Increased urban traffic congestion makes it necessary to improve intersection capacity and traffic running conditions.An approach to finding the boundary of a signal group for traffic analysis of the effects of certain intersection improvements is discussed in this paper.Firstly,models are presented for two types of degrees of correlation that form the basis of intersection grouping.The degree of saturation and free-flow travel time are considered in the model for the degree of correlation between two adjacent intersections,and the degree of correlation between any two intersections in the network is analyzed using a Laplacian matrix algorithm.Secondly,a new approach to finding an intersection group boundary is proposed.Thirdly,two measures are adopted for quantifying the results of intersection grouping: the minimum average cut degree of correlation and the minimum traffic influence on intersections outside the boundary.Finally,the method is demonstrated using a city road network.The results of the case study confirm the validity of the proposed approach.