摘要
针对管廊环境影响指标的多样性、测量数据的波动性以及传感器可信度带来的不确定性,提出一种基于多传感器数据融合的管廊环境评估方法。首先,构建了以舒适性和安全性2个综合指标及6个环境指标为主体的管廊环境评估指标体系;其次,考虑到传感器测量数据的波动性和不确定性,将测量数据用区间数表示,经箱线图将异常值剔除后,与评估等级特征值比较,得出各环境指标的等级置信度;然后利用证据推理方法将指标等级置信度转化为mass函数并按组合规则融合成综合指标等级置信度,根据决策准则,判断得到管廊环境等级状态。最后对某市地下综合管廊两分区环境进行评估,与模糊综合评价法相比,所提方法的评估结果更符合现场实际情况。
Considering the diversity of the environmental impact indexes, the volatility of the measured data and the uncertainty caused by the sensor reliability, an environmental assessment method for underground pipe gallery based on multi-sensor data fusion is proposed. Firstly, an environmental assessment index system for underground pipe gallery based on two comprehensive indexes of comfort and safety and six environmental indexes was constructed. Secondly, considering the volatility and uncertainty of sensor measurement data, the measurement data of each sensor is represented by interval number. After removing the outlier data by using the box-plot and comparing with the features of evaluation grade, the belief degrees grade of each environmental index is acquired. Thirdly, the belief degrees grade of index is transformed into mass function by evidence reasoning and fused into the comprehensive index belief degrees grade by the combining rule. According to the decision rule, the environmental condition of the underground pipe gallery is judged. Finally, the environment of two zones of underground pipe gallery in a city is evaluated. Compared with the fuzzy comprehensive evaluation method, the evaluation results of the proposed method are more in line with the actual situation.
作者
杨辉
夏贤隆
陆荣秀
YANG Hui;XIA Xian-long;LU Rong-xiu(School of Electrical and Electronic Engineering,East China Jiaotong University,Nanchang 330013,China;Key Laboratoryof Advanced Control&Optimization of Jiangxi Province,Nanchang 330013,China)
出处
《控制工程》
CSCD
北大核心
2020年第10期1665-1671,共7页
Control Engineering of China
基金
国家自然科学基金资助项目(61733005,61863014,61673172)。
关键词
环境评估
证据推理
多传感器数据融合
不确定性
Environmental assessment
evidence reasoning
multi-sensor data fusion
uncertainty
