机器人多指手抓取爆炸物能力研究  

Investigation into the Ability of Multi-Fingered Dexterous Hand to Grasp an Explosive

作　　者：莫海军[1] 林志生[1] 

机构地区：[1]华南理工大学机械与汽车工程学院,广东广州510640

出　　处：《华南理工大学学报（自然科学版）》2015年第7期124-129,共6页Journal of South China University of Technology(Natural Science Edition)

基　　金：国家自然科学基金资助项目(51175182)~~

摘　　要：应用力螺旋理论建立了多指手抓取矩阵与外力螺旋之间的关系,对多指手抓取爆炸物能力进行研究.首先建立多指手抓取爆炸物的力学模型,确定影响抓取的参数;然后以最大抓取重量为目标函数,建立满足力封闭约束条件下的抓取数学模型,并采用神经网络和优化方法对多指手抓取爆炸物进行了仿真,研究多指手不同抓取位置与爆炸物重量之间的对应关系,获得了多指手抓取爆炸物时最有利的抓取位置.On the basis of the force screw theory , the relationship between the grasp matrix of multi-fingered dexter-ous hand and the external force screw is established to investigate the ability of multi -fingered dexterous hand to grasp an explosive .Firstly, a mechanical model of multi-fingered dexterous hand which can grasp an explosive is constructed and the parameters influencing the grasp are determined .Then, with the maximum grasp being the goal, a mathematical model of the grasp is constructed under the constraint of force closure .Finally, through the neural network and optimization methods , the multi-fingered dexterous hand grasping an explosive is simulated , so as to reveal corresponding the relationship between the grasp position and the explosive weight .Thus, the best grasp position is obtained .

关 键 词：机器人多指手 抓取能力 神经网络 爆炸物 

分 类 号：TG242[金属学及工艺—铸造]