Share Email Print
cover

Proceedings Paper

Error compensation for truck scale based on complex BP neural networks
Author(s): Haijun Lin; Zhaosheng Teng; Rangzhou Liu; Dan Zheng; Jinbao Yang
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

Truck scale is widely applied to many fields such as storage, trade, transport, communication, industry and mine. Conventional method of error compensation for truck scales is fussy and the accuracy of weighing result is low. An error compensation method based on complex BP neural networks (CBPNNs) is proposed in this paper. Considering the character of error in scale's different weighing zones, the sub-BPNNs are constructed, and each one is used to compensate the weighing error of a weighing zone. The adaptive choice network is founded to automatically choose suitable sub-BPNN, and then the chosen sub-BPNN optimally compensates the error of different weighing ranges. Emulational experiments show that the weighing error of the truck scale with this proposed method is less than that of using a single RBFNN to compensate the error of total weighing range, and all results of field verification show that the maximum eccentric error is -4kg and the maximum repeatability error is +6kg, which is far less than that of the 3th class scales defined by International Recommendation OIML R76 "Nonautomatic Weighing Instruments".

Paper Details

Date Published: 28 December 2010
PDF: 8 pages
Proc. SPIE 7544, Sixth International Symposium on Precision Engineering Measurements and Instrumentation, 75445V (28 December 2010); doi: 10.1117/12.885366
Show Author Affiliations
Haijun Lin, Polytechnic College of Hunan Normal Univ. (China)
Zhaosheng Teng, Hunan Univ. (China)
Rangzhou Liu, Hunan Univ. (China)
Dan Zheng, Hunan Univ. (China)
Jinbao Yang, Polytechnic College of Hunan Normal Univ. (China)


Published in SPIE Proceedings Vol. 7544:
Sixth International Symposium on Precision Engineering Measurements and Instrumentation
Jiubin Tan; Xianfang Wen, Editor(s)

© SPIE. Terms of Use
Back to Top