基于BP神经网络的车辆动态称重技术  

A Dynamic Weighting System for Vehicle Based on BP Neural Network

作　　者：熊少康 王凌川[2] 章家岩[1] 郭许林[1] 冯旭刚[1] 

机构地区：[1]安徽工业大学电气与信息工程学院,安徽马鞍山243032 [2]马鞍山钢铁股份有限公司物资公司,安徽马鞍山243000

出　　处：《安徽工业大学学报（自然科学版）》2014年第1期76-79,共4页Journal of Anhui University of Technology（Natural Science）

基　　金：安徽省教育厅自然科学基金重点项目(KJ2013A054)

摘　　要：车辆动态称重对于控制超载运输或物流仓储等要求连续称重的场合具有重要作用。针对车辆动态称重的技术特点和工艺要求,分析动态称重过程的动力学模型,确定位移信号阶跃响应的最大超调量与车重之间的单调函数关系,提出1种基于BP神经网络预测最大超调量的动态称重计算方法,并在MATLAB环境下进行仿真分析,通过系统的学习和训练,完成了动态称重过程的车辆静态重量计算。结果表明,提出的动态称重算法可有效地提高车辆动态称重系统的称量精度,具有较高的实用价值。Weigh-in-motion plays an important role in the occasions which need continuous weighing such as the control of vehicle overloading transportation, logistics warehousing and so on. Aiming at the technical characteristics of vehicle dynamic weighing and the technological requirements, analysis of the dynamic model of the dynamic weighing process, determine the monotone function relationship between the weight and the maximum overshoot of a displacement signal step response. A dynamic weighting calculation method of forecasting the maximum overshoot based on BP neural network is proposed. The simulation and analysis are carried out in MATLAB. Through learning and training of the system, the vehicle static weight with dynamic weighing process calculation was calculated. The result shows that the proposed dynamic weighing algorithm can effectively improve the weighing accuracy of the vehicle dynamic weighing system and with a high practical value.

关 键 词：动态称重 最大超调量 神经网络 MATLAB仿真 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]