摘要
提出了一种基于GA-SVM的汽车追尾预测方法。该方法选取行车间距、后车车速、前后车速差、汽车制动时间、制动减速度等5个因素作为预测模型的输入指标,选取追尾概率作为输出向量,通过建立SVM预测模型,并用GA进行参数优化,对汽车追尾概率进行预测。通过仿真值与预测值对比,证明了该方法的准确性。
In this paper,a GA-SVM-based approach is proposed to predict rear-end collision.A SVM-based prediction model is built and parameterized with GA to predict the probability of rear-end collision,with five factors including vehicle distance,following vehicle speed,speed difference between the front vehicle and the following vehicle,braking time and braking deceleration selected as input indexes and probability of collision as output vector.Through the simulation value and predicted value contrast,this method proves the accuracy.
出处
《汽车技术》
北大核心
2012年第11期24-26,49,共4页
Automobile Technology
基金
江苏省高校研究生科研创新计划(CXLX12_0628)
国家质量监督检验检疫总局资助项目(2010IK084)
