基于沙蚁pol-神经元模型的航向角处理方法 an angle processing method based on the pol-neuron model of desert ant.pdfVIP

第 40 卷第 10 期 光电工程 Vol.40, No.10 2013 年 10 月 Opto-Electronic Engineering Oct, 2013 文章编号：1003-501X(2013)10-0028-07 基于沙蚁 POL-神经元模型的航向角处理方法 王 波，高 隽，范之国，王 昕 ( 合肥工业大学 计算机与信息学院，合肥 230009 ) 摘要：沙蚁POL-神经元对天空偏振光的感知机理为研究仿生偏振光导航方法提供了仿生基础。针对其信号处理方 法中存在的测角歧义性问题，通过分析 POL-神经元模型的测角原理，提出了一种基于高阶多项式拟合的航向角处 理方法，并根据航向角随模型参考轴旋转角度的变化规律，将模型的有效测角范围由[-45°,45°]线性地增加到 [0°,180°]。最后，利用四通道偏振光导航传感器对该方法进行了实验验证。理论与实测结果表明，该方法能有效 解决 POL-神经元模型中存在的航向角歧义性问题，并可显著提高模型的测角范围。 关键词：沙蚁；偏振光导航；POL-神经元模型；仿生导航 中图分类号：TP212.3 文献标志码：A doi：10.3969/j.issn.1003-501X.2013.10.006 An Angle Processing Method Based on the POL-neuron Model of Desert Ant WANG Bo ，GAO Jun ，FAN Zhiguo ，WANG Xin ( School of Computer and Information, Hefei University of Technology, Hefei 230009, China ) Abstract : Polarization sensitive mechanism of the POL-neuron model of desert ant (Cataglyphis) offers the basis for the study on bionic polarization navigation. Referring to the angle ambiguity error of the existing signal processing method of POL-neuron model, a new angle method is proposed based on high-order polynomial fitting after analyzing the working principle of the model. According to the relationship between navigation direction and the rotation angle of reference axis, so the effective angle range can be increased from [-45°, 45°] to [0°, 180°] linearly. Finally, the corresponding experiment is also done to test the performance of this new angle method using a 4-channel polarized light navigation sensor in real skies. The result demonstrates that this method can solve the angle ambiguity error of POL-neuron model effectively, and improve the an