신경회로망과 필터뱅크를 기반한 AGV용 가이드센서 설계

다국어 초록 (Multilingual Abstract)

In order to improve the accuracy and path detection performance of the guide sensor that is used in the AGV, this study is proposed the signal processing method based on filter bank and neural networks, and its performance were evaluated using by an implemented hardware device. In the results of the performance evaluation, the maximum error of the neural network’s output for the left branch path was 1.8㎜, and the error for the right path was 0.2 ㎜. And we compared and analyzed the performance of the proposed method and the method using the fuzzy inference method previously studied. As a result, the proposed method showed two times more precision than the previous method(fuzzy inference system). Therefore, if the proposed method in this study is applied to the design of the guide sensor, the problems of AGV driving that are the stability and out-of-track is thought to be improved.

참고문헌 (Reference)

1 박문호, "양극성 자기유도센서의 성능 향상을 위한 퍼지 추론 시스템" 한국지능시스템학회 24 (24): 58-63, 2014

2 조현학, "실외 자기유도 무인운반차를 위한 자기 위치측정 장치 개발" 한국지능시스템학회 24 (24): 259-264, 2014

3 Iris F. A. Vis, "Survey of research in the design and control of automated guided vehicle systems" 170 (170): 677-709, 2006

4 N. J. Flieget, "Multirate digital signal processing : multirate systems, filter banks, wavelets" Chichester Wiley 1994

5 logiworks, "LWM50-1 magnetic tape manual"

6 bhtech, "Introduction of production"

7 R. Chen, "Design of magnetic navigation automatic guided vehicle system" 1311 (1311): 012040-, 2019

8 L. Gao, "Automatic recognition about Pallet based on tracking algorithm of lidar and SLAM" 196-201, 2016

9 M. W Gardner, "Artificial neural networks(the multilayer perceptron)—a review of applications in the atmospheric sciences" 32 (32): 2627-2636, 1998