‐objective crow search algorithm (MOCSA) has been employed to deal with MV/LV transformer allocation issues by minimizing power losses and overall investment costs. The research demonstrates a trade‐off between power loss and investment costs using a series of non‐dominated solutions.…”