Abstract: This paper is based on the least-squares minimization algorithm and expound selectment of resisted dimission error factor. From resisting dimission error, this algorithm improves estimative precision of model. By using data of forest survey in Guangdong the paper introduces the method of the parametric identification of the model in detail. The result indicates that the algorithm has a great effect in resisting dimission error. In the course of collecting data,due to all kinds of causes, the data collected must have abnormity. If general leastsquares minimization algorithm was taken,the result of parametric identification have better resisting. In fitting model,general mothed is deleting abnormity using separated point graph, which need to do more work of data treatment. The new method of parametric identification with dimission error can overcome the difficulty and get the better estimation.