The scope of this project is to investigate the feasibility
of a self-balancing system using a fuzzy logic controller. This project
serves as a prelude to a master’s project to be started in May 2005.
The project, entitled the Mechanical Legs, is a self-balancing system
consisting of two legs and a representative upper body mass above them
connected by way of a hip joint.
This system is to mimic the human body in simplistic
form. The real trick is to keep the center of mass balanced above the
legs. This problem is classically referred to as the inverted pendulum
problem. The system should be able to respond to small perturbations,
such as lightly pushing the body in a particular direction with a finger,
by rebalancing itself. In addition, the unit will have the ability to
drive forward or reverse by remote control. A host computer running a
fuzzy logic controller will serve as the brains of the unit.
The balancing system is a control system which begins
at the Fuzzy Logic Controller (FLC). The FLC is the brains of the system,
which outputs a duty cycle to the control system hardware (the PIC microcontroller)
where the duty cycle is modified so that the motor controller can be
told whether to make the motor move forward or reverse at the appropriate
speed. A tilt sensor provides a signal to the microcontroller where
it is modified so that it can be read by the host computer. The mechanical
part is the physical realization of the body, the size, weight, and
shape of which plays a part in the determination of hardware. The block
diagram of the system is shown in the figure below. From this it can
be seen how each block is related. Please click on the part for which
you want more information.