基于轮地作用参数和PLSDA方法的月壤力学性能评估

发布时间：2018-05-15 22:18

本文选题：地面力学 + 月球车　；参考：《航空学报》2015年11期

【摘要】：为评估在轨月球车周边环境月壤力学性能,以月球车辙信息、轮上载荷和滑转率作为基本参数,提出了16个二元及三元标识量,结合偏最小二乘判别方法(PLSDA)建立评估月壤力学状态模型。根据容重对模拟月壤力学状态进行分级,分别为松软状态、自然状态和紧实状态。应用轮壤相互作用试验台共采集247组试验数据,每种模拟月壤状态的试验数据按照2∶1比例随机划分为校正集样本和预测集样本,最终校正集和预测集样本个数分别为166和81个。考虑到原始数据值相差大和所提出部分标识量包含冗余信息的特点,因此在建立PLSDA模型时,应用均值中心化预处理方法对原始数据进行数据预处理,并优选10个标识量建立识别月壤力学状态的PLSDA模型,对应的校正集的准确率和预测集的准确率分别为90.96%和90.12%。结果表明,应用PLSDA方法并结合月球车的车辙信息、轮上载荷和滑转率以及优选标识量所建立的评估月壤力学状态判别模型,其计算快速准确,可用于月球车在轨评估车轮前方月壤的力学特性和通过性能评估。
[Abstract]:In order to evaluate the mechanical properties of lunar soil around the orbiting lunar rover, 16 binary and ternary marking quantities were proposed with lunar rutting information, wheel load and slip rate as the basic parameters. Combined with partial least squares discriminant method (PLSDA), a model for evaluating the mechanical state of lunar soil was established. According to bulk density, the mechanical state of the simulated lunar soil is classified as soft state, natural state and compacted state. A total of 247 sets of experimental data were collected by using the rotor-soil interaction test rig. Each experimental data of simulating lunar soil state was randomly divided into correction set samples and prediction set samples according to 2:1 scale. The number of samples in the final correction set and prediction set is 166 and 81, respectively. Considering the large difference of the original data value and the redundancy information of the proposed partial identifiers, the mean centralization preprocessing method is used to preprocess the original data when the PLSDA model is established. The PLSDA model for identifying the mechanical state of lunar soil was established by selecting 10 markers. The accuracy of correction set and prediction set were 90.96% and 90.12% respectively. The results show that the model of evaluating the mechanical state of lunar soil established by PLSDA method, combined with the rutting information of the lunar rover, the load on the wheel, the slip rate and the optimal marking quantity, is fast and accurate. It can be used to evaluate the mechanical properties of the lunar soil in front of the wheel and to evaluate the performance of the lunar rover in orbit.
【作者单位】：吉林大学工程仿生教育部重点实验室;华东交通大学机电工程学院;中国空间技术研究院北京空间飞行器总体设计部;
【基金】：国家自然科学基金(51375199) 中国空间技术研究院总体部资助项目(2012M520676) 吉林大学国防预研基金(JDXJY20130207) 工程仿生教育部重点实验室开放基金(K201406)~~
【分类号】：P184.5;V476.3

【参考文献】