Improving the Performance of an Ultrasonic Sensor Using Soft Computing Techniques for 2D Localization

Ultrasonic sensors (HCSR04) are used for distance measurements whose range extends up to a distance of 4 m. These kinds of sensors are found being used in car reverse parking mechanisms and also helping blind people in navigation. But, there always exists a trade-off between distance and accuracy. This paper presents a method to improve the accuracy of an ultrasonic sensor using soft computing techniques for 2-D indoor localization. Advanced neuro fuzzy inference system (ANFIS) was used to minimize the error in distance measurement. Finally, the simulated FIS model was implemented in hardware for real-time evaluation. The error analysis showed that the distance estimation accuracy was improved by 77%.