Optimal offshore wind farms' collector design based on the multiple travelling salesman problem and genetic algorithm

The capital cost of the electrical network of a large offshore wind farm constitutes a significant proportion of the total cost. Finding the optimal design of electrical network is imperative task and it is addressed in this paper. The objective of this paper is to present a methodology for the optimal design for the offshore wind farms' collector system; it is based on the Multiple Travelling Salesman Problem and Genetic Algorithm. A cost model has been developed that includes a more realistic treatment of the cost of step-up transformers and undersea cables. These improvements make this cost model more detailed than others that are currently in use. Optimization model is specifically designed for offshore wind farms' collector system and it considers different cable cross sections when designing the radial arrays. A novel optimization method is used; it is based on an improved Genetic Algorithm and includes a specific application of the Open-Multiple Traveling Salesmen Problem (fsomTSP) considering a special gene coding developed for this specific formulation. The proposed approach is tested with a hypothetical wind farm where the convergence is examined versus number of wind turbines.