Diesel partial premixed compression ignition (PPCI) is sensitive to air-fuel mixing state and boundary conditions, so it is difficult to build a control-oriented combustion prediction model. In order to solve the above-referred problem, the sensitivity analysis of PPCI fuel injection control parameters was carried out, and the Weber function linearization combustion model of PPCI diesel engine was established. The results show that the change of fuel injection parameter has a great sensitivity to combustion parameter. The combustion prediction model established by using Weber function linearization method can predict important parameters such as cylinder pressure and heat release rate in real time, and the correlation coefficient between the calculated value and the calibration value of model parameter is greater than 0.98.