A preventive control method of power system transient stability based on a Convolutional Neural Network (CNN) is presented to better realize the control. The sensitivities of the CNN output variable are calculated to select control generators and determine the control amount, and the corresponding control scheme is verified by the transient stability assessment method based on the CNN combined with time-domain simulation to obtain the control scheme. This can stabilize the system under anticipated contingencies. A particular provincial power grid is analyzed to verify the effectiveness of the method. The results reveal that effective control measures are obtained with the proposed CNN-based transient stability preventive control method, and they can restore the system to a stable state. This work is supported by National Key Research and Development Program of China (No. 5442XT180029).