A Data Driven Approach to Wind Plant Control using Moth-Flame Optimization (MFO) Algorithm

One of the main issues of the wind plant power generation nowadays is that the current stand alone controller of each turbine in the wind plant is not able to cope with chaotic nature of wake aerodynamic effect. Therefore, it is necessary to re-tune the controller of each turbine in the wind plant such that the total power generation is improved. This article presents an investigation of a data driven approach using moth-flame optimization algorithm (MFO) to the problem of improving wind plants power generation. The MFO based technique is applied to search the turbine’s optimum controller such that the aggregation power generation of a wind plant is maximized. The MFO is a population based optimization method that mimics the behavior of moths that navigate on specific angle with respect to the moon location. Here, it is expected that the MFO can solve the control accuracy problem in the existing algorithms for maximizing wind plant. A row of wind turbines plant with wake aerodynamic effect among turbines is adopted to demonstrate the effectiveness of the MFO based technique. The model of the wind plant is derived based on the real Horns Rev wind plant in Denmark. The performance of the proposed MFO algorithm is analyzed in terms of the statistical analysis of the total power generation. Numerical results show that the MFO based approach generates better total wind power generation than spiral dynamic algorithm (SDA) based approach and safe experimentation dynamics (SED) based approach