In this paper, a dynamic model of two-wheeled balancing robot has been created, and the two types of FLC has been designed. The Mamdani methods used on both FLC. The first FLC uses pendulum tilt angle theta (θ) as the input and it requires the motor torque to keep the robot remains balanced as the output. The second FLC uses two inputs, the first input is theta (θ) and the second input is the change in the value of theta (θ) which is the output torque of the motor. The second plant model and FLC built by using Matlab Simulink. The first case is one input using 5 membership functions (mf). The second case is two inputs using the 5 and 7 mf. The characteristics and effects of the changes in the input and mf have been simulated in the Simulink and compared. By expanding the number of the inputs can reduce motor specification required in balancing robot. Meanwhile, by increasing the number mf, it can improve the performance of the controller much faster to reach the settling time.

Navigation equipment specifications differs in the update rate, accuracy, budget, reliability, size and mass. In some applications in order to meet navigation system requirement, a dead reckoning equipment i.e. Inertial Navigation System INS is could be integrated with one or many position fixing equipment, i.e. Global Navigation Satellite System GNSS. INS and GPS have different benefits and drawbacks, and they complement each other when integrating them to provide a navigation solution with higher bandwidth, and long-term and short-term accuracy. This research investigates the performance of an integrated system GPS/INS(MEMS) when changing the algorithm update rate, and compare between different integration algorithm namely loosely and tightly integration.