The estimation of parameter and nonparameter in a semiparametric regression model with a fixed design are investigated based on the semiparametric regression model research. Under convergence system error, the depth-weighted technique is used to obtain depth-weighted wavelet estimation in the semi -parametric regression model, and the independent errors are extended to the case of convergence system. It verifies convergence rate of the estimation.