Non Negative Linear Combination Grasp Force Optimization

In order to improve the adaptability of multi-fingered grasp, a force optimization method was proposed to calculate the contact force of each finger, and an adaptive force tracking impedance control algorithm was designed to improve the compliance during contact. The force optimization was realized through a non-negative linear combination method for initial force calculation, combined with gradient algorithm to optimize the contact force. This method was put forward according to the property that the set of grasp force is a convex cone. First, the external force was decentralized as a linear combination of 12 unit force vectors, and then, the initial basic feasible solution for every unit force vector was calculated offline, at last, the initial value of the contact force was obtained by non-negative linear combination; based on the above calculation, the gradient algorithm was called to optimize the contact force online. The force-tracking impedance control method was applied by considering the contact compliance and the force-tracking precision. The indirect adaptive algorithm was applied to estimate the parameters of the object. Take the three fingered point contact with friction for example, the force can be optimized efficiently and the impedance algorithm can track the desired contact force with a relatively high accuracy.