[go: up one dir, main page]

CN115526123A - Method and system for forecasting jet flow development of smoke ceiling in tunnel fire based on data assimilation - Google Patents

Method and system for forecasting jet flow development of smoke ceiling in tunnel fire based on data assimilation Download PDF

Info

Publication number
CN115526123A
CN115526123A CN202211025127.XA CN202211025127A CN115526123A CN 115526123 A CN115526123 A CN 115526123A CN 202211025127 A CN202211025127 A CN 202211025127A CN 115526123 A CN115526123 A CN 115526123A
Authority
CN
China
Prior art keywords
head
temperature
smoke
tunnel
fire
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202211025127.XA
Other languages
Chinese (zh)
Other versions
CN115526123B (en
Inventor
阳东
何苗
郭鑫
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Chongqing University
Original Assignee
Chongqing University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Chongqing University filed Critical Chongqing University
Priority to CN202211025127.XA priority Critical patent/CN115526123B/en
Publication of CN115526123A publication Critical patent/CN115526123A/en
Application granted granted Critical
Publication of CN115526123B publication Critical patent/CN115526123B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/28—Design optimisation, verification or simulation using fluid dynamics, e.g. using Navier-Stokes equations or computational fluid dynamics [CFD]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2113/00—Details relating to the application field
    • G06F2113/08—Fluids
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
    • G06F2119/02—Reliability analysis or reliability optimisation; Failure analysis, e.g. worst case scenario performance, failure mode and effects analysis [FMEA]

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Fluid Mechanics (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Computing Systems (AREA)
  • Pure & Applied Mathematics (AREA)
  • Computer Hardware Design (AREA)
  • Evolutionary Computation (AREA)
  • Geometry (AREA)
  • General Engineering & Computer Science (AREA)
  • Algebra (AREA)
  • Fire-Detection Mechanisms (AREA)

Abstract

The invention discloses a tunnel fire smoke ceiling jet flow development forecasting method and system based on data assimilation, wherein the forecasting method comprises the following steps: the method is used for forecasting the head temperature, the head propagation speed and the head thickness in the process that the head is propagated to each position behind the position of the current temperature sensor in super real time. The system comprises a temperature sensor, a data acquisition unit, a server, a control system and a fire fighting device arranged in the tunnel, wherein the temperature sensor, the data acquisition unit, the server and the control system are sequentially connected. The method is based on data assimilation, so that the model error is effectively reduced, and the forecasting result of the simple physical model is more reliable; based on a simple physical model, the time cost is effectively reduced, the super real-time prediction of the jet flow development of the tunnel fire smoke ceiling is realized, and the method has adaptability to different tunnel fire scenes.

Description

基于数据同化隧道火灾烟气顶棚射流发展预报方法及系统Method and system for forecasting jet flow development of smoke ceiling in tunnel fire based on data assimilation

技术领域technical field

本发明涉及一种基于数据同化的隧道火灾烟气顶棚射流发展预报方法及系统，可以实现顶棚射流发展过程烟气参数的超实时预报。The invention relates to a data assimilation-based method and system for forecasting the jet flow development of the roof of the tunnel fire smoke, which can realize the super real-time forecast of the smoke parameters during the development process of the roof jet flow.

背景技术Background technique

隧道具有狭长结构，与外界的连通口较少，其火灾风险与处置难度比一般地面建筑大。高温、有毒烟气是隧道火灾致死致伤的主要因素。隧道火灾早期，烟气以顶棚射流的形式发展，提前获得顶棚射流动态发展过程中的关键参数，可为火灾早期人员疏散路线的及时选择、排烟及救援方案的高效制定提供依据，对于隧道火灾的正确处置及减少人员伤亡来说意义重大。The tunnel has a long and narrow structure with fewer connections with the outside world, and its fire risk and disposal difficulty are greater than those of ordinary ground buildings. High temperature and toxic fumes are the main factors causing death and injury in tunnel fires. In the early stage of a tunnel fire, the smoke develops in the form of ceiling jets. Obtaining the key parameters in the dynamic development process of the ceiling jets in advance can provide a basis for the timely selection of evacuation routes for early fire personnel and the efficient formulation of smoke exhaust and rescue plans. It is of great significance for the correct disposal and the reduction of casualties.

布置在隧道中的传感器虽然可监测火灾烟气发展的实时信息，但是并不具备对监测时刻之后的烟气发展过程进行预报的能力。CFD场模拟虽然可获得事先设定的火灾场景的烟气动态过程，但由于描述隧道火灾的数学模型很复杂，模拟隧道烟气发展动态过程往往需要耗费大量计算机资源，导致执行CFD场模拟所需时间长于烟气实际发展时间，目前还不能达到预报的目的。有学者通过智能算法学习大量CFD场模拟案例的数据结果，从而建立基于智能算法的快速预报模型。但事先通过CFD模拟结果构建的数据库庞大，且难以覆盖所有火灾场景，这导致这种方法不能灵活地应用于不同的隧道火灾场景中。Although the sensors arranged in the tunnel can monitor the real-time information of the fire smoke development, they do not have the ability to predict the smoke development process after the monitoring time. Although the CFD field simulation can obtain the smoke dynamic process of the pre-set fire scene, due to the complexity of the mathematical model describing the tunnel fire, the simulation of the dynamic process of the tunnel smoke development often requires a lot of computer resources, resulting in the need for CFD field simulation. The time is longer than the actual development time of the flue gas, and the purpose of forecasting cannot be achieved at present. Some scholars learn the data results of a large number of CFD field simulation cases through intelligent algorithms, so as to establish a rapid prediction model based on intelligent algorithms. However, the database constructed by CFD simulation results in advance is huge, and it is difficult to cover all fire scenarios, which makes this method unable to be flexibly applied to different tunnel fire scenarios.

针对隧道火灾烟气顶棚射流，已经存在形式简单、计算效率很高的简单物理数学模型，能在很短的计算时间内获得烟气早期发展的动态信息，在效率上具备预报的潜力。但这类模型的预报精度却依赖于模型输入参数的准确性，在真实火灾场景中，这些参数难以被及时、准确地获取。输入参数的误差会转移到模型计算结果中，从而降低模型预报精度。另外，这些输入参数在顶棚射流发展过程中本身也是动态变化的，其影响因素多样且相互耦合。因此，单独使用简单模型很难保证预报精度，限制了其在实际隧道火灾中的应用。A simple physical-mathematical model with simple form and high calculation efficiency already exists for the ceiling jet flow of smoke in tunnel fires. It can obtain the dynamic information of the early development of smoke in a short calculation time, and has the potential for prediction in terms of efficiency. However, the prediction accuracy of this type of model depends on the accuracy of the model input parameters. In real fire scenarios, these parameters are difficult to obtain timely and accurately. The error of the input parameters will be transferred to the model calculation results, thereby reducing the accuracy of the model forecast. In addition, these input parameters are also dynamically changing during the development of the ceiling jet, and the influencing factors are diverse and coupled with each other. Therefore, it is difficult to guarantee the prediction accuracy by using the simple model alone, which limits its application in actual tunnel fires.

发明内容Contents of the invention

本发明的第一个目的在于提供一种基于数据同化的隧道火灾烟气顶棚射流发展预报方法，能够实现对隧道火灾早期顶棚射流阶段烟气发展的可靠预报。The first object of the present invention is to provide a data assimilation-based prediction method for the development of jet flow on the ceiling of tunnel fire smoke, which can realize reliable prediction of the development of smoke on the ceiling jet flow stage in the early stage of tunnel fire.

