An Optimization Method Based on Random Fork Tree Coding for the Electrical Networks of Offshore Wind Farms

Abstract The electrical network optimization is an important aspect of reducing the development cost of offshore wind farm because it accounts for nearly 25% of total investment. This paper presents an integration of the random fork tree coding scheme, a union-finding algorithm and electrical parameter calculating models for electrical network optimization. The coding scheme is developed for the first time for integrated optimization of the tree connection topology, substation positions, and cable cross sections to solve the defects of conventional step-by-step approaches such as minimum spanning tree, to achieve the most economical solution. The case studies clearly show that the proposed method, when coupled with the Omni-optimizer, achieves optimal economical matching solutions and can be applied for optimization of the tree-structure electrical network with any number of wind turbines and substations. However, the proposed method is not applicable for other topological structures, where the coding schemes need to be re-modelled. Also, gradient algorithms may be used in the optimization for convergence enhancement for hundreds of decision variables of the large-scale offshore wind farm.