A connected environment is crucial for improving road traffic safety and efficiency. However, it remains unclear how different connected environments affect the interaction between vehicles and their impact on driving safety and traffic efficiency in scenarios with potential risks, such as forced lane changes during emergency events. To investigate the effects of different connected environments on drivers' interaction characteristics and their impact on driving safety and traffic efficiency, a group of simulated driving test was implemented in a multi-agent interactive intelligent connected vehicle driving simulation platform. Four types of connected environments were designed, Non-Connected Vehicles (NCV), Front Vehicle Single-Connected Vehicles (FCV), Rear Vehicle Single-Connected Vehicles (RCV), and Double-Connected Vehicles (DCV). Additionally, four different initial headways were tested (10 m, 20 m, 30 m, and 40 m). 40 drivers were recruited to participate in driving simulation experiments, and simulated driving data were collected. The research results indicate that for the front vehicle (FV), connectivity significantly reduces the collision risk with the accident vehicle (TTCFCV = 4.238 s, TTCDCV = 4.385 s), decreases the maximum longitudinal deceleration of FV (FCV = -1.212 m/s2, DCV = -1.022 m/s2), and reduces the speed fluctuation of FV (FCV = 4.748 km/h, DCV = 3.784 km/h). For the rear vehicle (RV), benefits are observed only in the FCV environment, where connectivity helps reduce the maximum deceleration of RV (FCV = -1.545 m/s2), decrease its speed fluctuation (FCV = 3.852 km/h), and enhance overall traffic efficiency (FCV = 12.133 s). Additionally, the minimum time difference to collision (TDTC) in the RCV environment (2.679 s) is significantly higher compared to other connected environments, and the number of cases with TDTC < 1.5 s (49) is notably lower than in other connected environments (NCV = 101, FCV = 107, DCV = 80). This suggests that the RCV environment effectively reduces the lateral collision risk during lane changes. Overall, while single-vehicle connectivity may help reduce driving risks and improve traffic efficiency, DCV may not significantly enhance vehicle safety and traffic efficiency. When the vehicle headway between FV and RV is 20 m (1.651 s), lateral conflicts between the vehicles are most severe. The maximum longitudinal deceleration of FV and RV also significantly decreases with increasing vehicle headway, and when the vehicle headway exceeds 30 m, the maximum longitudinal deceleration of RV nearly ceases to decrease (-1.993 m/s2 at 30 m, -1.948 m/s2 at 40 m). As the distance between the front and rear vehicles (DHWFV-RV) increases, the speed of FV becomes more stable, particularly when DHWFV-RV is 40 m (M = 4.204 km/h), where the speed fluctuations of FV are significantly lower compared to other vehicle headways. A 30-meter vehicle headway (M = 5.684 km/h) is more effective in maintaining speed stability for RV. Although travel time increases with the increase in DHWFV-RV, this change does not show a significant difference. Overall, to ensure traffic efficiency, a vehicle headway of 30 m generally satisfies lane-change safety requirements and provides more stable vehicle speed and acceleration.