本发明的第一个目的通过以下技术措施实现：一种基于数据同化的隧道火灾烟气顶棚射流发展预报方法，其特征在于具体包括以下步骤：The first object of the present invention is achieved through the following technical measures: a method for forecasting the development of jet flow on the ceiling of tunnel fire smoke based on data assimilation, which is characterized in that it specifically includes the following steps:

S1、沿隧道纵向中轴线布设若干个用于记录烟气温度数据的温度传感器，且温度传感器沿隧道整个长度方向布满；S1. Arrange several temperature sensors for recording flue gas temperature data along the longitudinal central axis of the tunnel, and the temperature sensors are distributed along the entire length of the tunnel;

S2、通过温度传感器响应时间对火源位置进行判断，即认为火源处于最早发生温度异常的两个温度传感器的正中间位置；S2. Judging the position of the fire source based on the response time of the temperature sensor, that is, the fire source is considered to be in the middle of the two temperature sensors where the temperature abnormality occurred first;

S3、利用蒙特卡洛方法生成火源的热释放速率，并将热释放速率代入火羽流模型，计算得到烟气初始头部质量流量和烟气初始头部温度，再利用同化算法集合卡尔曼滤波同化最早发生温度异常的两个温度传感器所记录的烟气温度数据的平均值，对热释放速率、烟气初始头部质量流量和烟气初始头部温度进行校正；S3. Use the Monte Carlo method to generate the heat release rate of the fire source, and substitute the heat release rate into the fire plume model to calculate the initial mass flow rate of the smoke head and the initial head temperature of the smoke gas, and then use the assimilation algorithm to gather Kalman Filter and assimilate the average value of the flue gas temperature data recorded by the two temperature sensors with the earliest abnormal temperature, and correct the heat release rate, flue gas initial head mass flow rate and flue gas initial head temperature;

S4、利用蒙特卡洛方法生成卷吸比率和换热修正系数，和校正后的烟气初始头部质量流量和烟气初始头部温度一起作为输入参数代入顶棚射流模型，实现对头部到达不同位置时的头部温度、头部传播速度和头部厚度的超实时预报。在头部蔓延过程中，利用同化算法集合卡尔曼滤波同化头部所达位置处温度传感器记录的烟气温度数据，实时校正卷吸比率、换热修正系数、头部温度和头部质量流量，再将这些校正参数作为输入参数代入顶棚射流模型中，重新对头部蔓延到当前温度传感器所在位置之后的各位置时的头部温度、头部传播速度和头部厚度进行超实时预报。S4. Use the Monte Carlo method to generate the entrainment ratio and heat transfer correction coefficient, and substitute the corrected flue gas initial head mass flow rate and flue gas initial head temperature into the ceiling jet model as input parameters to achieve different head arrivals. Super real-time forecasts of head temperature, head propagation velocity and head thickness at time of location. During the head spreading process, the assimilation algorithm is used to integrate the Kalman filter to assimilate the flue gas temperature data recorded by the temperature sensor at the position where the head reaches, and the entrainment ratio, heat transfer correction coefficient, head temperature and head mass flow rate are corrected in real time. These correction parameters are then substituted into the ceiling jet model as input parameters, and the head temperature, head propagation velocity, and head thickness are re-predicted in real time when the head spreads to each position after the current temperature sensor.

本发明通过基于集合卡尔曼滤波的数据同化算法将温度传感器数据与隧道火灾烟气顶棚射流阶段简单模型融合。简单模型的应用相比于CFD场模拟来说，保证了预报的时间提前量。本发明方法所需执行时间远小于CFD应用于长大隧道时的执行时间，同时远小于隧道火灾早期烟气实际蔓延所需时间，因此具备对烟气发展态势进行预报的能力；本预报方法相比于基于智能算法的快速预报方法，不依赖于提前构建的数据库，能够适用于更多的火灾场景；通过同化温度传感器数据对模型关键输入参数进行实时校正，利用少量传感器测量数据补充模型所需关键信息，使得模型输入参数更准确，对火场的变化响应更及时，大幅提高了简单模型的预报精度，也拓展了隧道内传感器的用途。因此，本发明能够实现对隧道火灾早期顶棚射流阶段烟气发展的可靠预报。The invention fuses the temperature sensor data with the simple model of the tunnel fire smoke ceiling jet flow stage through the data assimilation algorithm based on the ensemble Kalman filter. Compared with the CFD field simulation, the application of the simple model guarantees the time advance of the forecast. The execution time required by the method of the present invention is much shorter than the execution time when CFD is applied to a long tunnel, and at the same time it is far shorter than the time required for the actual spread of smoke in the early stage of tunnel fire, so it has the ability to predict the development trend of smoke; the prediction method is relatively Compared with the rapid prediction method based on intelligent algorithms, it does not rely on the database built in advance, and can be applied to more fire scenarios; the key input parameters of the model are corrected in real time by assimilating temperature sensor data, and a small amount of sensor measurement data is used to supplement the model. The key information makes the input parameters of the model more accurate, responds more timely to changes in the fire scene, greatly improves the prediction accuracy of the simple model, and expands the use of sensors in the tunnel. Therefore, the present invention can realize the reliable prediction of the smoke development in the ceiling jet flow stage in the early stage of the tunnel fire.

本发明所述步骤S3具体包括：Step S3 of the present invention specifically includes:

(一)利用蒙特卡洛方法生成火源热释放速率的初始背景场：(1) Using the Monte Carlo method to generate the initial background field of the fire source heat release rate:

Figure BDA0003815396540000031
Figure BDA0003815396540000031

式中，

Figure BDA0003815396540000032
代表热释放速率，
Figure BDA0003815396540000033
代表参数集合应满足的正态分布，
Figure BDA0003815396540000034
分别代表所给参数集合的期望与标准方差，下标j代表第j个集合成员，上标f代表预测值。In the formula,
Figure BDA0003815396540000032
represents the heat release rate,
Figure BDA0003815396540000033
Represents the normal distribution that the set of parameters should satisfy,
Figure BDA0003815396540000034
Represent the expectation and standard deviation of the given parameter set respectively, the subscript j represents the jth set member, and the superscript f represents the predicted value.

(二)将由公式(1)生成的热释放速率集合成员分别代入火羽流模型的羽流流量计算公式得到火源处的羽流流量

Figure BDA0003815396540000035
考虑火源上方的烟气在撞击顶棚后会均匀地流向隧道两侧，继而得到烟气初始头部质量流量：(2) Substitute the members of the heat release rate set generated by formula (1) into the plume flow calculation formula of the fire plume model to obtain the plume flow at the fire source
Figure BDA0003815396540000035
Considering that the smoke above the fire source will flow evenly to both sides of the tunnel after hitting the ceiling, then the initial head mass flow rate of the smoke can be obtained:

Figure BDA0003815396540000036
Figure BDA0003815396540000036

Figure BDA0003815396540000037
Figure BDA0003815396540000037

Figure BDA0003815396540000038
Figure BDA0003815396540000038

Figure BDA0003815396540000039
Figure BDA0003815396540000039

Figure BDA00038153965400000310
Figure BDA00038153965400000310

其中，

Figure BDA00038153965400000311
为火源处的羽流流量，
Figure BDA00038153965400000312
为总热释放速率，
Figure BDA00038153965400000313
为对流热释放速率，一般取值范围为
Figure BDA00038153965400000314
到
Figure BDA00038153965400000315
ht为隧道高度，z0为虚点火源位置，
Figure BDA00038153965400000316
为火焰平均高度，D′为火源的当量直径，
Figure BDA00038153965400000318
为顶棚射流质量流量，上标f代表预测值，下标j代表第j个集合成员；in,
Figure BDA00038153965400000311
is the plume flow at the fire source,
Figure BDA00038153965400000312
is the total heat release rate,
Figure BDA00038153965400000313
is the convective heat release rate, and the general value range is
Figure BDA00038153965400000314
arrive
Figure BDA00038153965400000315
h t is the height of the tunnel, z 0 is the position of the virtual ignition source,
Figure BDA00038153965400000316
is the average flame height, D' is the equivalent diameter of the fire source,
Figure BDA00038153965400000318
is the ceiling jet mass flow rate, the superscript f represents the predicted value, and the subscript j represents the jth set member;

(三)将热释放速率和羽流流量代入火羽流模型的能量守恒公式，得到火源附近烟气温度，即为烟气初始头部温度：(3) Substituting the heat release rate and plume flow rate into the energy conservation formula of the fire plume model, the temperature of the flue gas near the fire source is obtained, which is the initial head temperature of the flue gas:

Figure BDA00038153965400000317
Figure BDA00038153965400000317

式中，Ts是火源上方平均烟气温度，

Figure BDA0003815396540000041
为对流热释放速率，Cp为定压比热容(1004J/(kg·K))，
Figure BDA0003815396540000042
为火源处的羽流流量，T0是环境温度，上标f代表预测值，下标j代表第j个集合成员，下标i代表头部到达第i个位置；In the formula, T s is the average flue gas temperature above the fire source,
Figure BDA0003815396540000041
is the convective heat release rate, C p is the specific heat capacity at constant pressure (1004J/(kg K)),
Figure BDA0003815396540000042
is the plume flow rate at the fire source, T 0 is the ambient temperature, the superscript f represents the predicted value, the subscript j represents the j-th set member, and the subscript i represents the head reaches the i-th position;

(四)通过同化算法集合卡尔曼滤波同化最早发生温度异常两温度传感器所记录的烟气温度数据的平均值，得到热释放速率、初始头部质量流量以及初始头部温度的校正值：(4) Assimilate the average value of the flue gas temperature data recorded by the two temperature sensors that first occurred temperature anomalies by assembling the Kalman filter through the assimilation algorithm, and obtain the correction value of the heat release rate, the initial head mass flow rate and the initial head temperature:

