Fuzzy controller for an autonomous bicycle system

In this thesis we implemented a fuzzy logic controller for stabilisation of an autonomous bicycle. We revised concept of fuzzy logic and researched possibilities of its usage. For our bicycle model, made from Lego NXT pieces, we defined its physical model and with simulation searched for velocity, at which model is most stable. We then implemented PID and fuzzy controller and compared their implementation and performance by analyzing data gathered from our test runs. We implemented a bluetooth link with computer, which can be used for steering the bicycle remotely by user. After analyzing our data we concluded that fuzzy logic in our case provides some advantages and flexibility that PID can not offer. Implementation and understanding of multiple input systems is greatly improved using fuzzy logic. At the end we have gathered some possible future additions to our system