基于人工神经网络的驾驶室人机界面匹配度评价权重分配方法  

Allocation Means of Weights for the Matching Extent Evaluation of Man-Machine Interface of Driver Cab Based on Artificial Neural Network

作　　者：白雅娟[1] 张浩[2] 张海燕[3] 

机构地区：[1]北京特种车辆研究所,北京100072 [2]中北大学机械工程与自动化学院,山西太原030051 [3]中国北方车辆研究所,北京100072

出　　处：《中北大学学报（自然科学版）》2010年第5期481-485,共5页Journal of North University of China(Natural Science Edition)

摘　　要：对于任何综合评价系统,确定评价指标的权重分配系数都是非常重要的,它直接影响评价结果的准确性.针对特种车辆人机界面匹配度综合评价,设计了权重分配系数验证修正实验,基于人工神经网络理论,对主观赋权的权重系数进行了修正.实验数据分析结果表明:修正后的权重系数更加符合驾驶员的实际感受.It is very important to determine the weight allocated coefficient of evaluation index for any synthetical evaluation system, which affects the accuracy of evaluation results. The verified and modified experiment of weight allocated coefficient was designed to evaluate synthetically matching extent of the special vehicle＇s human-machine interface. Based on artificial neural networks theory, the subjective weight coefficient was modified. The experimental data show that the modified weights coefficient is more accordance with the practical experience of driver.

关 键 词：人工神经网络 综合评价 权重系数 人机界面 

分 类 号：TP311.5[自动化与计算机技术—计算机软件与理论] TB18[自动化与计算机技术—计算机科学与技术]