In order to explore the influence of the oxidation and spontaneous combustion process of fractured coal with different burial depths under uniaxial stress, an experimental platform for the spontaneous combustion characteristics of coal under load and uses bituminous coal samples from the Liuhuanggou mining area in Xinjiang to carry out uniaxial stress temperature programmed test in an oxygen-poor environment under 0~8 MPa is built in this paper. According to the relationship between the gas generated in the experiment and the temperature, the apparent activation energy and oxygen consumption rate of the coal sample under uniaxial stress are calculated. Combining coal spontaneous combustion oxidation kinetics and pyrolysis parameters, the nonlinear development process of coal from slow oxidation to rapid oxidation under uniaxial stress is described, and the catastrophe temperature and the critical temperature of coal oxidation-combustion process under test conditions are calculated based on catastrophe theory, which determine four characteristic parameters, TCO,THY, TCO′, and THY′ characterized by CO and oxygen consumption rate, and the variation law of different characteristic parameters with uniaxial stress is analyzed. The analysis results show that the pyrolysis gas concentration, apparent activation energy and oxygen consumption rate show a cubic function law that first increases, then decreases and then increases with the increase of uniaxial stress (the critical axial pressures at 1.8 and 5.5 MPa). At 1.8 MPa, the apparent activation energy and various characteristic parameter values are the lowest, the oxygen reaction rate of coal is the fastest, and the oxygen consumption rate is the highest; when the uniaxial stress is 5.5 MPa, the oxygen consumption rate is the highest, and the number of new cracks is created the most and the characteristic parameters of TCO have the greatest impact. The temperature index of coal spontaneous combustion slowly transitions to rapid oxidation, and the catastrophe temperature TCO characterized by the CO concentration is the most accurate. Important theoretical guiding significance for the prediction of the development stage of fire areas with different burial depths in the production process and the prevention and control of coal fire disasters is concluded in this research.