Xa＝Xf+Kx，y(y-HXf)X a =X f +K x,y (y-HX f )

Figure BDA0003815396540000043
Figure BDA0003815396540000043

Figure BDA0003815396540000044
Figure BDA0003815396540000044

Xf＝[x1 f，x2 f，K，xj f，K，xm f]X f = [x 1 f , x 2 f , K, x j f , K, x m f ]

Figure BDA0003815396540000045
Figure BDA0003815396540000045

Figure BDA0003815396540000046
Figure BDA0003815396540000046

Figure BDA0003815396540000047
Figure BDA0003815396540000047

其中，X为m个向量组成的集合矩阵，H为观测算子，包含了观测值与预测值向量的相对位置关系，R为观测值误差，P为预测协方差矩阵，Kx，y为卡尔曼增益矩阵，是由m个预报值向量组成的预报矩阵，x为预报值向量，y为传感器温度，Ts是火源上方平均烟气温度，

Figure BDA0003815396540000048
为顶棚射流质量流量，
Figure BDA0003815396540000049
为总热释放速率，上标a代表同化值，上标f代表预测值，下标j代表第j个集合成员。Among them, X is a set matrix composed of m vectors, H is an observation operator, which includes the relative positional relationship between the observed value and the predicted value vector, R is the error of the observed value, P is the prediction covariance matrix, and K x, y is Karl Mann gain matrix is a forecast matrix composed of m forecast value vectors, x is the forecast value vector, y is the sensor temperature, T s is the average smoke temperature above the fire source,
Figure BDA0003815396540000048
is the ceiling jet mass flow rate,
Figure BDA0003815396540000049
is the total heat release rate, the superscript a represents the assimilated value, the superscript f represents the predicted value, and the subscript j represents the jth ensemble member.

本发明所述步骤S4包括：Step S4 of the present invention comprises:

(一)利用蒙特卡洛方法生成卷吸比率和换热修正系数的初始背景场，(1) Using the Monte Carlo method to generate the initial background field of the entrainment ratio and heat transfer correction coefficient,

Figure BDA00038153965400000410
Figure BDA00038153965400000410

Figure BDA0003815396540000051
Figure BDA0003815396540000051

式中，λα，λce分别代表换热修正系数与卷吸比率，

Figure BDA0003815396540000052
Figure BDA0003815396540000053
分别代表参数集合应满足的正态分布，
Figure BDA0003815396540000054
分别代表λα集合的期望与标准方差，
Figure BDA0003815396540000055
分别代表λce集合的期望与标准方差，下标j代表第j个集合成员；where λ α and λ ce represent the heat transfer correction coefficient and the entrainment ratio, respectively,
Figure BDA0003815396540000052
Figure BDA0003815396540000053
Respectively represent the normal distribution that the parameter set should satisfy,
Figure BDA0003815396540000054
Represent the expectation and standard deviation of the λ α set, respectively,
Figure BDA0003815396540000055
Represent the expectation and standard deviation of the λ ce set, respectively, and the subscript j represents the jth set member;

(二)将卷吸比率和换热修正系数以及校正后的初始头部温度和初始头部质量流量代入顶棚射流模型的顶棚射流速度公式，得到顶棚射流头部传播速度、头部在相邻位置间的传播时间和头部到达不同位置的时间：(2) Substitute the entrainment ratio, heat transfer correction coefficient, corrected initial head temperature and initial head mass flow rate into the ceiling jet velocity formula of the ceiling jet model, and obtain the ceiling jet head propagation velocity and the head at adjacent positions The propagation time between and the time for the head to reach different positions:

Figure BDA0003815396540000056
Figure BDA0003815396540000056

Figure BDA0003815396540000057
Figure BDA0003815396540000057

Δti→i+1＝(Li+1-Li)/Us，i Δ ti→i+1 = (L i+1 -L i )/U s,i

ti+l＝ti+Δt i→i+1(i＝0，1，...n，-1) 公式(6)t i+l =t i + Δt i→i+1 (i=0, 1,...n, -1) formula (6)

式中，

Figure BDA0003815396540000058
Ts，i和Us，i分别代表了顶棚射流头部热流量、头部烟气温度以及头部纵向蔓延速度W，T0和ρ0分别为隧道宽度、环境空气温度以及环境空气密度；Li为i位置距离火源中心的距离，n为待预报位置的总数。In the formula,
Figure BDA0003815396540000058
T s, i and U s, i respectively represent the heat flow at the head of the ceiling jet, the temperature of the smoke at the head and the longitudinal spreading speed W at the head, and T 0 and ρ 0 are the tunnel width, ambient air temperature and ambient air density, respectively; L i is the distance from position i to the center of the fire source, and n is the total number of positions to be predicted.

(三)将卷吸比率和换热修正系数以及校正后的初始顶棚射流头部温度和初始头部质量流量代入顶棚射流模型的质量守恒公式，得到头部厚度：(3) Substituting the entrainment ratio and heat transfer correction coefficient, the corrected initial ceiling jet head temperature and initial head mass flow rate into the mass conservation formula of the ceiling jet model, the head thickness is obtained:

Figure BDA0003815396540000059
Figure BDA0003815396540000059

Figure BDA00038153965400000510
Figure BDA00038153965400000510

(四)计算头部从位置i到位置i+1蔓延过程中的换热损失，在计算过程中引入换热修正系数：(4) Calculate the heat transfer loss during the spreading process of the head from position i to position i+1, and introduce the heat transfer correction coefficient in the calculation process:

Figure BDA0003815396540000061
Figure BDA0003815396540000061

Figure BDA0003815396540000062
Figure BDA0003815396540000062

Figure BDA0003815396540000063
Figure BDA0003815396540000063

式中，αc为烟气与隧道衬砌结构间的换热系数；k0为空气导热系数；v0为空气的运动粘度系数，α0为空气的热扩散系数；h′t为隧道断面的水力学直径；

Figure BDA0003815396540000064
为顶棚射流头部从位置i到i+1的换热量；Tw为隧道壁面温度。In the formula, α c is the heat transfer coefficient between the flue gas and the tunnel lining structure; k 0 is the thermal conductivity of the air; v 0 is the kinematic viscosity coefficient of the air, α 0 is the thermal diffusivity of the air; h′ t is the thermal conductivity of the tunnel section hydraulic diameter;
Figure BDA0003815396540000064
is the heat transfer rate of the ceiling jet head from position i to i+1; T w is the temperature of the tunnel wall.

(五)将位置i处头部温度、头部质量流量以及换热损失代入顶棚射流模型的能量守恒公式计算位置i+1处的头部温度，在计算过程中引入卷吸比率：(5) Substitute the head temperature, head mass flow rate and heat transfer loss at position i into the energy conservation formula of the ceiling jet model to calculate the head temperature at position i+1, and introduce the entrainment ratio in the calculation process:

Figure BDA0003815396540000065
Figure BDA0003815396540000065

Figure BDA0003815396540000066
Figure BDA0003815396540000066

(六)通过同化算法集合卡尔曼滤波同化头部到达位置处温度传感器所记录的烟气温度数据，实时校正头部温度、头部质量流量、卷吸比率和换热修正系数；(6) Assimilating the flue gas temperature data recorded by the temperature sensor at the arrival position of the head through the Kalman filter assimilation algorithm, and correcting the head temperature, head mass flow rate, entrainment ratio and heat transfer correction coefficient in real time;

(七)将校正后的头部温度、头部质量流量、卷吸比率以及换热修正系数作为输入参数代入顶棚射流模型中，更新对后续顶棚射流头部温度、头部传播速度和头部厚度的超实时预报。(7) Substitute the corrected head temperature, head mass flow rate, entrainment ratio, and heat transfer correction coefficient into the ceiling jet model as input parameters, and update the head temperature, head propagation velocity, and head thickness of subsequent ceiling jets. super real-time forecast.

本发明的第二个目的在于提供一种使用上述基于数据同化的隧道火灾烟气顶棚射流发展预报方法的系统。The second object of the present invention is to provide a system using the above-mentioned data assimilation-based method for forecasting the jet flow development of the roof of tunnel fire smoke.

