基于模糊神经网络的车辆运行状态多维智能监测方法研究
Research on Multi-Dimensional Intelligent Monitoring Method for Vehicle Operation Status Based on Fuzzy Neural Network
范文明 1马宏伟 2杨晓峰2
作者信息
- 1. 国能铁路装备有限责任公司,北京 100010
- 2. 天津哈威克科技有限公司,天津 301799
- 折叠
摘要
车辆自动驾驶领域,其运行状态数据种类较多,特征模糊化明显,难以准确采集,导致车辆运行状态监测存在准确度低、精度低、时效慢等问题.为此,提出了基于模糊神经网络的车辆运行状态多维智能监测方法.首先,通过多传感器采集车辆运行状态数据,并使用自适应加权平均算法对采集到的数据实行融合处理;其次,通过自适应遗传算法和浮动搜索算法,获取车辆运行状态多融合数据的最优特征子集;最后,将车辆运行状态最优特征子集输入至模糊神经网络模型完成车辆运行状态多维智能监测.实验结果表明,所提方法能够实现对减速、正常、加速、抗蛇行四种车辆运行状态的准确监测,且对车辆运行状态的监测时效高,适用于实际应用.
Abstract
In the field of vehicle autonomous driving,there are many types of operating status data,with obvious feature fuzziness and difficulty in accurately collecting,which leads to problems such as low accuracy,low accuracy,and slow time efficiency in vehicle operating status monitoring.Therefore,a multi-dimensional intelligent monitoring method for vehicle operation status based on fuzzy neural network is proposed.Firstly,vehicle operating status data is collected through multi-ple sensors,and the collected data is fused using an adaptive weighted average algorithm.Secondly,by using adaptive genet-ic algorithm and floating search algorithm,the optimal feature subset of vehicle operating state multi fusion data is obtained.Finally,the optimal feature subset of vehicle operation status is input into the fuzzy neural network model to complete multi-dimensional intelligent monitoring of vehicle operation status.The experimental results show that the proposed method can achieve accurate monitoring of four vehicle operating states:deceleration,normal,acceleration,and anti hunting,and has high monitoring efficiency for vehicle operating states,making it suitable for practical applications.
关键词
车辆运行状态监测/多传感器/自适应加权平均算法/最优特征子集/模糊神经网络模型Key words
vehicle operation status monitoring/multiple sensors/adaptive weighted average algorithm/optimal feature subset/fuzzy neural network model引用本文复制引用
出版年
2024
国家科技期刊平台
NSTL
万方数据