Aiming at the drawback of large temperature error of the pressure sensor,a network method making use of the radial basis function (RBF) neural network which has the strong ability of non linear approximation was proposed to realize the non linear correction and temperature compensation,and the method was also improved in this work.The simulation results showed that the non linear errors of the system corrected by this improved method could be compensated automatically.The improved RBF neural network method has the features of small error,high precision and thus is feasible to compensate the non linear error the pressure sensor.