本发明的第二个目的通过以下的技术措施来实现：一种使用上述基于数据同化的隧道火灾烟气顶棚射流发展预报方法的系统，其特征在于，它包括依次连接的温度传感器、数据采集器、服务器、控制系统和设置在隧道内的消防装置，所述温度传感器在隧道纵向中轴线上布设若干个且沿隧道整个长度方向以一定间距布设，所述温度传感器将采集的烟气温度数据传送给数据采集器，由其处理后发送给服务器，所述服务器对烟气温度数据进行同化，计算并输出超实时预报结果，所述控制系统根据超实时预报结果提供消防策略参考并可以控制消防装置。The second object of the present invention is achieved by the following technical measures: a system using the above-mentioned method for forecasting the development of the tunnel fire smoke ceiling jet flow based on data assimilation is characterized in that it includes temperature sensors and data collectors connected in sequence , a server, a control system and a fire-fighting device arranged in the tunnel, the temperature sensors are arranged on the longitudinal central axis of the tunnel and arranged at a certain interval along the entire length of the tunnel, and the temperature sensors transmit the collected flue gas temperature data To the data collector, after processing, send it to the server, the server assimilates the flue gas temperature data, calculates and outputs the super real-time forecast result, and the control system provides fire fighting strategy reference and can control the fire fighting device according to the super real-time forecast result .

本发明各温度传感器等间距设置，且设置间距不超过20m。The temperature sensors of the present invention are arranged at equal intervals, and the interval between them is not more than 20m.

本发明各温度传感器距离顶棚不超过20cm。Each temperature sensor of the present invention is no more than 20cm away from the ceiling.

与现有技术相比，本发明具有如下显著的效果：Compared with prior art, the present invention has following remarkable effect:

本发明通过数据同化算法集合卡尔曼滤波将温度传感器的烟气温度数据与隧道火灾烟气顶棚射流阶段简单模型融合。简单物理模型的应用相比于CFD技术保证了预报的时间提前量，本发明所需执行时间远小于CFD在长大隧道模拟中的执行时间，也远小于隧道火灾早期烟气实际蔓延时间，因此能够实现对烟气发展态势进行预报的目的；本发明基于简单物理模型相比于基于智能算法的快速预报方法能够适用于更多的火灾场景；通过同化可靠的温度传感器数据对模型关键输入参数进行实时估计，实现了传感器数据对于火场信息的补充，将关键输入参数的实时估计值带入模型后能够减小由于输入参数不确定带来的模型误差，在减少传感器类型与布置密度的同时，大幅提高了简单模型在预报过程中的精度，也拓展了隧道洞内环境参数实时监测数据的用途。另外，对于物理模型简化带来模型固有的计算误差，本发明在模型表达复杂物理过程的公式中引入修正系数作为模型输入参数，将固有的模型误差转化为了输入参数的误差，并通过上述数据同化的方式，利用可靠的传感器数据减少了这一误差对于预报结果的影响。因此，本发明能够实现对隧道火灾早期顶棚射流阶段烟气发展的可靠预报。The invention fuses the smoke temperature data of the temperature sensor with the simple model of the jet flow stage of the smoke ceiling of the tunnel fire through the data assimilation algorithm set Kalman filter. Compared with CFD technology, the application of simple physical model guarantees the time advance of prediction. The execution time required by the present invention is much shorter than the execution time of CFD in long tunnel simulation, and it is also far shorter than the actual smoke spread time in the early stage of tunnel fire. Therefore, The purpose of predicting the development trend of smoke can be realized; the present invention is based on a simple physical model and can be applied to more fire scenarios than the rapid prediction method based on an intelligent algorithm; the key input parameters of the model are assimilated by assimilating reliable temperature sensor data Real-time estimation realizes the supplement of sensor data to fire scene information. After bringing the real-time estimated values of key input parameters into the model, it can reduce the model error caused by uncertain input parameters. While reducing the sensor type and layout density, it can significantly It improves the accuracy of the simple model in the forecasting process, and also expands the use of real-time monitoring data of environmental parameters in the tunnel. In addition, for the calculation error inherent in the model caused by the simplification of the physical model, the present invention introduces the correction coefficient into the formula for expressing the complex physical process as the model input parameter, converts the inherent model error into the error of the input parameter, and through the above data assimilation In this way, reliable sensor data is used to reduce the influence of this error on the forecast results. Therefore, the present invention can realize the reliable prediction of the smoke development in the ceiling jet flow stage in the early stage of the tunnel fire.

附图说明Description of drawings

下面结合附图和具体实施例对本发明作进一步的详细说明。The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.

图1是本发明方法的流程框图；Fig. 1 is a block flow diagram of the inventive method;

图2是本发明方法具体实施例所用隧道模型及火源位置示意图；Fig. 2 is the tunnel model used in the specific embodiment of the method of the present invention and the position schematic diagram of fire source;

图3是烟气动态发展模拟温度分布图；Figure 3 is a simulated temperature distribution diagram of flue gas dynamic development;

图4是不同化次数下预报效果评估图；Figure 4 is the evaluation chart of the forecast effect under different times;

图5是同化三次后对顶棚射流蔓延过程中头部温度参数变化的预报图；Fig. 5 is a forecast diagram of the head temperature parameter change during the roof jet spreading process after assimilating three times;

图6是同化三次后对顶棚射流蔓延过程中头部界面高度变化的预报图；Fig. 6 is a forecast map of the head interface height change during the roof jet spreading process after assimilating three times;

图7是本发明系统的结构组成示意图。Fig. 7 is a schematic diagram of the structural composition of the system of the present invention.

具体实施方式detailed description

下面结合实施例及其附图对本发明进行详细说明，以帮助本领域的技术人员更好的理解本发明的发明构思，但本发明权利要求的保护范围不限于下述实施例，对于本领域的技术人员来说，在不脱离本发明之发明构思的前提下，没有做出创造性劳动所获得的所有其他实施例，都属于本发明的保护范围。The present invention is described in detail below in conjunction with embodiment and accompanying drawing thereof, to help those skilled in the art better understand the inventive concept of the present invention, but the scope of protection of the claims of the present invention is not limited to following embodiment, for those skilled in the art For those skilled in the art, on the premise of not departing from the inventive concept of the present invention, all other embodiments obtained without creative work all belong to the protection scope of the present invention.

如图1～6所示，是本发明一种基于数据同化的隧道火灾烟气顶棚射流发展预报方法，具体包括以下步骤：As shown in Figures 1 to 6, it is a data assimilation-based method for predicting the development of the jet flow development of the ceiling of the tunnel fire smoke, which specifically includes the following steps:

S1、沿隧道纵向中轴线布设若干个用于记录烟气温度数据的温度传感器，且温度传感器沿隧道整个长度方向布满；S1. Arrange several temperature sensors for recording flue gas temperature data along the longitudinal central axis of the tunnel, and the temperature sensors are distributed along the entire length of the tunnel;

S2、通过温度传感器响应时间对火源位置进行判断，即认为火源处于最早发生温度异常的两个温度传感器的正中间位置；S2. Judging the position of the fire source based on the response time of the temperature sensor, that is, the fire source is considered to be in the middle of the two temperature sensors where the temperature abnormality occurred first;

S3、利用蒙特卡洛方法生成火源的热释放速率，并将热释放速率代入火羽流模型，计算得到烟气初始头部质量流量和烟气初始头部温度，再利用同化算法集合卡尔曼滤波同化最早发生温度异常的两个温度传感器所记录的烟气温度数据的平均值，对热释放速率、烟气初始头部质量流量和烟气初始头部温度进行校正；S3. Use the Monte Carlo method to generate the heat release rate of the fire source, and substitute the heat release rate into the fire plume model to calculate the initial mass flow rate of the smoke head and the initial head temperature of the smoke gas, and then use the assimilation algorithm to gather Kalman Filter and assimilate the average value of the flue gas temperature data recorded by the two temperature sensors with the earliest abnormal temperature, and correct the heat release rate, flue gas initial head mass flow rate and flue gas initial head temperature;

具体包括：Specifically include:

(一)利用蒙特卡洛方法生成火源的热释放速率的初始背景场：(1) Using the Monte Carlo method to generate the initial background field of the heat release rate of the fire source:

Figure BDA0003815396540000081
Figure BDA0003815396540000081

式中，

Figure BDA0003815396540000082
代表热释放速率，
Figure BDA0003815396540000083
代表参数集合应满足的正态分布，
Figure BDA0003815396540000084
分别代表所给参数集合的期望与标准方差，下标j代表第j个集合成员，上标f代表预测值。In the formula,
Figure BDA0003815396540000082
represents the heat release rate,
Figure BDA0003815396540000083
Represents the normal distribution that the set of parameters should satisfy,
Figure BDA0003815396540000084
Represent the expectation and standard deviation of the given parameter set respectively, the subscript j represents the jth set member, and the superscript f represents the predicted value.

