哈尔滨工程大学学报2016，Vol.37Issue(5)：724-731,8.DOI:10.11990/jheu.201501037
采用P ELM的阵列式皮带秤称重误差建模与补偿
Measurement error modeling and compensation for array belt weigher using process extreme learning machine
摘要
Abstract
To further improve and maintain the dynamic measuring accuracy of an electronic belt weighing system o⁃ver the long term, the weighing force error was investigated by comprehensively considering the potential sources of error in the belt⁃weighing system. A weighing force error model for the single⁃roller belt weigher was established. An internal force theory for an array⁃type belt weigher was derived that suggests that the errors in weighing accuracy of the weighing system are mainly associated with the conveyor belt tension and idler misalignments on both ends of the idler roller set. The process extreme learning machine, where the process neurons are in the output layer, not in the hidden layer, is put forward by improving the process neural network and introducing the extreme learning machine. The extreme learning machine combined with the internal force theory were employed to compensate for the meas⁃urement error of the array⁃type belt weigher. Finally, experimental data confirmed the success of the method and showed that the weighing accuracy of an array⁃type belt weigher can reach ± 0. 1% with the error compensation method. A new method has been found for error compensation of weighing continuous bulk materials.关键词
变分原理/阵列式皮带秤/误差补偿/过程神经网络/极限学习机Key words
variational principles/array belt weigher/error compensation/neural network/extreme learning ma-chine ( ELM)分类
机械制造引用本文复制引用
朱亮,吴绍锋,何非,李东波,童一飞,袁延强..采用P ELM的阵列式皮带秤称重误差建模与补偿[J].哈尔滨工程大学学报,2016,37(5):724-731,8.基金项目
科技型中小企业技术创新基金（13C26213202062）． （）
