A Simple Ultrasonic GPS System for Indoor Mobile Robot System using Kalman Filtering

This paper introduces a simple indoor GPS system using ultrasonic sensor. Conventional ultrasonic GPS systems require at least 3 transmitters, which restricts update time of location information by 5 cycles/sec so that the ultrasonic waves from the 3 transmitters do not overlap. This system consists of 1 transmitter having ultrasonic and RF and 2 receivers. The transmitter irradiates RF and ultrasonic, and the receivers calculate corresponding distance with reference to the RF signal. The 2 distance values are used for determining the location of the transmitter by trigonometrical functions. Due to the characteristics of ultrasonic sensors, noise occurs in sensor values by surrounding temperature or obstacle. Location error is minimized by prediction and correction of the noise with linear Kalman filter. In order to prove the effectiveness of this system, experiments were carried out in a room dimensioned by 3.5 m*2.2 m, wherein the location error showed 2 cm max