Optimization-based hexapod robot locomotion generation

The purpose of this paper is specifically to provide a more intelligent locomotion planning method for a hexapod robot based on trajectory optimization, which could reduce the complexity of locomotion design, shorten time of design and generate efficient and accurate motion.

The authors generated locomotion for the hexapod robot based on trajectory optimization method and it just need to specify the high-level motion requirements. Here the authors first transcribed the trajectory optimization problem to a nonlinear programming problem, in which the specified motion requirements and the dynamics with complementarity constraints were defined as the constraints, then a nonlinear solver was used to solve. The leg compliance was taken into consideration and the generated motions were deployed on the hexapod robot prototype to prove the utility of the method and, meanwhile, the influence of different environments was considered.

The generated motions were deployed on the hexapod robot and the movements were demonstrated very much in line with the planning. The new planning method does not require lots of parameter-tuning work and therefore significantly reduces the cycle for designing a new locomotion.

A locomotion generation method based on trajectory optimization was constructed for a 12-degree of freedom hexapod robot. The variable stiffness compliance of legs was considered to improve the accuracy of locomotion generation. And also, different from some simulation work before, the authors have designed the locomotion in three cases and constructed field tests to demonstrate its utility.