Abstract: Pulverized coal (PC) is widely used in thermal power,chemical and metallurgical fields. In the process of preparation and transportation,due to its increased oxidation activity,it is easy to spontaneously ignite and induce explosion accidents,which seriously threatens the life safety of operators. In order to prevent the occurrence of PC disasters,it is of great significance to study the influencing factors and prediction of the critical parameters of PC spontaneous combustion. The key critical parameters such as the minimum ignition temperature (tM) and the corresponding ignition delay time (ti) for the self-ignition of PC were studied. Through the establishment of the self-ignition evolution experimental system of PC,the pulverized coal with two levels of metamorphism from the Liuhuanggou mine and the Hongliu mine were collected and placed in an equidistant cylindrical reaction tank for a constant temperature oxidation experiment. The influencing factors of the critical parameters of PC spontaneous combustion were analyzed. The mathematical prediction model for the critical parameters of PC spontaneous combustion was established to predict the critical parameters. The results show that oxygen concentration and accumulation thickness are important factors affecting the critical parameters of PC spontaneous combustion. With the decrease in oxygen concentration,MIT and ti of the two types of coal powder increase remarkably. The decrease of oxygen concentration inhibits the reaction of the active structure of PC with oxygen,slows down the heating rate of PC,and then increases the MIT and ti of pulverized coal. As the thickness of pulverized coal increases,the heat released by pulverized coal oxidation is more likely to accumulate. The minimum ignition temperature of pulverized coal gradually decreases,and the maximum drop can reach 34 ℃. The increase in the thickness of the PC accumulation can slow down the diffusion rate of oxygen into the PC,which shows an increasing trend corresponding to ti. A linear regression mathematical prediction model is established using linear regression method,which can be used to calculate the critical parameters of pulverized coal spontaneous combustion and provide a basis for the prevention and control of deflagration accidents caused by PC spontaneous combustion.