The new intelligent transportation system can issue a collision warning based on driving habits
chrispeng [ad_1]
The mechanical engineers of the Massachusetts Institute of Technology are developing a new type of anti-collision intelligent safety system. The difference is that its intelligent algorithm can issue a collision warning to the driver based on the driver’s driving habits, and finally take corresponding measures. Control the vehicle to prevent car accidents.
So far, engineers have developed a variety of safety systems to prevent collisions, including automatic cruise control systems, radar or laser-based sensor systems, systems that can slow down when approaching other vehicles, and blind spot warning systems. These systems generally use flashing lights or whistle to remind drivers to pay attention to other vehicles, or automatically brake when the vehicle slips or loses control of steering.
However, the above systems all face a common challenge, that is, how to deal with the problem of excessive security protection in the design. If a security system is too sensitive, it will alert every car on the way. If the driver believes that the safety system will continue to issue unnecessary warnings, they will tend to think that the safety system is unreliable. This will cause the driver to ignore the warnings of the safety system, which will cause the safety system to exist in name only.
Therefore, the study of human driving patterns has become an important basic work for the development of automotive safety systems. In order to establish the algorithmic model of the safety system, De Vekwo and Woma of the Massachusetts Institute of Technology divided human driving behavior into two modes: braking and accelerating. They first establish a limited set of vehicle positions in different driving modes within a specific time, such as a tenth of a second or ten seconds later; then combine the driver’s driving habits, such as acceleration or deceleration measures at an intersection, Constructed a new algorithm model of automobile safety system. In this algorithm model, intersections are defined as dangerous areas. When a car equipped with an intelligent anti-collision safety system encounters an intersection, it can judge the actions of other vehicles based on the sensors on the roadside or on the traffic lights, and then take certain anti-collision measures. If two cars equipped with an anti-collision intelligent safety system arrive at an intersection at the same time, they will exchange location information and pass through the intersection in a cooperative manner to avoid traffic accidents.
In order to test its new algorithm, De Vekwo and Woma built a small simulated traffic road in the laboratory, on which two small mini cars drove. One of the cars is fully autonomous, and the other is manned. In order to reflect different driving habits, the experimenters also selected 8 volunteers. In 100 tests, the system avoided 97 crashes. After analysis, the researchers found that the three crashes were mainly caused by information transmission delays, because the road infrastructure required a certain amount of time to capture the position and speed of vehicles not equipped with intelligent systems and transmit them to the intelligent collision avoidance system, resulting in system errors. Sentenced. Researchers are trying to improve the system to solve the problem of information delay.
At present, researchers have begun to conduct field real-time testing of the system, focusing on obtaining human driving response data to determine when the smart safety system issues an alarm and when safety measures are taken. In addition, the researchers also hope that the system can take road conditions and weather conditions into account so that they can make more informed decisions.
[ad_2]
