仿生沙基机器人二维运动C形腿数学模型的建立及自适应控制的研究

The light and flexible sand Robert can not only complete the exploration of the desert, it can also be applied to explore the surface of the moon. Since the current slip rate of sand Robert is still too high when it walks on soft sand，this project focuses on the mathematical model of wheel-sand interaction of the two dimensional motion C shape leg according to the Terradynamics and the principle of bionics as well as the discrete element method, and the Terradynamics test achieved by interaction between wheel and sand, it revealed the mechanism of the wheel-sand interaction and provides the theory basis to optimize the two dimensional motion C shape leg. By using the intelligent information fusion method of Support Vector Machine（SVM）and the fuzzy reasoning, and the experiment exploration method, it studies the sand Robert online identification and explores the maximum control strategy of two dimensional motion C shape leg bionic six-legged robot, and solves the self-adaptation control. Its innovation point is as follows: ①Change the original one-dimensional C shape legs into translational and rotational two-dimensional motion C shape legs by using the slippery course and the crank connecting rod mechanism without any increase in drive motor cases, which can overcome the shortcomings of too high slip rate of the C shape legs on the soft sand, and improve the ability of the robot to get through the soft sand. ②Establish the mathematical model of two dimensional motion C shape legs wheel-sand interaction.③Regards the sand and the Robert as a whole system, and establish the ground system based on SVM according to the experimental data of training SVM ground identification model set.

轻巧灵便的沙基机器人不但可以完成沙漠的探险，还可用于月球表面探测。鉴于目前沙基机器人在松软沙基行走时滑转率仍然过大的问题，本项目根据地面动力学理论和仿生学原理，结合离散元法，通过轮与沙作用的地面动力学试验，着重研究二维运动C形腿轮与沙作用的数学模型，揭示轮沙作用的机理，为优化设计二维运动C形腿提供理论依据。用基于支持向量机与模糊推理的智能信息融合方法，运用实验探索法，研究沙基在线辨识，探索二维运动C形腿仿生六足机器人最优控制策略，解决其自适应控制问题。创新点：①利用滑轨和反装曲柄连杆机构，在不增加驱动电机情况下，将原来只有一维运动的C形腿变为既有平动又有转动的二维运动C形腿，可克服C形腿在松软沙基中滑转率过大的缺点，提高机器人通过松软沙基的能力；②建立二维运动C形腿轮与沙作用的数学模型；③将沙基与机器人看成是一个系统，用实验数据训练支持向量机地面辨识模型集，建立基于支持向量机的地面辨识系。

轻巧灵便的沙基机器人不但可以完成沙漠的探险，还可用于月球表面探测。鉴于目前沙基机器人在松软沙基行走时滑转和沉陷仍然很大，本项目以地面动力学理论和仿生学原理为基础，借鉴地面力学的研究方法，以实验为中心展开了一系列研究，着重研究沙基机器人轮足与沙相互作用的数学模型，揭示轮与沙作用的机理，同时为优化设计沙基机器人轮足提供理论依据。主要研究内容、重要结果如下：①研制出了直接测量C形腿轮缘正应力和剪应力的测试方法和实验装置；②研制出了直接测量C形腿挂钩牵引力和地面支持力的测试方法和装置；③构建了C形腿环形实验平台；④分析了C形腿轮缘正应力和剪应力分布及变化规律并给出了解析表达式；⑤测得了C形腿挂钩牵引力和地面支持力，建立了C形腿的动态模型及轮缘正应力和剪应力与挂钩牵引力的关系方程；⑥ 设计出了类C形腿沙基自适应机器人。本项目所建立的C形腿环形试验平台，可测出C形腿轮缘应力分布变化规律和C形腿轮轴获得的挂钩牵引力和地面支持力，该试验装置类似于空气动力学中的风洞试验装置，各种沙基机器人轮足可以安装在环形试验平台上，测量其与沙基作用时的动力学特性，为解决沙基机器人在松软沙基上行走的滑转和沉陷问题提供理论依据和技术上的支持。