(二)将由公式(1)生成的热释放速率集合成员分别代入火羽流模型的羽流流量计算公式得到火源处的羽流流量

Figure BDA0003815396540000085
由于火源上方的烟气在撞击顶棚后会均匀地流向隧道两侧，继而得到烟气初始头部质量流量：(2) Substitute the members of the heat release rate set generated by formula (1) into the plume flow calculation formula of the fire plume model to obtain the plume flow at the fire source
Figure BDA0003815396540000085
Since the smoke above the fire source will flow evenly to both sides of the tunnel after hitting the ceiling, then the initial head mass flow rate of the smoke is obtained:

Figure BDA0003815396540000091
Figure BDA0003815396540000091

Figure BDA0003815396540000092
Figure BDA0003815396540000092

Figure BDA0003815396540000093
Figure BDA0003815396540000093

Figure BDA0003815396540000094
Figure BDA0003815396540000094

Figure BDA0003815396540000095
Figure BDA0003815396540000095

其中，

Figure BDA0003815396540000096
为火源处的羽流流量，
Figure BDA0003815396540000097
为总热释放速率，
Figure BDA0003815396540000098
为对流热释放速率，一般取值范围为
Figure BDA0003815396540000099
到
Figure BDA00038153965400000910
ht为隧道高度，z0为虚点火源位置，
Figure BDA00038153965400000911
为火焰平均高度，D′为火源的当量直径，
Figure BDA00038153965400000916
为顶棚射流质量流量，上标f代表预测值，下标j代表第j个集合成员。in,
Figure BDA0003815396540000096
is the plume flow at the fire source,
Figure BDA0003815396540000097
is the total heat release rate,
Figure BDA0003815396540000098
is the convective heat release rate, and the general value range is
Figure BDA0003815396540000099
arrive
Figure BDA00038153965400000910
h t is the height of the tunnel, z 0 is the position of the virtual ignition source,
Figure BDA00038153965400000911
is the average flame height, D' is the equivalent diameter of the fire source,
Figure BDA00038153965400000916
is the ceiling jet mass flow rate, the superscript f represents the predicted value, and the subscript j represents the jth set member.

(三)将热释放速率和羽流流量代入火羽流模型的能量守恒公式，得到火源附近烟气温度，即为烟气初始头部温度：(3) Substituting the heat release rate and plume flow rate into the energy conservation formula of the fire plume model, the temperature of the flue gas near the fire source is obtained, which is the initial head temperature of the flue gas:

Figure BDA00038153965400000912
Figure BDA00038153965400000912

式中，Ts是火源上方平均烟气温度，

Figure BDA00038153965400000913
为对流热释放速率，Cp为定压比热容(1004J/(kg·K))，
Figure BDA00038153965400000914
为火源处的羽流流量，T0是环境温度，上标f代表预测值，下标j代表第j个集合成员，下标i代表头部到达第i个位置；In the formula, T s is the average flue gas temperature above the fire source,
Figure BDA00038153965400000913
is the convective heat release rate, C p is the specific heat capacity at constant pressure (1004J/(kg K)),
Figure BDA00038153965400000914
is the plume flow rate at the fire source, T 0 is the ambient temperature, the superscript f represents the predicted value, the subscript j represents the j-th set member, and the subscript i represents the head reaches the i-th position;

(四)通过同化算法集合卡尔曼滤波同化最早发生温度异常两温度传感器所记录的烟气温度数据的平均值，得到热释放速率、烟气初始头部质量流量以及烟气初始头部温度的校正值：(4) Assimilate the average value of the flue gas temperature data recorded by the two temperature sensors that first occurred temperature anomalies by assembling the Kalman filter through the assimilation algorithm, and obtain the correction of the heat release rate, the mass flow rate of the initial head of the flue gas, and the initial head temperature of the flue gas value:

Xa＝Xf+Kx，y(y-HXf)X a =X f +K x,y (y-HX f )

Figure BDA00038153965400000915
Figure BDA00038153965400000915

Figure BDA0003815396540000101
Figure BDA0003815396540000101

Xf＝[x1 f，x2 f，K，xj f，K，xm f]X f = [x 1 f , x 2 f , K, x j f , K, x m f ]

Figure BDA0003815396540000102
Figure BDA0003815396540000102

Figure BDA0003815396540000103
Figure BDA0003815396540000103

Figure BDA0003815396540000104
Figure BDA0003815396540000104

其中，X为m个向量组成的集合矩阵，H为观测算子，包含了观测值与预测值向量的相对位置关系，R为观测值误差，P为预测协方差矩阵，Kx，y为卡尔曼增益矩阵，是由m个预报值向量组成的预报矩阵，x为预报值向量，y为传感器温度，Ts是火源上方平均烟气温度，

Figure BDA0003815396540000105
为顶棚射流质量流量，
Figure BDA0003815396540000106
为总热释放速率，上标a代表同化值，上标f代表预测值，下标j代表第j个集合成员。Among them, X is a set matrix composed of m vectors, H is an observation operator, which includes the relative positional relationship between the observed value and the predicted value vector, R is the error of the observed value, P is the prediction covariance matrix, and K x, y is Karl Mann gain matrix is a forecast matrix composed of m forecast value vectors, x is the forecast value vector, y is the sensor temperature, T s is the average smoke temperature above the fire source,
Figure BDA0003815396540000105
is the ceiling jet mass flow rate,
Figure BDA0003815396540000106
is the total heat release rate, the superscript a represents the assimilated value, the superscript f represents the predicted value, and the subscript j represents the jth ensemble member.

S4、利用蒙特卡洛方法生成卷吸比率和换热修正系数，和校正后的烟气初始头部质量流量和烟气初始头部温度一起作为输入参数代入顶棚射流模型，实现对头部到达不同位置时的头部温度、头部传播速度和头部厚度的超实时预报，而在头部蔓延过程中，利用同化算法集合卡尔曼滤波同化头部所达位置处温度传感器记录的烟气温度数据，实时校正卷吸比率、换热修正系数、头部温度和头部质量流量，再将这些校正参数作为输入参数代入顶棚射流模型中，重新对头部蔓延到当前温度传感器所在位置之后的各位置时的头部温度、头部传播速度和头部厚度进行超实时预报。S4. Use the Monte Carlo method to generate the entrainment ratio and heat transfer correction coefficient, and substitute the corrected flue gas initial head mass flow rate and flue gas initial head temperature into the ceiling jet model as input parameters to achieve different head arrivals. Super real-time forecast of head temperature, head propagation velocity and head thickness at the position, and in the process of head spread, use the assimilation algorithm ensemble Kalman filter to assimilate the flue gas temperature data recorded by the temperature sensor at the position where the head reaches , real-time correction of entrainment ratio, heat transfer correction coefficient, head temperature and head mass flow rate, and then these correction parameters are substituted into the ceiling jet model as input parameters, and the head spreads to each position after the current temperature sensor position Super real-time prediction of head temperature, head propagation velocity and head thickness.

具体包括：Specifically include:

(一)利用蒙特卡洛方法生成卷吸比率和换热修正系数的初始背景场，(1) Using the Monte Carlo method to generate the initial background field of the entrainment ratio and heat transfer correction coefficient,

Figure BDA0003815396540000107
Figure BDA0003815396540000107

Figure BDA0003815396540000108
Figure BDA0003815396540000108

式中，λα，λce分别代表换热修正系数与卷吸比率，

Figure BDA0003815396540000109
Figure BDA0003815396540000111
分别代表参数集合应满足的正态分布，
Figure BDA0003815396540000112
分别代表λα集合的期望与标准方差，
Figure BDA0003815396540000113
分别代表λce集合的期望与标准方差，下标j代表第j个集合成员。where λ α and λ ce represent the heat transfer correction coefficient and the entrainment ratio, respectively,
Figure BDA0003815396540000109
Figure BDA0003815396540000111
Respectively represent the normal distribution that the parameter set should satisfy,
Figure BDA0003815396540000112
Represent the expectation and standard deviation of the λ α set, respectively,
Figure BDA0003815396540000113
Represent the expectation and standard deviation of the λ ce set, respectively, and the subscript j represents the jth set member.

(二)将卷吸比率和换热修正系数以及校正后的烟气初始头部温度和烟气初始头部质量流量代入顶棚射流模型的顶棚射流速度公式，得到头部传播速度、头部在相邻位置间的传播时间和头部到达不同位置的时间：(2) Substituting the entrainment ratio and heat transfer correction coefficient, the corrected initial head temperature of the flue gas and the mass flow rate of the initial head of the flue gas into the ceiling jet velocity formula of the ceiling jet model, the head propagation velocity, the head in phase The propagation time between adjacent positions and the time for the head to reach different positions:

Figure BDA0003815396540000114
Figure BDA0003815396540000114

Figure BDA0003815396540000115
Figure BDA0003815396540000115

Δti→i+1＝(Li+1-Li)/Us，i Δt i→i+1 = (L i+1 -L i )/U s,i

ti+1＝ti+Δti→i+1(i＝0，1…n-1) 公式(6)t i+1 ＝t i +Δt i→i+1 (i=0, 1...n-1) Formula (6)

式中，

Figure BDA0003815396540000116
Ts，i和Us，i分别代表了顶棚射流头部热流量、头部烟气温度以及头部纵向蔓延速度W，T0和ρ0分别为隧道宽度、环境空气温度以及环境空气密度；Li为i位置距离火源中心的距离，n为待预报位置的总数。In the formula,
Figure BDA0003815396540000116
T s, i and U s, i respectively represent the heat flow at the head of the ceiling jet, the temperature of the smoke at the head and the longitudinal spreading speed W at the head, and T 0 and ρ 0 are the tunnel width, ambient air temperature and ambient air density, respectively; L i is the distance from position i to the center of the fire source, and n is the total number of positions to be predicted.

