为准确地判定紧急制动状态,以便及时启动座椅坐垫倾角调节机构,对汽车制动模式判别策略进行研究.选取在汽车中低速行驶下的缓慢、点刹、紧急制动3种模式,采用正交试验设计方法,选用不同驾驶员以不同车速进行了制动实验.在对实验数据进行分析的基础上,采用了踏板力峰值法与踏板力变化率法两种判别算法.实验结果发现,两种算法都存在局限性.为此,又尝试了基于RBF神经网络模式识别算法.结果表明,这种算法能准确地判别制动模式.%For accurately ascertaining the state of emergency braking in order to promptly actuate seat cushion inclined angle adjuster, a research on vehicle braking pattern recognition strategy is conducted.Firstly by using orthogonal experimental design method, different drivers are chosen to carry ont braking experiments with slow, intermittent and emergency three braking modes at different medium and low speeds.On the base of analyzing test data, two discrimination algorithms, based on peak pedal force and the changing rate of pedal force respectively, are adopted.The results of experiment indicate that both algorithms have certain limitation.Accordingly a RBF neural network-based pattern recognition algorithm is attempted.The results show that the algorithm can accurately discriminate braking patterns.
展开▼