Implementation Of The Detecting Mechanism Of Dangerous Driving States And System Construction

With the increasing development of road traffic, the occurrence ratio of road traffic accidents shows an rising trend. The dangerous driving state is a major cause of road traffic accidents. Fatigue driving state and anger driving state are the two most common dangerous driving states. When a driver is in the fatigue driving state, he cannot operate the vehicle in time as a result of the decreased reaction rate, which will make it easy to cause a traffic accident. When a driver is in the anger driving state, he tends to ignore the risk of anger driving as he cannot calm down, which may cause a traffic accident.A sensor can sense a certain change. The acceleration sensor can effectively perceive the acceleration of the steering wheel, the pulse sensor can effectively perceive the change of drivers’ pulses. The acceleration of the steering wheel has certain relationship with a driver’s fatigue driving state. The change of the pulses has some close relationships with a driver’s fatigue state and anger state. This paper studies how to identify drivers’ fatigue driving state and anger driving state based on taking the drivers’ pulses and the acceleration of the steering wheel as research objects. This paper mainly studies the topic from the following aspects.1. We collect the acceleration of the steering wheel by the the acceleration sensor embedded in the mobile terminal. With the theory of the steering wheel not moving in four seconds as the foundation, this paper proposes a method of detecting drivers’ fatigue driving states based on dynamic threshold of the acceleration. We collect pulses form the pulse sensor. This paper proposes a method for detecting drivers’ anger driving states based on dynamic threshold of the pulses. Based on the above two methods of detecting drivers’ fatigue driving states, this paper puts forward a method of detecting drivers’ fatigue driving states based on the data fusion algorithm, which greatly improves the detection accuracy rate.2. By observing the change of the pulses of the drivers who are in anger driving states and conducting contrast experiments, we identify three characteristics associated with the emotion of anger, namely the pulse amplitude variation rate, the decline from the main wave to the dicrotic wave and the change rate of pulse rate. We can get a learning evaluation model about the drivers’ anger driving states by the training process of the BP neural network. Based on the model, we proposes a method to identify drivers’ anger driving states.3. On the basis of the method of detecting drivers’ fatigue driving states based on data fusion algorithm and the method of detecting drivers’ anger driving states based on BP neural network, this paper designs and constructs a system to detect drivers’ dangerous driving states. We elaborate the function of the system architecture and the realization process of separate modules of the system.