The Application of Multi-Agent-Based Dynamic Immunity Model in the Car Safety-Detection

This paper presents a safety-detection model to improve the car's safety. It includes three sub-models: the model of describing the state of car in Multi-Agent system, setting vehicle detecting-area and the model of identifying the abnormal agent-cars. After identifying the abnormal vehicle, an adjusting-action is given in order to guarantee the normal driving. Some simulations are done and the results show that the model's response-time to the dangerous signal that has appeared in the past period is less than other model's. Meanwhile, the average decision-making time is decreased. The result also indicates that the threshold influenced the correctness of decision-making. According to the results, 0.55 is better than other threshold value.