Abstract: This study investigates the average power generation efficiency of an array consisting of three vertical axis wind turbines through two-dimensional computational fluid dynamics (CFD) numerical simulations. Research has shown that the flow spacing between the blade tip speed ratio and the upstream and downstream wind turbines has a certain impact on the average power generation efficiency of the array: when in the low blade tip speed ratio stage (TSR = 1.0), excessive angle of attack causes flow separation on the blade surface, leading to dynamic stall and reducing the aerodynamic performance of the wind turbine. As the blade tip speed ratio increases, the flow separation phenomenon gradually disappears until the optimal blade tip speed ratio (TSR = 2.0) is reached, at which point the aerodynamic performance is optimal. When the blade tip speed ratio continues to increase (TSR = 3.0), the blade angle of attack is no longer the optimal value, the lift to drag ratio decreases, and the aerodynamic performance decreases; On the basis of a blade tip speed ratio of 2.0 in the wind turbine array, the downstream wind turbine is moved in the direction of the flow. At this time, the obstruction of the upstream wake by the downstream wind turbine is weakened, and the power generation efficiency of the two upstream wind turbines increases. When it moves to 3.7 times the diameter (10.36 m), the accelerated airflow passes between the two upstream wind turbines, and the downstream wind turbine is less affected by the upstream wind turbine wake, and the system power generation efficiency reaches its peak; The downstream wind turbine continues to move in the direction of the flow, and is more affected by the wake of the upstream wind turbine, resulting in a decrease in the power generation efficiency of the downstream wind turbine; Continuing to change the blade tip speed ratio of downstream wind turbines, when its blade tip speed ratio is 2.25, the three wind turbine array becomes the optimal operating condition for this study, corresponding to a peak average power generation efficiency of 31.90%.