Distance Measurement Method Using Neural Network Learning of Microwave Reflection Signals

In this paper, we present a novel non-contact distance measurement method using microwave reflection signals and artificial neural networks. Based on data learning, this method can effectively predict the distance of an object placed in a complex environment. In particular, by using a two-step neural network, we propose a method of maintaining precision while reducing the data used for training. Through an experimental test, microwave reflection signals for each distance are acquired and the two-step neural network is trained. Finally, the distance is estimated from the microwave reflection signal measured for an arbitrary distance. Using the proposed method, we have successfully demonstrated the distance measurement of an object placed in an underwater environment.

I. Introduction

As the demand for precise distance measurement continues to increase, distance measurement technology is advancing. The non-contact measurement method that does not require physical contact is generally used more than the contact measurement method because the inspection method is simple and does not cause damage to the measurement object. Non-contact measurement methods include methods using an ultrasonic sensor, an infrared sensor, radar, lidar, and a stereo camera.