Artificial -neural -network and genetic -algorithm for optimization of helical -blade -vertical -axis -wind -turbine

Wind energy as a renewable and sustainable type of energy has been attractive from past eras. Three helical blade vertical axis wind turbine (VAWT-3-HB) is suitable for the use in urban areas with low-speed wind flow due to its low required amount of torque for self-starting and its low noise generation. The optimization of VAWT-3-HB with application of Artificial -Neural -Network (ANN) and Genetic -Algorithm (GA) which are very important tools for proper design and improving the performance and of this category of wind turbine still is not covered in literature. For GA optimization procedure, the average power coefficient ($C_{p-ave}$) was the objective function which had to be maximized. Design variables were airfoil chord length, helical angle, and the blade tip speed ratio which were selected after extensive 3-D-CFD simulation runs and examining all effective parameters. The optimal values of these parameters were obtained 0.42 m, 30 $°$, and 1.4 respectively. $C_{p-ave}$ at the optimum point was 0.1845 with 218 % rise (in comparison with 0.058 before optimization). Results of a 3-D-CFD simulation run with optimal values of design variables showed a good match between average power coefficients predicted by ANN-GA and predicted by 3-D-CFD simulation run with about 0.21 % difference.