In this paper, we propose a lane changing model based on the collision cone approach. Specifically, we show how a vehicle decides whether to change lanes using the collision cone algorithm based on the information of the velocity and the location of surrounding vehicles. The model compares the current and target lane with a new measure of driving advantages. It determines if there are any existing driving advantages such as free space and speed by lane changing. Moreover, a new methodology of lane changing for collision avoidance, which is based on line of sight (LOS) with a new leader in a target lane, is suggested with a model predictive control (MPC) controller. Additionally, we show that the model makes reliable decisions and generates acceptable lane changing trajectories.