For the existing 10 kW high-power industrial microwave oven, a relay is used as the control actuator. When using traditional control methods for heating, there is a large overshoot and obvious temperature oscillation, and the system temperature stability is low. To solve the above problems, back propagation neural network PID control is introduced into the microwave heating temperature control of the installation, and simulation comparison and experimental verification are conducted using tap water as the heating object. Firstly, using existing input and output experimental data, establish a temperature control model for industrial microwave ovens; Secondly, use MATLAB/SIMULINK to build a high-power industrial microwave oven temperature control system and conduct simulation comparative experiments; Finally, experimently verify the temperature control performance of the back propagation neural network PID control method in industrial microwave ovens when heating 5 kg of tap water. The experimental results show that this method has smaller overshoot and no significant temperature oscillation compared to conventional PID and fuzzy PID control in the medium temperature control during microwave heating process, effectively improving the system temperature stability during the operation of high-power industrial microwave ovens, and helping to improve product quality and safety performance.