In order to optimize the parameters of nonlinear models of crop growth and development, particle swarm optimization (PSO) algorithm is used in the parameter estimation of the nonlinear system model, and is verified though parameter estimation of tomato growth and development model. With the help of good man-machine interface, the method will be more accurate, effective and convenient. Experiments show that: with this method, after the parameter optimization of crop model and the simulation of dynamic process of tomato growth, compare the simulation values with measured values, the both having a good closeness of fit.