Trajectory optimization inspired pneumatic locomotion on compliant terrains

Meyer, Jacques (2022-04)

Thesis (MEng)--Stellenbosch University, 2022.

Thesis

ENGLISH ABSTRACT: In order to achieve true autonomy, robots have to be able to handle complex and rough terrain generally found outside of the lab. Legged robotics has become the focal point in recent years, aided by the developments in trajectory optimization methods. However, a major problem in legged robotics is dealing with hybrid contacts with different terrain types. The difference in dynamics due to the interaction between the foot and the ground makes it increasingly difficult to design controllers that successfully execute on multiple surfaces. This work investigates trajectory optimization methods for a pneumatically actuated mono-pod on rigid and compliant terrain. Trajectory optimization was utilized to obtain trajectories for acceleration, steady-state and deceleration hopping on compliant terrain, as well as rigid terrain surfaces. For the compliant terrain trajectories a novel method was developed to model the specific characteristics of the compliant terrain. Trajectories were generated using this method for two different compliant terrain types, namely: rough gravel and fine gravel. To mimic the pneumatic actuation in the trajectory optimization problem, a simplified mathematical model was developed to accommodate the bang-bang force of the pneumatic actuator. This model used complementarity constraints and node bucketing techniques to mimic the behaviour of a real pneumatic actuator with damping and delay. Once these methods and models were implemented, trajectories were executed, in open-loop, on a fixed body robotic platform that was designed and built for this thesis. The executions were compared to the trajectory results. The rigid terrain trajectories were executed successfully on a hard surface, but failed on gravel surfaces. The compliant terrain trajectories executed successfully on gravel surfaces, indicating that the method developed to model compliant terrain is a more accurate representation of the gravel surfaces compared to the rigid terrain trajectories. After these results showed that the methods used to describe the compliant terrain proved to be accurate, a free body mono-pod robot and support rig was designed and built. The support rig limited the movement of the mono-pod to the sagittal plane to mimic the limitations of the trajectory optimization model. Acceleration, steady-state and deceleration trajectories were generated for the free body mono-pod on compliant terrain surfaces and a rigid terrain surface. From these trajectories a controller was designed with the main sources of feedback being the height of the robot and the angle of the free moving body of the robot. The free body mono-pod robot used the controller to execute hopping from rest back to rest with three steady-state hops in between. For each terrain type the controller was adjusted based on the generated trajectories. The results show successful execution of the trajectories on all terrain types using the controller. Lastly multi-surface hopping was executed on the mono-pod robot platform. The controller was adjusted to hop from a hard surface to a compliant surface and executed these trajectories successfully.

AFRIKAANSE OPSOMMING: Om ware outonomie te bewerkstellig, moet robotte komplekse en rowwe terreine, wat gewoonlik buite die laboratorium voorkom, kan hanteer. Been aangedrewe robotika het die afgelope paar jaar ’n fokuspunt geword in die literatuur, aangehelp deur ontwikkelinge in trajek-optimaliseringsmetodes. ’n Groot probleem in been aangedrewe robotika is egter die hantering van verskillende terreintipes. Die verskil in dinamika as gevolg van die interaksie tussen die voet en die grond maak dit steeds moeiliker om beheerders te ontwerp wat suksesvol op verskeie oppervlaktes uitgevoer kan word. Hierdie werk ondersoek optimaliseringsmetodes van trajekte vir ’n pneumaties aange drewe eenbenige robot op stewige en sagte terrein. Trajek-optimalisering is gebruik om versnelling-, bestendige- en vertraging-trajekte te genereer wat op sagte terreine sowel as stewige terreinoppervlakke spring. Hierdie trajekte is uitgevoer op ’n robot platvorm met ’n vaste liggaam en is vergelyk met die trajek resultate. Die stewige terrein trajek is suksesvol uitgevoer op ’n harde oppervlak, maar het op ’n gruisoppervlak misluk. Die sagte terrein is suksesvol uitgevoer op twee gruisoppervlaktes, wat aandui dat dit ’n meer akkurate voorstelling van die gruisoppervlakke is in vergelyking met die rigiede terrein trajek. Om die pneumatiese kragte in die tajekoptimalisering na te boots, is ’n vereenvoudigde wiskundige model ontwikkel om die drukstootbeweging van die pneumatiese aandrywer te akkommodeer. Hierdie model het komplementariteitsbeperkings en knooppunte gebruik om die gedrag van ’n werklike pneumatiese aandrywer na te boots. Nadat hierdie resultate getoon het dat die metodes wat gebruik is om die terrein te beskryf akkuraat is, is ’n eenbenige robot en ’n ondersteuningsplatvorm vir ’n vrye liggaam robot ontwerp en gebou. Die ondersteuningsplatvorm het die beweging van die eenbenige robot tot die sagittale vlak beperk om die beperkings van die trajekoptimaliseringsmodel na te boots. Versnelling-, bestendige- en vertragings-trajekte is gegenereer vir die eenbenige robot met ’n vrye liggaam op sagte terrein en op stewige terrein. Uit hierdie trajekte is ’n beheerder ontwerp, met die belangrikste bronne van terugvoer die hoogte van die robot en die hoek van die vry bewegende liggaam van die robot. Die eenbenige robot het die beheerder gebruik om van rus na rus te spring, met drie bestendige hoppe tussenin. Vir elke terrein tipe is die beheerder aangepas op grond van die gegenereerde trajekte. Die resultate toon ’n suksesvolle uitvoering van die trajekte op alle terreintipes met behulp van die beheerder. Laastens is ’n twee-oppervlakte-hop uitgevoer op die mono-pod robotplatform. Die beheerder is aangepas om van ’n harde oppervlak na ’n sagte oppervlak te spring en hierdie trajekte was suksesvol uitgevoer.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/124602
This item appears in the following collections: