基于复合传感器阵列的多火源融合定位方法研究

项目名称： 基于复合传感器阵列的多火源融合定位方法研究

项目编号： No.61503174

项目类型： 青年科学基金项目

立项/批准年度： 2016

项目学科： 自动化技术、计算机技术

项目作者： 王国勇

作者单位： 洛阳理工学院

项目金额： 22万元

中文摘要： 基于复合传感器的火源定位是现代突发公共事件防范与快速处置中高层和地下建筑消防技术发展的基础。由于受多火源、非静风空间和非高斯数据干扰等因素影响，加之大型复杂空间内传感器阵列布局的多样性，使得现有室内火源定位方法性能将显著降低，严重时甚至会出现方法失效。本项目以多类复合式传感器阵列探测系统为基础，以多源信息融合技术、模糊推理技术、非静风空间建模、统计分析理论为手段，开展复杂环境下适应多种复合传感器阵列组合模式的多火源目标融合定位方法研究，重点解决如下关键问题：1）复合式传感器阵列模式及非静风空间情形下的火源目标关联问题，2）在方位估计统计特性建模不精准、火源动态扩散、非静风空间等情形下的多火源融合定位问题，3）研究多火源融合定位性能评估方法及其对定位系统性能的反馈优化方法。以期获得一系列基于复合传感器信息融合技术的多火源探测定位方法，为大型空间建筑的智能火灾探测与消防安全提供一定的理论储备。

中文关键词： 复合传感器阵列；目标关联；多火源融合定位；非静风空间建模；性能评估

英文摘要： Composite sensor fire positioning is the foundations of modern public emergency prevention and rapid treatment of high-rise and underground building fire control technology and equipment.Due to the multiple fire sources, non-static space and non-Gaussian interference and other factors, combined with the diversity of sensor array in the large complex space layout, performance of existing indoor fire location will significantly reduce, and even failure can occur.This project is based on many types of complex sensor array detection system, leveraging multi-source information fusion theory, fuzzy inference, statistical analysis, to conduct the study of fire target positioning which adapt to multi detector array combination in complex environment, focus on the following key questions:1)Fire target correlation problem under compound sensor array mode and non-static space condition,2)Multiple fire fusion positioning problem when direction estimation of statistical modeling is not accurate, fire diffuse dynamically, or non-static space,3) research on performance evaluation method of multiple fire fusion positioning and optimization methods of the position system performance feedback.Hoping to obtain a range of fire detection positioning method based on multi-sensor information fusion technology, and so as to provide intelligent fire detection and fire safety of large buildings with theoretical reserve.

英文关键词： Composite sensor array;target association;Multi source fusion positioning;Non static wind space modeling;Performance evaluation