船舶真火模拟训练控制系统的研究

发布时间：2018-05-23 11:48

本文选题：船舶真火 + FDS　；参考：《江苏科技大学》2017年硕士论文

【摘要】：船舶火灾事故与损失呈逐年上升趋势,各种类型船舶火灾、爆炸事故频发,造成了大量的财产损失与人员伤亡。由于船舶消防事故是小概率事件,单纯依靠在实际的消防事故中取得的经验是有限的,传统消防训练机构的训练环境与船舶火灾训练环境差异很大。国内的消防训练系统多采用传统方式,使用真实的燃料如油、柴火等生成消防训练环境,具有污染大,可控性差等不利因素。国外的消防机构根据消防训练的特点,使用燃气作为燃料,配合训练道具和场景的布置来模拟火灾场景,但是并没有针对船舶火灾事故开展消防训练系统的研究。因此,有必要建立一套能逼真地模拟船舶真实火灾环境的消防训练系统。首先,本文以船舶机舱油池火为研究对象,采用火灾动力学分析软件FDS,建立了船舶机舱油池火物理模型,对船舶机舱火灾发展过程进行了研究。根据FDS仿真结果,分析了火源功率、风机流量、舱室体积和消防水对舱室火灾的影响,分别建立了相应的多项式数学模型,为设计消防训练系统、分析船舶火灾危害程度提供了理论基础。其次,针对船舶火灾安全评估,构建了多组火灾场景,以FDS仿真结果为训练样本,采用遗传算法优化支持向量机参数(GA-SVM)理论,建立了船舶火灾的安全评估模型。为了验证本文模型的准确性,分别探讨了支持向量机(SVM)模型和BP神经网络模型对船舶机舱火灾温度的预测效果,仿真结果表明,GA-SVM模型具有较好的泛化能力和预测效果,为船舶火灾温度间接计算及预测提出了一种新方法。然后,针对船舶火灾消防训练,采用多项式对以FDS的仿真结果进行数据拟合,建立了多项式燃烧数学模型和灭火数学模型,在理想条件下(一定的风机流量、忽略消防水损失等因素),实验结果表明这两种模型可以满足某些特定条件的消防训练需要,但存在一定的局限性。鉴于此,提出了基于火场氧气浓度、火盆温度、环境温度以及火源功率的灭火有效性标准,构建了多组火灾场景,验证了这种消防训练灭火有效性标准的准确性。最后以FDS仿真结果,采用GA-SVM模型,建立了不同维数的灭火模型,仿真结果证明,二维灭火模型不但可以缩短模型运行时间,而且可以满足消防训练的精度需求,为消防控制系统设计提供了依据。最后,根据训练系统要求,分别对训练控制系统和监控系统的总体架构进行了设计。采用STEP 7-Micro/WIN SMART编程软件对自检、点火、燃烧和灭火等控制程序进行了设计。采用力控ForceControl 7.0对训练监控、曲线分析、事件记录和系统设置等功能界面进行了设计。采用OPC技术实现力控与MATLAB之间的信息交互,解决了训练模型控制算法实现的问题。
[Abstract]:Ship fire accidents and losses are rising year by year. Various types of ship fire and explosion accidents occur frequently, resulting in a large number of property losses and casualties. Because the ship fire accident is a small probability event, it is limited to rely solely on the experience gained in the actual fire fighting accident. The training environment of the traditional fire fighting training institution is very different from the ship fire training environment. The fire fighting training system in our country mostly adopts the traditional way, uses the real fuel such as oil, firewood and so on to form the fire fighting training environment, has the pollution big, the controllability is poor and so on unfavorable factor. According to the characteristics of fire fighting training, foreign fire control agencies use gas as fuel to simulate fire scene with training props and scene layout, but there is no research on fire training system for ship fire accident. Therefore, it is necessary to set up a fire fighting training system which can simulate the real fire environment of ships. Firstly, taking the oil pool fire in the engine room of the ship as the research object, the physical model of the oil pool fire in the engine room of the ship is established by using the fire dynamic analysis software FDS, and the development process of the fire in the engine room of the ship is studied. According to the simulation results of FDS, the influences of fire source power, fan flow rate, cabin volume and fire water on the cabin fire are analyzed, and the corresponding polynomial mathematical models are established to design the fire fighting training system. It provides a theoretical basis for analyzing the damage degree of ship fire. Secondly, according to the ship fire safety assessment, a multi-group fire scene is constructed. Taking the FDS simulation results as the training sample, using the genetic algorithm to optimize the parameters of support vector machine (SVM) theory, the safety assessment model of ship fire is established. In order to verify the veracity of this model, the prediction effect of support vector machine (SVM) model and BP neural network model on ship engine room fire temperature is discussed respectively. The simulation results show that the GA-SVM model has better generalization ability and prediction effect. A new method for indirect calculation and prediction of ship fire temperature is proposed. Then, in view of the fire fighting training of ship, the polynomial is used to fit the data of the simulation result of FDS, and the mathematical model of polynomial combustion and fire extinguishing is established. Under the ideal conditions (a certain fan flow rate, The experimental results show that the two models can meet the fire training needs of some special conditions, but there are some limitations. In view of this, a fire extinguishing efficiency standard based on oxygen concentration, fire basin temperature, ambient temperature and fire source power is proposed, and several fire scenarios are constructed, and the accuracy of the fire fighting effectiveness standard for fire fighting training is verified. Finally, based on the simulation results of FDS and GA-SVM model, the fire extinguishing model with different dimensions is established. The simulation results show that the two-dimensional fire extinguishing model can not only shorten the running time of the model, but also meet the precision requirement of fire fighting training. It provides the basis for the design of fire control system. Finally, according to the requirements of the training system, the overall structure of the training control system and the monitoring system are designed. The control programs of self-inspection, ignition, combustion and fire extinguishing are designed by using STEP 7-Micro/WIN SMART programming software. The function interface of training monitoring, curve analysis, event recording and system setting is designed by force control ForceControl 7.0. The information exchange between force control and MATLAB is realized by OPC technology, and the problem of training model control algorithm is solved.
【学位授予单位】：江苏科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U698.4

【参考文献】