A velocity azimuth display (VAD) wind retrieval method based on conjugate gradient algorithm is proposed, in which the conjugate gradient algorithm is used to replace the solution of Fourier series expansion in original VAD method by using optimization theory. In order to solve the problem that the algorithm does not converge to the optimal solution in wind retrieval, the Hessian matrix is used to modify the algorithm. At the same time, a synchronous comparison experiment had been carried out with Doppler lidar and standard wind cup anemometer in accord with IEC 61400-12-1 international standards for 43 days. The results show that, when the azimuth scanning range and radial number of lidar are 60° and 7 respectively, the correlation coefficients of wind speed and wind direction are 0.991 and 0.998, the standard deviations of wind speed and wind direction are 0.52 m/s and 5.1°, the deviations of wind speed and wind direction are -0.02 m/s and 3.6°. The results of comparison experiment prove that the VAD wind retrieval method based on conjugate gradient algorithm can guarantee the measurement accuracy to meet international standard in case of low azimuth scanning range with strong applicability. Meanwhile, the measurement performance of lidar system is confirmed, which provides a better choice for the monitoring of dynamic complex wind field.