Multi-objective Optimization for Preventive Tree Trimming Scheduling in Overhead Electric Power Distribution Networks

Abstract

One of the main causes of power supply interruption in distribution networks is the vegetation. The high cost of tree trimming schedules encourages utilities to look for more efficient ways to manage vegetation. In this work, the vegetation maintenance planning is formulated as a combinatorial multi-objective optimization problem, which determines the most appropriate tree trimming schedules in the zones of the distribution feeders. This formulation considers two conflicting objectives: one is the social cost given by the customer interruption cost, and another is the vegetation maintenance cost of utility. The constraints comprise the availability of human resources, distribution system reliability and priority zones. The solution technique of the proposed optimization problem is the elitist non-dominated sorting genetic algorithm II for two different application cases. In the first case, failure rates from vegetation in the zones are obtained using a reliability index and a vegetation growth model. In the second application case, failure rate multipliers are used to quantify the impact of maintenance on system reliability. Results from the application of the optimization algorithm for both cases in a real-world distribution system composed of 11 feeders are presented and discussed.