摘要
以路面识别为目的,利用自适应神经模糊网络(ANFIS)进行路面不平度激励时域估测研究.首先建立车辆1/4模型运动微分方程,并使用白噪声信号激励车辆模型,利用激励产生的模型动力学响应进行自适应神经模糊系统训练.之后对训练获得的逆向车辆动力学模型进行分析并利用随机路面激励产生的系统响应进行随机路面时域估测.最后对自适应神经模糊网络系统隶属函数个数及输入数据组合进行分析比较.仿真结果显示,自适应模糊神经网络系统能够以较高的精度完成路面时域估测.
Based on adaptive neuro fuzzy inference system（ANFIS）,a time domain estimation for road profile input was presented.Differential equations of quarter vehicle model were created,and white noise signal was employed to stimulate the model and the dynamic response was used to train the inverse dynamic model with ANFIS.In the simulation,different kinds of random excitations were used to verify the accuracy of ANFIS,and the effect of the number of membership function and combination of input data were also discussed.The result shows that ANFIS can be used for road estimation and its time domain reproduction.
出处
《北京理工大学学报》
EI
CAS
CSCD
北大核心
2015年第5期481-484,489,共5页
Transactions of Beijing Institute of Technology
关键词
自适应神经模糊网络
路面识别
时域估测
路面不平度
adaptive neuro fuzzy inference system（ANFIS）
road estimation
time domain estimation
road profile
