In order to improve the initiative of lane changing game between Connected and Automated Vehicles and drivers vehicles in the upstream section of the signalized intersection, the left-turn Connected and Automated Vehicle is taken as the research object to analyze the characteristics of the forced lane change game in this section. Firstly, by analyzing the driving intention and lane change behavior of vehicles in the upstream section of the signalized intersections, the driver's expectation function was set to objectively reflect the driving demand of the vehicle, and the safety and driving efficiency of the vehicle were taken as the benefits and quantified. Under the assumption of complete information, the optimal lane changing decision was obtained through the game equilibrium solution to maximize the lane changing benefit. Secondly, in order to improve the comfort of lane changing, a five-degree polynomial was used to plan the lane change trajectory and the game lane changing process between Connected and Automated Vehicles and driver vehicles is realized. Finally, the simulation test was used to verify the model, and the influence of different lane changing positions and remaining green time on the decision-making of Connected and Automated Vehicles were analyzed. The results indicate that in the process of non-cooperative game forced lane changing at signalized intersection, the lane changing probability increases by an average of 0.69 % with the increase of the unit lane change position, and the vehicle lane changing probability decreases by an average of 0.82 % with the increase of the unit green light remaining time. The game lane changing characteristics and decision tendency of vehicles on upstream sections of signalized intersections are analyzed by simulation, which is conducive to providing decision guidance for lane changing of Connected and Automated Vehicles.