Research on an expected coal feed rate model for train rapid quantitative loading systems based on multi-source information fusion
基于多源信息融合的火车快速定量装车系统期望给煤量模型研究

摘要

火车快速定量装车系统在C70/C80大标载车型装车作业时，缓冲仓煤量存在短时波动大、供需失衡的问题。针对此问题，提出一种基于多源信息融合的期望给煤量模型。该模型首次将缓冲仓煤量、带式输送机缓存煤量、装车进度、列车编组信息进行在线融合，构建了以时间为自变量、期望给煤量为因变量的分段解析函数，作为上游给料控制系统的实时跟踪目标。工业试验结果表明，期望给煤量模型较传统“仓位-阈值”策略，平均装车时间由65秒/节缩短至58秒/节，给煤机启停次数由3次/列降至1次/列，显著提升了装车效率与设备寿命。

Abstract

During the loading operation of C70/C80 large-rated-load freight cars with the rapid quantitative train loading system, the coal volume in the buffer bin suffers from large short-term fluctuations and supply-demand imbalance. To address this problem, an expected coal feeding rate model based on multi-source information fusion is proposed. For the first time, this model performs online fusion of the coal volume in the buffer bin, the cached coal volume on the belt conveyor, loading progress, and train marshalling information, and constructs a piecewise analytical function with time as the independent variable and the expected coal feeding rate as the dependent variable, which serves as the real-time tracking target for the upstream feeding control system. Industrial test results show that compared with the traditional "bin level-threshold" strategy, the expected coal feeding rate model reduces the average loading time per car from 65 seconds to 58 seconds and the number of start-stop cycles of the coal feeder from 3 times per train to 1 time per train, significantly improving loading efficiency and equipment service life.