自主车辆导航系统中的多传感器融合技术

分析了目前农业领域中几种常用的自主车辆导航方式，并说明了多传感器融合的必要性。着重探讨了卡尔曼滤波、模糊逻辑推理以及人工神经网络3种融合方法，分析了未来的发展方向。

Multi-sensor Fusion in Navigation of Autonomous Vehicle

In this paper, some general navigation methods of autonomous vehicle used in the agriculture were analyzed concisely, and the essentiality of multi sensor fusion was derived. Then some methods of multi sensor fusion were introduced, and the attention was paid to the method of Kalman filter, fuzzy logic inference and artificial neural network. Kalman filter can provide an optimal statistical result, and is a potential means in the field of multi sensor fusion. Both fuzzy logic inference and artificial neural network can successfully establish a complex non linear map between the sensor data and the data needed. It showed that all of them could improve the precision and reliability of the system by reviewing many navigation practices in recent years. Finally, the prospect of multi sensor fusion technology in the navigation of autonomous vehicle was discussed.