Central Pattern Generators augmented with virtual model control for quadruped rough terrain locomotion

Abstract

Abstract-We present a modular controller for quadruped locomotion over unperceived rough terrain. Our approach is based on a computational Central Pattern Generator (CPG) model implemented as coupled nonlinear oscillators. Stumbling correction reflex is implemented as a sensory feedback mechanism affecting the CPG. We augment the outputs of the CPG with virtual model control torques responsible for posture control. The control strategy is validated on a 3D forward dynamics simulated quadruped robot platform of about the size and weight of a cat. To demonstrate the capabilities of the proposed approach, we perform locomotion over unperceived uneven terrain and slopes, as well as situations facing external pushes.

Modular control of limit cycle locomotion over unperceived rough terrain

“…A CPG-only control was partially successful on the randomized uneven terrain. As from our previous study on a stiff quadruped [18], we were expecting the open-loop control to perform badly, however, a 56% success rate was obtained. This partial success is due to the compliance, which prevents minor stumblings by passive deflection of the legs, and moderately self-stabilizes the roll and pitch oscillations.…”

“…This simulated quadruped is a detailed model of the Oncilla robot ( Figure 1) which will be used for a full validation in near future. This paper is an extension of our previous study [18] on a stiff torque controlled simulated quadruped.…”

“…As we have shown in our previous study [18], attitude control regularizes the role-pitch variations (RPV) and stabilizes them. So we set the k att such that it stabilizes the RPV (Figure 7).…”

“…The posture control feedback signal ξ p is the sum of the attitude and direction control feedbacks ξ att and ξ dir . 1) Attitude control: In [18] we used Virtual Model Control (VMC) [23] to convert posture control virtual forces to joint torques. VMC uses the Jacobian transpose method [24] to generate torques representing the desired virtual forces.…”

Cheetah-cub-S: Steering of a quadruped robot using trunk motion

“…Matos and Santos applied a CPG to a SONY Aibo to make omnidirectional locomotion possible by controlling each leg individually [9]. Ajallooeian developed a controller for a cat-sized quadruped robot and made turning motion possible by using ab-/adduction at the hip [10]. WildCat is another quadruped which uses ab-/adduction to [11].…”