Abstract: For the existing engine start-stop system, the engine will stop as long as the engine meets the basic conditions of idle stop, resulting in frequent starting and stopping of the vehicle, which will not only aggravate the wear and tear of engine parts, but also increase the fuel consumption of the whole vehicle. To solve this problem, a hybrid electric vehicle engine start-stop trigger system (HEV-EST) combined with road information was designed. HEV-EST start-stop conditions were determined according to HEV engine idle and reboot fuel consumption chart. BP neural network number recognition algorithm and color recognition algorithm were used to construct the traffic signal recognition model. The queuing theory of overall planning was used to analyze and calculate traffic flow, constructing a traffic flow analysis model. Finally, HEV-EST was designed by combining the traffic signal recognition model and the traffic flow analysis model. Based on MATLAB simulation verification, the results demonstrate that the proposed system significantly reduces the number of invalid vehicle starts and stops and improves the overall vehicle fuel economy.

Key words: vehicle engineering; hybrid electric vehicles; traffic signal recognition; traffic flow; engine start-stop

摘要： 现有的发动机起停系统只要发动机满足怠速停止的基本条件发动机就会停止，造成车辆频繁起停，这样不仅会加剧发动机部件磨损，还会增加整车油耗。为解决这一问题，设计了一种结合道路信息的混合动力汽车发动机起停触发系统（HEV-EST）。根据HEV发动机怠速和重起油耗图，确定HEV-EST起停条件；利用BP神经网络数字识别算法和颜色识别算法构建交通信号灯识别模型；运用统筹学的排队论对交通流进行分析计算，构建交通流分析模型；最后结合交通信号灯识别模型和交通流分析模型，设计了HEV-EST；并进行了MATLAB仿真验证。结果表明：该系统明显降低了车辆无效起停的次数，提高了整车燃油经济性。

关键词: 车辆工程； 混合动力汽车；交通信号灯识别；交通流；发动机起停