Aiming at the problem of temperature drift of micro-electro-mechanical system(MEMS) piezoresistive pressure sensor affected by ambient temperature, based on the analysis of the common temperature compensation methods, a temperature compensation model of pressure sensor based on the particle swarm optimization radial basis function (PSO-RBF) is proposed in this paper. Combined with the sample data collected in the calibration experiment, the nonlinear mapping relationship between the calibration pressure and the output voltage and temperature of the sensitive element is established, and the temperature compensation is realized. The results show that the proposed algorithm has better compensation prediction effect than the traditional least squares method, cubic spline interpolation method, standard RBF, PSO-BP, ELM neural network and other methods, and does not need to normalize the sample data, which has good engineering practice significance.