【摘要】：現(xiàn)代建筑的設(shè)計(jì)隨著科技的發(fā)展愈加復(fù)雜化與大型化,而建筑火災(zāi)的發(fā)生往往具有突發(fā)性,由于缺乏有效的疏散指引以及疏散人員的恐慌心理導(dǎo)致疏散效率低下,因此如何對處在建筑火災(zāi)中的人員進(jìn)行合理高效的疏散也成為了研究重點(diǎn)。由于傳統(tǒng)的疏散方案僅以路徑最短或者時(shí)間最短這種單一目標(biāo)對疏散人員的路徑進(jìn)行規(guī)劃,而在實(shí)際建筑火災(zāi)人員疏散過程中,單一目標(biāo)由于其局限性往往不能反映出實(shí)際應(yīng)急疏散。為了保證安全且快速的疏散,需要對疏散人員所選擇的路徑上的多個(gè)目標(biāo)進(jìn)行決策。本文以全體疏散人員疏散時(shí)間、擁堵度,風(fēng)險(xiǎn)作為路徑?jīng)Q策的三個(gè)目標(biāo),構(gòu)建出相應(yīng)的多目標(biāo)應(yīng)急疏散模型,并以建筑的數(shù)據(jù)模型為基礎(chǔ),添加相應(yīng)的路徑參數(shù)構(gòu)建動(dòng)態(tài)路徑模型。由于基本粒子群算法對解決實(shí)際多目標(biāo)問題效果不是很好,將粒子群算法中適應(yīng)度函數(shù)進(jìn)行修改;粒子的位置和速度公式進(jìn)行修正以及使用線性慣性權(quán)值等方式優(yōu)化。將改進(jìn)后的多目標(biāo)粒子群算法為每個(gè)疏散人員的火條動(dòng)態(tài)路徑進(jìn)行決策,同時(shí)對三個(gè)目標(biāo)進(jìn)行優(yōu)化,從疏散人員整體考慮,為疏散人員個(gè)體提供適合的疏散路徑。本文通過對盛京銀行虛擬逃生指示系統(tǒng)的設(shè)計(jì)與實(shí)現(xiàn),將改進(jìn)的多目標(biāo)粒子群算法應(yīng)用于三維的逃生指示系統(tǒng)中的疏散路徑智能優(yōu)化模塊,同時(shí)優(yōu)化三個(gè)目標(biāo)對疏散人員的路徑進(jìn)行分析決策,將決策后的路徑以三維可視化的方式顯示在終端,為決策人員以及疏散人員在火災(zāi)發(fā)生時(shí)提供一個(gè)合理的疏散方案。
[Abstract]:The design of modern architecture is more complicated and larger with the development of science and technology, and the occurrence of building fire is often bursty. Therefore, how to reasonably and efficiently evacuate the personnel in the building fire has also become the focus of the research. Because the traditional evacuation plan is only the shortest path or the shortest time, the path of the scattered personnel is planned, and in the course of the evacuation of the actual building fire, the single goal cannot reflect the actual emergency evacuation because of its limitation. In order to ensure a safe and rapid evacuation, it is necessary to make a decision on a plurality of objects on the path selected by the evacuee. The corresponding multi-objective emergency evacuation model is constructed based on the data model of the building and the corresponding path parameters are added to construct the dynamic path model based on the data model of the building. Because the basic particle swarm optimization is not good for solving the actual multi-objective problem, the fitness function in the particle swarm optimization is modified, the position and velocity formula of the particles are modified, and the linear inertia weight is used to optimize the particle swarm optimization. The improved multi-target particle swarm algorithm is used to make a decision for the dynamic path of the fire bar of each of the evacuees, and the three objectives are optimized, and the suitable evacuation route is provided for the individual of the evacuees. In this paper, through the design and implementation of the virtual escape instruction system of Shengjing Bank, the improved multi-objective particle swarm optimization algorithm is applied to the intelligent optimization module of the evacuation route in the three-dimensional escape instruction system, and the path of the three targets is optimized for analysis and decision. The decision-making path is displayed on the terminal in a three-dimensional visual way, and a reasonable evacuation plan is provided for the decision-maker and the dispersion personnel in the event of a fire.
【學(xué)位授予單位】：東北大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2015
【分類號(hào)】：TU998.1;TP18;TP311.52

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 李海江;;特別重大火災(zāi)人員傷亡統(tǒng)計(jì)與分析[J];中國減災(zāi);2013年05期

2 要進(jìn)輝;趙慶海;李建華;;虛擬現(xiàn)實(shí)技術(shù)在消防特勤模擬訓(xùn)練中的應(yīng)用[J];安全;2012年03期

3 趙蔚;段紅;;虛擬現(xiàn)實(shí)軟件研究[J];計(jì)算機(jī)技術(shù)與發(fā)展;2012年02期

4 李遠(yuǎn)鑫;蔣海鷗;徐亦飛;徐芝琦;;基于Web3D的交互式虛擬社區(qū)[J];計(jì)算機(jī)工程;2011年11期

5 秦全德;李榮鈞;;基于生物寄生行為的雙種群粒子群算法[J];控制與決策;2011年04期

6 劉軍;方源敏;苗芳芳;;礦井火災(zāi)避災(zāi)決策支持系統(tǒng)的研究與設(shè)計(jì)[J];科學(xué)技術(shù)與工程;2011年02期

7 唐斌;;虛擬現(xiàn)實(shí)技術(shù)在消防戰(zhàn)訓(xùn)工作中的應(yīng)用[J];中國新技術(shù)新產(chǎn)品;2010年16期

8 ;An adaptive hybrid optimizer based on particle swarm and differential evolution for global optimization[J];Science China(Information Sciences);2010年05期

9 崔光照;李小廣;張勛才;王延峰;李翠玲;;基于改進(jìn)的粒子群遺傳算法的DNA編碼序列優(yōu)化[J];計(jì)算機(jī)學(xué)報(bào);2010年02期

10 公茂果;焦李成;楊咚咚;馬文萍;;進(jìn)化多目標(biāo)優(yōu)化算法研究[J];軟件學(xué)報(bào);2009年02期

，

本文編號(hào)：2483961