Primary vertex reconstruction based on the Kalman filter technique at BES

Primary vertex reconstruction is crucial to estimate the beam profile in collision experiments. We study the principle of an iterative process, called the Kalman filter method, and apply it to primary vertex reconstruction at BESⅢ. A Newton procedure to find the zero point of the distance function＇s gradient is used for primary vertex finding in 3-dimensional space. Results are obtained based on raw data at BESⅢ.