Fuzzy Control for Cyclist Robot Stability Using FPGAs

Abstract

This paper presents a fuzzy controller implementation in FPGA (Field Programmable Gate Array) for a robot that rides a bicycle using the well-known Acrobot model. The overall system presents a hardware/software codesign approach and it was achieved by means of a Microblaze FPGA embedded processor and a fuzzy controller, which was implemented directly in hardware. Both the microprocessor and the controller are connected via the Fast Simplex Link - FSL bus. The proposed design methodology involves firstly the fuzzy controller design in software for simulation and testing issues, taking into account the mathematical model of the plant. Afterwards, the controller was synthesized to the hardware description language VHDL using the Xfuzzy 2.0 tool. The fuzzy controller has 2 modules, each one producing a torque control variable. The first module receives both the position and angular speed of the first link of the Acrobot system whereas the second module receives the position and angular speed of the second link. The final torque variable is calculated in the Microblaze taking into account two gains. Each gain represents a priority that is applied to each fuzzy module. These gains were experimentally calculated through several simulation executed in the Matlab computational environment.