中文摘要：

月壤力学特性参数的研究可以使人们了解更多的星球地质学信息，也是进行月球探测车等设备开发以及未来从事人类月球活动的工程基础。利用月球车轮地作用测试平台和模拟月壤对6种不同尺寸和轮刺的车轮进行试验，利用传统压板试验和剪切试验测量土壤力学参数。针对月球车轮地作用地面力学积分模型进行耦合度和参数敏感度分析，进而将8个力学参数分为3组，即接触角系数、承压特性参数和剪切特性参数。提出一种循环迭代的参数辨识方法，利用相关度最大的挂钩牵引力、沉陷量和前进阻力矩分别对3组参数依次进行辨识。采用试验数据进行验证，结果表明，此方法可以高精度地辨识土壤的3个剪切参数，组合沉陷模量可以设定为一典型值，接触角系数和沉陷指数系数与车轮相关，反映了车轮的轮刺效应和尺寸效应。该方法避免了简化模型带来的参数辨识误差，实现了对于月壤参数的全面辨识，既可以估计月壤特性，还可以提高轮地作用力学模型的预测精度。

英文摘要：

The research of lunar soils＇ mechanical property parameters can both improve our scientific knowledge of the moon＇s geological properties, and provide engineering knowledge required for development of exploration rovers or future human settlement activities. Experiments are carried out for six kinds of wheels with different dimensions and wheel lugs using the wheel-terrain interaction test-bed developed for lunar rovers and lunar soil stimulant, the mechanical property parameters of which are measured by conventional plate-sinkage experiments and shearing experiments. The degree of coupling and parameter sensitivity are analyzed for the integrated wheel-soil interaction terramechanics model, based on which the eight mechanical parameters are divided into three groups, i.e., contact angle coefficients, bearing performance parameters and shearing performance parameters. A cyclic iterative parameter identification approach is brought forward to estimate the three groups of parameters step by step, using the measured data that have maximum correlativity with them respectively, i.e., drawbar pull, wheel sinkage, and resistance moment. The experimental data are adopted to verify the approach, it is proved that three of the shearing parameters of soil can be identified with high precision; the lumped sinkage modulus can be set to a typical value; the sinkage exponent coefficients and contact angle coefficients are correlated to the wheel, reflecting the effects of wheel dimensions and lugs. The parameter identification error caused by the simplified mod- els is decreased effectively, and mechanical parameters of the lunar soil can be identified comprehensively. This approach can be used for estimating soil property and it can improve prediction precision of wheel-soil interaction mechanics model.