(三)将卷吸比率和换热修正系数以及校正后的烟气初始头部温度和烟气初始头部质量流量代入顶棚射流模型的质量守恒公式，得到头部厚度：(3) Substituting the entrainment ratio and heat transfer correction coefficient, the corrected initial head temperature of the flue gas, and the mass flow rate of the initial head of the flue gas into the mass conservation formula of the ceiling jet model to obtain the head thickness:

Figure BDA0003815396540000117
Figure BDA0003815396540000117

Figure BDA0003815396540000118
Figure BDA0003815396540000118

(四)计算头部从位置i到位置i+1蔓延过程中的换热损失，在计算过程中引入换热修正系数：(4) Calculate the heat transfer loss during the spreading process of the head from position i to position i+1, and introduce the heat transfer correction coefficient in the calculation process:

Figure BDA0003815396540000121
Figure BDA0003815396540000121

Figure BDA0003815396540000122
Figure BDA0003815396540000122

Figure BDA0003815396540000123
Figure BDA0003815396540000123

式中，αc为烟气与隧道衬砌结构间的换热系数；k0为空气导热系数；ν0为空气的运动粘度系数，α0为空气的热扩散系数；h′t为隧道断面的水力学直径；

Figure BDA0003815396540000124
为顶棚射流头部从位置i到i+1的换热量；Tw为隧道壁面温度。In the formula, α c is the heat transfer coefficient between the flue gas and the tunnel lining structure; k 0 is the thermal conductivity of the air; ν 0 is the kinematic viscosity coefficient of the air, α 0 is the thermal diffusivity of the air; h′ t is the thermal conductivity of the tunnel section hydraulic diameter;
Figure BDA0003815396540000124
is the heat transfer rate of the ceiling jet head from position i to i+1; T w is the temperature of the tunnel wall.

(五)将位置i处头部烟气温度、头部烟气质量流量以及换热损失代入顶棚射流模型的能量守恒公式计算位置i+1处的头部烟气温度，在计算过程中引入卷吸比率：(5) Substituting the head flue gas temperature at position i, the mass flow rate of head flue gas, and the heat transfer loss into the energy conservation formula of the ceiling jet model to calculate the head flue gas temperature at position i+1, and introduce volume in the calculation process Suction ratio:

Figure BDA0003815396540000125
Figure BDA0003815396540000125

Figure BDA0003815396540000126
Figure BDA0003815396540000126

循环步骤(二)～(五)直至计算到位置i＝n处的头部烟气参数，并输出预报结果。Steps (2) to (5) are repeated until the head smoke parameters at position i=n are calculated, and the forecast results are output.

(六)通过同化算法集合卡尔曼滤波同化头部到达位置处温度传感器所记录的烟气温度数据，实时校正头部烟气温度、头部烟气质量流量、卷吸比率和换热修正系数；(6) Assimilate the flue gas temperature data recorded by the temperature sensor at the arrival position of the head through the assimilation algorithm and gather the Kalman filter, and correct the flue gas temperature at the head, the mass flow rate of the flue gas at the head, the entrainment ratio and the heat transfer correction coefficient in real time;

即当位置i处温度传感器发生异常时，通过集合卡尔曼滤波同化该异常温度，实现对于关键参数λα，λce，以及头部温度的实时校正，同化算法采用公式(4)，其中，预报向量中参数是：That is, when the temperature sensor at position i is abnormal, the abnormal temperature is assimilated through the ensemble Kalman filter to realize real-time correction of the key parameters λ α , λ ce , and head temperature. The assimilation algorithm adopts formula (4), where the forecast The parameters in the vector are:

Figure BDA0003815396540000127
Figure BDA0003815396540000127

(七)将校正后的头部烟气温度、头部烟气质量流量、卷吸比率以及换热修正系数作为输入参数代入顶棚射流模型中，更新对后续头部烟气温度、头部烟气传播速度和头部烟气厚度进行超实时预报。(7) Substitute the corrected head smoke temperature, head smoke mass flow rate, entrainment ratio, and heat transfer correction coefficient into the ceiling jet model as input parameters, and update the subsequent head smoke temperature, head smoke Propagation velocity and head smoke thickness are predicted in real time.

当位置n处的传感器烟气温度数据发生异常被感知时，终止运行。When the sensor flue gas temperature data at position n is abnormal and sensed, the operation is terminated.

为了验证本发明的可行性，下面通过FDS仿真数据对本发明的效果进行验证。In order to verify the feasibility of the present invention, the effects of the present invention will be verified through FDS simulation data below.

通过隧道火灾模拟工况对该预报模型的效果进行验证，模型的物理尺寸、火源设置如图2所示，其中，隧道5宽度是13m，长度是210m，高度是5m，火源4距离隧道右侧出口105m，火源距离顶棚4m。The effect of the prediction model is verified through tunnel fire simulation conditions. The physical size and fire source settings of the model are shown in Figure 2. Among them, the tunnel 5 is 13m wide, 210m long, and 5m high, and the fire source 4 is far from the tunnel. The exit on the right is 105m away, and the fire source is 4m away from the ceiling.

图3是该工况下烟气动态发展模拟温度分布图。可以看到随着时间的增长，顶棚射流头部沿着纵向蔓延。距离火源位置越远，烟气的温度和传播速度都有明显的降低，整个蔓延过程大概持续了80s左右。Figure 3 is a simulated temperature distribution diagram of flue gas dynamic development under this working condition. It can be seen that the roof jet head spreads longitudinally with time. The farther away from the fire source, the temperature and propagation speed of the smoke decreased significantly, and the whole spreading process lasted about 80s.

仅通过同化温度测量对顶棚射流烟气蔓延过程进行预报。每一次同化后，本发明对于头部到达不同位置时的预报提前时间量以及预报值相对测量值的平均误差如图4所示。可以看出，第三次同化后，本发明预报方法已经展现了较好的预报效果。对不同位置处的平均预报相对误差均低于10％，且预报时间提前量高达54秒。对于顶棚射流头部蔓延时间只持续80秒的场景来说，这个时间提前量是相当可观的。The process of ceiling jet smoke propagation is predicted only by assimilation temperature measurement. After each assimilation, the forecast lead time and the average error of the forecast value relative to the measured value are shown in Figure 4 for the present invention when the head arrives at different positions. It can be seen that after the third assimilation, the forecasting method of the present invention has shown a better forecasting effect. The average forecast relative error for different positions is lower than 10%, and the forecast time advance is as high as 54 seconds. For a scene where the roof jet head spread lasts only 80 seconds, this is a considerable amount of time advance.

第三次同化后，本发明对顶棚射流后续阶段的超实时预报结果如图5、6所示，其中虚线点线代表对顶棚射流后续阶段头部参数发展态势的预报值，实线点线代表测量值带入预报过程之前的烟气发展情况。After the third assimilation, the super real-time prediction results of the present invention for the subsequent stage of the ceiling jet flow are shown in Figures 5 and 6, wherein the dashed dotted line represents the predicted value of the development trend of the head parameters in the subsequent stage of the ceiling jet flow, and the solid line dotted line represents Smoke development before measurements are brought into the forecasting process.

