基于BP-HMM的驾驶盲区危险等级预警模型研究

Research on Early Warning Model of Dangerous Level of Driving Blind Area Based on BP-HMM

文章提出了一种基于BP神经网络和隐马尔科夫模型的汽车驾驶盲区危险等级识别的防撞预警系统模型。立足于驾驶盲区产生根源,使用层次分析法对影响驾驶状态的主要因素进行分析,得出每个驾驶危险等级的初始决策向量;并建立BP网络模型,确定HMM参数,使用MATLAB对其进行联合仿真,通过Viterbi算法预测出当前汽车驾驶盲区危险等级,从而提前消除车辆行驶过程中汽车盲区所带来的潜在安全危害,实现驾驶盲区危险等级的预警,提高汽车的主动安全性。

In this paper, the collision warning system model based on BP neural network and Hidden Markov Model of car driving blind danger level recognition. Based on the root cause of the driving blind zone, the analytic hierarchy process is used to analyze the main factors affecting the driving state, and the initial decision vector of each driving hazard level is obtained.And the establishment of BP neural network model to determine the parameters of HMM, co-simulation using MATLAB by the Viterbi algorithm to predict the current car driving blind danger level, thereby eliminating potential safety hazards automotive vehicle blind spots caused by the process of moving ahead to achieve driving blind danger levels, which can improve the active safety of cars.