Bicycle robot has high-speed and high-energy efficiency on off-road terrains, but is unable to keep balance at low speed and run on uneven terrain. So we design a wheel-leg hybrid control strategy to provide assistive torque to balance the bikebot at low speed or off-road terrain conditions.
A model predictive control is designed for the leg assistive actuation to take advantage of the leg-ground interaction force and balance torque.
The following video shows the created simulation scenario in RaiSim.
The following figure is schematic of control system.
At first build the whole system dynamics model and get the state space equation.
We further use the mpc form to calculate the proper leg-ground interaction force and use qpOASES solver to make the computation faster.
First generate the 3D model of bikebot by solidworks and then use algorithm to control the robot in Raisim.
Finally we create rugged terrain with the height as a function of the xy coordinate as ℎ=|0.07sin(7𝑥)−0.07cos(4𝑦+2)| m.