如图7所示，一种使用上述基于数据同化的隧道火灾烟气顶棚射流发展预报方法的系统，它包括依次连接的温度传感器1、数据采集器2、服务器3、控制系统和设置在隧道5内的消防装置6，温度传感器1在隧道5纵向中轴线上布设若干个且沿隧道5整个长度方向布满，各温度传感器1距离顶棚不超过20cm，且各温度传感器1等间距设置，且设置间距不超过20m。温度传感器1将采集的烟气温度数据传送给数据采集器2，由其处理后发送给服务器3，服务器3对烟气温度数据进行同化，并计算输出超实时预报结果，控制系统根据超实时预报结果提供疏散、喷淋排烟等策略制定参考并可以控制消防装置6，消防装置6是喷淋排烟装置等。As shown in Figure 7, a system that uses the above-mentioned method for forecasting the development of jet flow on the ceiling of tunnel fire smoke based on data assimilation, it includes a temperature sensor 1, a data collector 2, a server 3, a control system and a tunnel 5 that are connected in sequence. In the fire-fighting device 6 inside, several temperature sensors 1 are arranged on the longitudinal central axis of the tunnel 5 and covered along the entire length direction of the tunnel 5. The distance between each temperature sensor 1 is no more than 20 cm from the ceiling, and each temperature sensor 1 is arranged at equal intervals, and set The distance shall not exceed 20m. The temperature sensor 1 transmits the collected flue gas temperature data to the data collector 2, which processes it and sends it to the server 3. The server 3 assimilates the flue gas temperature data, calculates and outputs the super real-time forecast results, and the control system according to the super real-time forecast The results provide a reference for formulating strategies such as evacuation, spraying and smoke exhausting, and can control the fire-fighting device 6, which is a spraying and smoke-exhausting device.

本系统的工作过程是：当隧道发生火灾时，温度传感器首先检测到温度异常升高，这一信息传递到服务器被判定为火灾发生，最早两个温度异常升高的传感器所在位置正中间被判定为火源位置，认为烟气从该位置向隧道两边蔓延，服务器自动开启预报程序对烟气参数发展过程进行预报。随着隧道两边的温度传感器依次发生异常温升，温度传感器的烟气温度数据不断被数据采集器收集后，输入预报程序，预报程序将通过集合卡尔曼滤波同化该观测值，并更新对于后续烟气蔓延参数的超实时预报。超实时预报结果可作为疏散、喷淋排烟的控制决策的参考。The working process of this system is: when a fire breaks out in the tunnel, the temperature sensor first detects an abnormal increase in temperature, and this information is transmitted to the server to be judged as a fire, and the location of the first two sensors with an abnormally high temperature is determined to be in the middle is the position of the fire source, and the smoke is considered to spread from this position to both sides of the tunnel, and the server automatically starts the forecasting program to forecast the development process of the smoke parameters. As the temperature sensors on both sides of the tunnel experience abnormal temperature rises sequentially, the flue gas temperature data from the temperature sensors are continuously collected by the data collector and input into the forecasting program, which will assimilate the observed values through the ensemble Kalman filter and update the data for subsequent smoke Super real-time forecasting of gas spread parameters. The super real-time forecast results can be used as a reference for the control decision-making of evacuation, spraying and smoke exhaust.

Claims (6)

1. A tunnel fire smoke ceiling jet flow development forecasting method based on data assimilation is characterized by comprising the following steps:
s1, arranging a plurality of temperature sensors for recording smoke temperature data along a longitudinal central axis of a tunnel, wherein the temperature sensors are fully distributed along the whole length direction of the tunnel;
s2, judging the position of the fire source through the response time of the temperature sensors, namely considering that the fire source is in the middle of the two temperature sensors with the earliest temperature abnormality;
s3, generating the heat release rate of the fire source by using a Monte Carlo method, substituting the heat release rate into a fire plume model, calculating to obtain the initial head mass flow and the initial head temperature of the smoke, and then utilizing an assimilation algorithm set Kalman filtering to assimilate the average value of the smoke temperature data recorded by two temperature sensors with the earliest abnormal temperature to correct the heat release rate, the initial head mass flow and the initial head temperature of the smoke;
and S4, generating a entrainment ratio and a heat exchange correction coefficient by using a Monte Carlo method, substituting the entrainment ratio and the heat exchange correction coefficient and the corrected initial head mass flow of the smoke and the corrected initial head temperature of the smoke into a ceiling jet flow model as input parameters to realize super real-time prediction of the head temperature, the head propagation speed and the head thickness when the head reaches different positions, in the head spreading process, integrating the smoke temperature data recorded by a temperature sensor at the position where the head reaches by using an assimilation algorithm set Kalman filtering, correcting the entrainment ratio, the heat exchange correction coefficient, the head temperature and the head mass flow in real time, substituting the correction parameters into the ceiling jet flow model as the input parameters, and performing super real-time prediction on the head temperature, the head propagation speed and the head thickness when the head spreads to each position behind the position where the current temperature sensor exists again.
2. The method for forecasting development of smoke ceiling jet flow of tunnel fire based on data assimilation according to claim 1, characterized in that the S3 comprises:
generating an initial background field of the heat release rate of the fire source by using a Monte Carlo method:
Figure FDA0003815396530000011
in the formula ,
Figure FDA0003815396530000012
which is representative of the rate of heat release,
Figure FDA0003815396530000013
represents the normal distribution that the parameter set should satisfy,
Figure FDA0003815396530000014
respectively representing the expected and standard variances of a given parameter set, wherein a subscript j represents a jth set member, and a superscript f represents a predicted value;
(II) respectively bringing the heat release rate set members generated by the formula (1) into a plume flow calculation formula of the fire plume model to obtain the plume flow at the fire source
Figure FDA0003815396530000015
And initial head mass flow of flue gas:
Figure FDA0003815396530000021
Figure FDA0003815396530000022
Figure FDA0003815396530000023
Figure FDA0003815396530000024
Figure FDA0003815396530000025
wherein ,
Figure FDA0003815396530000026
is the plume flow at the fire source,
Figure FDA0003815396530000027
as a result of the overall heat release rate,
Figure FDA0003815396530000028
for convective heat release rates, the general value range is
Figure FDA0003815396530000029
To
Figure FDA00038153965300000210
h t Is the tunnel height, z 0 The position of the virtual ignition source is shown,
Figure FDA00038153965300000211
d' is the equivalent diameter of the fire source,
Figure FDA00038153965300000212
the ceiling jet flow mass flow is represented by a superscript f which represents a predicted value and a subscript j which represents a jth set member;
and (III) substituting the heat release rate and the plume flow into an energy conservation formula of the fire plume model to obtain the initial head temperature of the smoke gas:
Figure FDA00038153965300000213
in the formula ,Ts Is the average flue gas temperature above the fire source,
Figure FDA00038153965300000214
to convective heat release rate, C p Is the specific heat capacity at constant pressure (1004J/(kg. K)),
Figure FDA00038153965300000215
plume flow at the fire source, T 0 Is ambient temperature, the superscript f represents the predicted value, the subscript j represents the jth set member, belowMark i represents the arrival of the head at the ith position;
and (IV) assimilating the average value of the flue gas temperature data recorded by the two temperature sensors with the earliest abnormal temperature through the assimilation algorithm set Kalman filtering to obtain the correction values of the heat release rate, the initial head mass flow of the flue gas and the initial head temperature of the flue gas:
X a ＝X f +K x，y (y-HX f )
Figure FDA00038153965300000216
Figure FDA00038153965300000217
X f ＝[x 1 f ，x 2 f ，K，x j f ，K，x m f ]
Figure FDA0003815396530000031
Figure FDA0003815396530000032
Figure FDA0003815396530000033
wherein, X is an aggregate matrix formed by m vectors, H is an observation operator and comprises the relative position relation between an observation value and a predicted value vector, R is an observation value error, P is a prediction covariance matrix, K x，y Is a Kalman gain matrix, which is a prediction matrix composed of m prediction value vectors, x is the prediction value vector, y is the sensor temperature, T s Is the average flue gas temperature above the fire source,
Figure FDA0003815396530000034
is the mass flow of the jet flow of the ceiling,
Figure FDA0003815396530000035
for the total heat release rate, superscript a represents the assimilation value, superscript f represents the predicted value, and subscript j represents the jth set member.
3. The data assimilation tunnel fire smoke ceiling jet development forecasting method based on the claim 2 is characterized in that the S4 comprises the following steps:
firstly, an initial background field of entrainment ratio and heat exchange correction coefficient is generated by using a Monte Carlo method,
Figure FDA0003815396530000036
Figure FDA0003815396530000037
in the formula ,λα ，λ ce Respectively represent the heat exchange correction coefficient and the entrainment ratio,
Figure FDA0003815396530000038
Figure FDA0003815396530000039
respectively representing the normal distribution, λ, that the parameter set should satisfy α，μ ，
Figure FDA00038153965300000310
Respectively represents lambda α Expectation and standard deviation of the set, λ ce，μ ，
Figure FDA00038153965300000311
Each represents lambda ce Expectation of aggregationStandard deviation, subscript j represents the jth set member;
substituting the entrainment ratio, the heat exchange correction coefficient, the corrected initial head temperature of the smoke and the corrected initial head mass flow into a ceiling jet flow velocity formula of a ceiling jet flow model to obtain the head propagation velocity, the propagation time of the head between adjacent positions and the time for the head to reach different positions:
Figure FDA00038153965300000312
Figure FDA00038153965300000313
Δt i→i+1 ＝(L i-1 -L i )/U s，i
t i+1 ＝t i +Δt i→i+1 (i =0, 1.., n-1) formula (6)
in the formula ,
Figure FDA0003815396530000041
T s，i and Us，i Respectively represents the heat flow of the roof jet head, the temperature of the smoke of the head and the longitudinal spreading speed W, T of the head 0 and ρ0 Tunnel width, ambient air temperature and ambient air density, respectively; l is i The distance between the position i and the center of the fire source is calculated, and n is the total number of the positions to be forecasted;
and (III) substituting the entrainment ratio, the heat exchange correction coefficient, the corrected initial head temperature of the smoke and the corrected initial head mass flow of the smoke into a mass conservation formula of the ceiling jet flow model to obtain the head thickness:
Figure FDA0003815396530000042
Figure FDA0003815396530000043
(IV) calculating the heat exchange loss of the head in the process of spreading from the position i to the position i +1, and introducing a heat exchange correction coefficient in the calculation process:
Figure FDA0003815396530000044
Figure FDA0003815396530000045
Figure FDA0003815396530000046
in the formula ,αc The heat exchange coefficient between the flue gas and the tunnel lining structure is adopted; k is a radical of 0 Is the air thermal conductivity; v. of 0 Is the kinematic viscosity coefficient of air, alpha 0 Is the thermal diffusivity of air; h' t Is the hydraulic diameter of the tunnel section;
Figure FDA0003815396530000047
the heat exchange quantity from the position i to the position i +1 is the ceiling jet flow head; t is a unit of w Is the tunnel wall temperature;
(V) substituting the head temperature, the head mass flow and the heat exchange loss at the position i into an energy conservation formula of the ceiling jet flow model to calculate the head temperature at the position i +1, and introducing entrainment ratio in the calculation process:
Figure FDA0003815396530000048
Figure FDA0003815396530000049
collecting flue gas temperature data recorded by a temperature sensor at the i +1 position of the Kalman filtering assimilation position through an assimilation algorithm, and correcting the head temperature, the head mass flow, the entrainment ratio and the heat exchange correction coefficient in real time;
and (seventhly), substituting the corrected head temperature, head mass flow, entrainment ratio and heat exchange correction coefficient into the ceiling jet flow model as input parameters, and updating the super-real-time prediction of the subsequent ceiling jet flow head temperature, head propagation speed and head thickness.
4. A system using the data assimilation-based tunnel fire smoke ceiling jet flow development forecasting method as claimed in claim 1, wherein the method comprises the following steps: the system comprises a plurality of temperature sensors, a data acquisition unit, a server, a control system and a fire fighting device, wherein the temperature sensors, the data acquisition unit, the server, the control system and the fire fighting device are sequentially connected, the temperature sensors are distributed on a longitudinal central axis of a tunnel and are fully distributed along the whole length direction of the tunnel, the temperature sensors transmit collected smoke temperature data to the data acquisition unit, the data acquisition unit processes the data and transmits the data to the server, the server assimilates the smoke temperature data and calculates and outputs a super real-time forecast result, and the control system provides fire fighting strategy reference according to the super real-time forecast result and can control the fire fighting device.
5. The system of claim 4, wherein the temperature sensors are equally spaced and are spaced no more than 20m apart.
6. The system of claim 5, wherein the temperature sensor is no more than 20cm from the ceiling.
CN202211025127.XA 2022-08-25 2022-08-25 Method and system for forecasting jet flow development of fire smoke ceiling of tunnel based on data assimilation Active CN115526123B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211025127.XA CN115526123B (en) 2022-08-25 2022-08-25 Method and system for forecasting jet flow development of fire smoke ceiling of tunnel based on data assimilation

