<p>Wind energy, a pollution-free and renewable new energy, has been widely promoted in recent years and has great development potential. An important stage of wind power generation is to solve the problem of wind farm layout optimization (WFLO), that is, to obtain the optimal location of wind turbines so as to obtain the maximum power output with the minimum energy consumption. Based on the Jansen wake model, a WFLO strategy based on the Harris Hawk Optimization Algorithm with Variable Weight Coefficients (ABHHO) was proposed based on the Jansen’s wake model. Two weight coefficients (Beta and Alpha) are used to improve the HHO algorithm. Beta can reduce the search step size, make the global search faster and more detailed, and improve the search efficiency. The vibration pattern of Alpha interferes with the escape energy of the prey, thus moving away from the local optimal. The combination of these two strategies improves the search performance and convergence speed of HHO algorithm. Then seven algorithms (HHO, WOA, BA, AOA, BOA, SCA and RSA) are selected to perform performance optimization simulation with ABHHO on CEC2017 test functions. The experimental results show that ABHHO performs better than other algorithms. Finally, ABHHO was used to optimize the layout of wind power plants, and the performance of Case 1–Case 4 was tested under two initial wind speeds, including the influence of single wake and multi-wake at the same time, shielding part of the wake area, and not shielding the wake area on the production cost and output energy. For the solutions of Case 1–Case 4, 9 algorithms such as HHO, WOA, SCA, PSO, GOA, DE, CSA, ACOR and TLBO, are selected for comparison with ABHHO. Simulation results show that ABHHO can obtain better results in most test schemes and achieve better wind farm layout than other algorithms under the same experimental conditions, as well as obtain higher output power and lower value cost.</p>