中文摘要：

为避免电机制动失效对驾驶安全的影响,提出了一种基于驾驶员意图识别的电动汽车电机制动失效监控系统。根据控制要求,将系统的硬件部分分为电源电路、信号采集调理电路、故障监测电路和故障处理电路并进行了设计。将驾驶员制动意图分为缓慢制动、紧急制动、正常制动。利用k-medoids聚类算法对数据进行离散化处理,利用双层隐形马尔科夫模型对驾驶员的制动操作意图进行识别。提出基于模糊推理的失效等级进行判断方法,并根据不同失效等级制订了车辆响应方式。利用Simulink与Simplorer对控制系统进行联合仿真,并对系统进行了试验验证。结果表明,该控制策略能正确识别驾驶员的制动操作意图,正确判断故障等级,并据此作出正确响应,可保证车辆制动时的安全性、操控性,实现既定的控制要求。

英文摘要：

In order to avoid the influence of motor brake failure on driving safety, a kind of system of monitoring motor brake failure of electric vehicle based on driver intention recognition is put forward. According to the control requirements, the designed hardware of the system is divided into power supply circuit, signal processing circuit, fault monitoring and failure processing circuit. The driver intention is divided into slow brake, emergency brake, and normal brake. The data is discretized by K-medoids clustering algorithm, and the driver＇s operation intention is recognized by double hidden Markov model. The vehicle response modes are respectively made based on the failure grade judged by fuzzy reasoning. The union simulation of control system is conducted by Simulink/Simplorer, and brake test on the system is conducted. The result shows that the control strategy can recognize the driver＇s brake intention correctly, judge the failure grade properly, and it makes correct response according to the failure grade. The control system can ensure the vehicle safety, handling in brake, and achieve the established control requirement.