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211025127.XA CN115526123B (en) 2022-08-25 2022-08-25 Method and system for forecasting jet flow development of fire smoke ceiling of tunnel based on data assimilation

Publications (2)

Publication Number Publication Date
CN115526123A true CN115526123A (en) 2022-12-27
CN115526123B CN115526123B (en) 2023-05-09

Family

ID=84698381

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211025127.XA Active CN115526123B (en) 2022-08-25 2022-08-25 Method and system for forecasting jet flow development of fire smoke ceiling of tunnel based on data assimilation

Country Status (1)

Country Link
CN (1) CN115526123B (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116882268A (en) * 2023-06-15 2023-10-13 重庆大学 Data-driven tunnel fire smoke development prediction method and intelligent control system
CN118395368A (en) * 2024-03-21 2024-07-26 重庆大学 Fire disaster prediction method and system based on data assimilation coupling physical information neural network

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102162375A (en) * 2010-12-27 2011-08-24 中国安全生产科学研究院 On-site hot smoke test equipment and method for subway station and inter-station tunnel
CN102519598A (en) * 2011-06-27 2012-06-27 杭州电子科技大学 Fire source positioning method based on statistic characteristics of sensor array
CN109632793A (en) * 2018-12-13 2019-04-16 国网陕西省电力公司 Experiment porch and method for the research of cable tunnel fire temperature field simulated behavior
KR20210044092A (en) * 2019-10-14 2021-04-22 울산대학교 산학협력단 Method and apparatus for real-time ensemble streamflow forecasting

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102162375A (en) * 2010-12-27 2011-08-24 中国安全生产科学研究院 On-site hot smoke test equipment and method for subway station and inter-station tunnel
CN102519598A (en) * 2011-06-27 2012-06-27 杭州电子科技大学 Fire source positioning method based on statistic characteristics of sensor array
CN109632793A (en) * 2018-12-13 2019-04-16 国网陕西省电力公司 Experiment porch and method for the research of cable tunnel fire temperature field simulated behavior
KR20210044092A (en) * 2019-10-14 2021-04-22 울산대학교 산학협력단 Method and apparatus for real-time ensemble streamflow forecasting

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
钟委;李兆周;吕金金;梁天水;: "FDS6对隧道火灾温度场模拟的适用性研究", 郑州大学学报(工学版) *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116882268A (en) * 2023-06-15 2023-10-13 重庆大学 Data-driven tunnel fire smoke development prediction method and intelligent control system
CN116882268B (en) * 2023-06-15 2024-02-06 重庆大学 Data-driven tunnel fire smoke development prediction method and intelligent control system
CN118395368A (en) * 2024-03-21 2024-07-26 重庆大学 Fire disaster prediction method and system based on data assimilation coupling physical information neural network

Also Published As

Publication number Publication date
CN115526123B (en) 2023-05-09

Similar Documents

Publication Publication Date Title
CN115526123B (en) Method and system for forecasting jet flow development of fire smoke ceiling of tunnel based on data assimilation
EP4195091A1 (en) Target available model-based environment prediction method and apparatus, program, and electronic device
CN105807685B (en) A kind of intelligent monitor-type curtain wall system
CN102174994A (en) Pipe burst accident on-line positioning system for urban water supply pipeline network
CN109992900A (en) A multi-field real-time online collaborative intelligent simulation method and system for mass concrete
CN103914622A (en) Quick chemical leakage predicating and warning emergency response decision-making method
CN118395368B (en) Fire disaster prediction method and system based on data assimilation coupling physical information neural network
CN103984980B (en) The Forecasting Methodology of temperature extremal in a kind of greenhouse
CN111505205A (en) Improved search algorithm for strong back calculation of gas leakage source
CN105844361A (en) Wind power prediction method, cable untwisting method and device for wind turbine generator
CN105756082B (en) The ecological retaining wall that can be monitored in real time
CN115933785B (en) Environmental control method and system for a solar thermal agricultural greenhouse
CN105843140B (en) A kind of underground piping monitoring system for oil exploitation
CN116882268B (en) Data-driven tunnel fire smoke development prediction method and intelligent control system
CN116485165B (en) Forest fire control strategy formulation method, system and storage medium based on fusion factors
CN118750848A (en) A smart fire protection system based on digital twin technology
CN105678439A (en) Power transmission line dynamic capacity-increasing operation risk assessment method based on BP neural network
CN101308364A (en) Modeling method of water supply network event model based on transient flow analysis
CN113674512B (en) On-line monitoring and early warning system and method for electrified crossing construction site
CN118643676B (en) A method and system for monitoring the health status of soil-covered storage tanks based on digital twins
CN119691670A (en) Communication data fusion method and system based on edge calculation
CN105138048A (en) Device and method for intelligently controlling temperature and humidity of small environment of concrete pouring storehouse surfaces
CN112182888A (en) Method and device for identifying mechanical parameters of main control structural plane of small-sized sliding dangerous rock mass
CN105841738B (en) Water channel, the real-time monitoring protection system of river course both sides side slope
CN118821397A (en) A method and system for obtaining air cooling temperature of continuous casting billet

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant