Existing commercial central air conditioning control systems only control the indoor temperature, and the setting temperature value depends on the experience of operators or managers, and cannot be intelligently controlled through the demand feedback of service objectives. This paper will introduce an intelligent air conditioning control system based on human heat radiation automatic adjustment of wind field distribution, this system applies convolutional neural network analysis for human heat radiation images, according to the actual sensory state of the human body and the distribution of personnel, by adjusting the air outlet angle to control the wind field distribution and other ways to intelligently control the air conditioning system, so that it can achieve the ideal indoor temperature field distribution. At the same time, advanced prediction algorithms are introduced in the control strategy to make the air conditioning system more energy-saving, healthy and comfortable.