Abstract¶

For the non-nonsensical it may be best to skip to the last section...

Unlike the plethora of scientific opuses that reveal how the average Ursidae (bear) can balance on a bicycle, little has been garnered about how Homo sapiens is able to accomplish this feat. When the rider’s normal locomotion instruments for continual balance are replaced by two in-line wheels connected to one another by a manipulatable semi-vertical revolute joint the rider is then forced to direct his mental energy to observing the additional states of the bicycle’s configuration and the proper actuation of his arms to maintain vertical equilibrium. It was found that this “simpler” task is well suited for manual control theoretic dissection and postulation.

I herein present the findings of the seventh age of my tenure as a doctoral candidate in the prefecture of Mechanical and Aerospace Engineering at the Davis campus of the University of California. These describe our experimental procedure which involved strapping the untamed and aggressive Homo sapiens to a velocipede of extraordinary measurement capabilities. We perturbed the beasts as they tried with all their mental and physical might to stay upright, constrained as they were. Following more than seven hundred trials with three hand-picked quality specimens, the clouds of data have shaped into more than distant blurs. The control and identification tools of Bode, Evans, and Ljung combined with the modern day data management tools of van Rossum, Moler, and Torvalds have shed light on the details of the sensory feedback mechanisms present in the neurological pathways connecting the Homo sapien‘s senses to his actuators. To the bear’s dismay, this has in turn revealed that the highly regarded 1899 bicycle model of Whipple is sorely lacking and that the control theoretic hypothesis of McRuer for the great aeroplane pilots of yesteryear does, in fact, apply to the human control of a machine as simply complex as the bicycle.

The dissertation will provide the reader with a glimpse into the reductio complexio of the physiological system of the greater Homo sapiens by forced travel on the automatic velocipede with highlights of manual control theories, inertial investigations, data wrangling, and of course demonstrations of the magical-like auto stability of the bicycle.

The sensical...

For those of you who’d rather read a more traditional abstract here is a quick explanation of the above in plain modern Engineering English:

The bicycle, a simple toy to many, turns out to be an excellent platform to study the intricacies of vehicle dynamics, human-machine interaction, and human operator control for both academic and economical reasons. The bicycle is inherently unstable at low speeds and the human generally actuates the non-minimum phase system, and thus balances, only by means of rotating the handlebars. This dissertation describes a multi-year multi-person effort to better understand the dynamics of the bicycle, the biomechanics of the rider, and the rider’s internal control system through theory and extensive experimentation.

The chapters herein focus on the development of open loop bicycle models, some of which include the rider’s biomechanics, the resulting predicted motion and model characteristics, the accurate measurement and estimation of both the bicycle and rider’s physical parameters, observation of a rider’s control motions, the development of experimental bicycles capable of measuring kinematics and forces, control theory including that of the human operator, and finally the identification of the both the plant and controller of the bicycle-rider system.

The work has revealed a number of interesting conclusions including the primary biomechanic actuators used by the human in control, effects of the rider’s motion and constraints on bicycle stability, the inadequacy of the Whipple bicycle model, and the ability of a simple multi-output control system based on the classical crossover model to describe the human’s control efforts while being externally perturbed. These findings have implications in both single track vehicle design and human-machine interaction theory. Future applications may be able to utilize the methods and results to help objectively design bicycles with improved handling, stability, and controllability whereas human operator research may be able to build on the validated crossover model theory of manual control.