CN113205669B - Fused fire-fighting Internet of things monitoring and early warning method - Google Patents

Fused fire-fighting Internet of things monitoring and early warning method Download PDF

Info

Publication number
CN113205669B
CN113205669B CN202110393906.4A CN202110393906A CN113205669B CN 113205669 B CN113205669 B CN 113205669B CN 202110393906 A CN202110393906 A CN 202110393906A CN 113205669 B CN113205669 B CN 113205669B
Authority
CN
China
Prior art keywords
fire
detector
alarm
delta
signal
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.)
Active
Application number
CN202110393906.4A
Other languages
Chinese (zh)
Other versions
CN113205669A (en
Inventor
关劲夫
袁宏永
王宇
梁光华
苏国锋
疏学明
吴津津
贺胜
王静舞
于淼淼
付明
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.)
Chen An Tianze Zhilian Technology Co ltd
Hefei Institute for Public Safety Research Tsinghua University
Original Assignee
Chen An Tianze Zhilian Technology Co ltd
Hefei Institute for Public Safety Research Tsinghua 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 Chen An Tianze Zhilian Technology Co ltd, Hefei Institute for Public Safety Research Tsinghua University filed Critical Chen An Tianze Zhilian Technology Co ltd
Priority to CN202110393906.4A priority Critical patent/CN113205669B/en
Publication of CN113205669A publication Critical patent/CN113205669A/en
Application granted granted Critical
Publication of CN113205669B publication Critical patent/CN113205669B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G08—SIGNALLING
    • G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
    • G08B29/00—Checking or monitoring of signalling or alarm systems; Prevention or correction of operating errors, e.g. preventing unauthorised operation
    • G08B29/18—Prevention or correction of operating errors
    • G08B29/185—Signal analysis techniques for reducing or preventing false alarms or for enhancing the reliability of the system
    • G08B29/188—Data fusion; cooperative systems, e.g. voting among different detectors
    • G—PHYSICS
    • G08—SIGNALLING
    • G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
    • G08B17/00—Fire alarms; Alarms responsive to explosion
    • G08B17/06—Electric actuation of the alarm, e.g. using a thermally-operated switch
    • G—PHYSICS
    • G08—SIGNALLING
    • G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
    • G08B17/00—Fire alarms; Alarms responsive to explosion
    • G08B17/10—Actuation by presence of smoke or gases, e.g. automatic alarm devices for analysing flowing fluid materials by the use of optical means
    • G08B17/117—Actuation by presence of smoke or gases, e.g. automatic alarm devices for analysing flowing fluid materials by the use of optical means by using a detection device for specific gases, e.g. combustion products, produced by the fire
    • G—PHYSICS
    • G08—SIGNALLING
    • G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
    • G08B17/00—Fire alarms; Alarms responsive to explosion
    • G08B17/12—Actuation by presence of radiation or particles, e.g. of infrared radiation or of ions
    • G08B17/125—Actuation by presence of radiation or particles, e.g. of infrared radiation or of ions by using a video camera to detect fire or smoke

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Business, Economics & Management (AREA)
  • Emergency Management (AREA)
  • Engineering & Computer Science (AREA)
  • Computer Security & Cryptography (AREA)
  • Chemical & Material Sciences (AREA)
  • Analytical Chemistry (AREA)
  • Multimedia (AREA)
  • Fire Alarms (AREA)

Abstract

The invention relates to the field of fire monitoring, in particular to a fused monitoring and early warning method for a fire-fighting Internet of things. The method comprises the following steps: s1, acquiring a theoretical fire source position theta through a detector which sends out an alarm signal; s2, forming actual time difference delta t by the time when any two detectors send out alarm signalsl' obtaining maximum likelihood position of theoretical fire source position theta
Figure DDA0003017800170000011
S3, calculating the time difference T formed by the time when any two detectors send out the alarm signal aiming at the fire source position in the step S2; s4, obtaining the corrected time difference
Figure DDA0003017800170000012
S5, calculating delta tl' and
Figure DDA0003017800170000013
relative error delta of formationk(ii) a S6, if δk≤δminIf yes, the alarm signal is true, namely fire disaster occurs, and fire disaster processing is carried out, otherwise, the step S7 is carried out; s7, if Δ tl′≤Δt′lmaxAnd k is less than or equal to kmaxIf yes, the alarm signal is true, and fire disaster processing is carried out, otherwise, the step S8 is carried out; and S8, informing the fire processing unit through an alarm system. The invention judges whether the fire really occurs or not through a plurality of detectors, and can reduce the misjudgment of the fire as much as possible.

Description

Fused fire-fighting Internet of things monitoring and early warning method
Technical Field
The invention relates to the field of fire monitoring, in particular to a fused monitoring and early warning method for a fire-fighting Internet of things.
Background
The fire detector is arranged in the building, so that whether a fire disaster happens in the building can be detected, and related workers can timely perform fire fighting and extinguishing treatment according to the alarm signal sent by the detector.
However, the existing detector can send out the alarm signal by mistake, and people outside the building can not identify the truth of the alarm signal sent out by the detector, when the alarm signal is false, related people still adopt the measure of fire fighting, and the waste of resources can be caused.
Disclosure of Invention
In order to solve the technical problems, the invention provides a fused monitoring and early warning method for a fire-fighting Internet of things, which can identify the truth of an alarm signal sent by a detector and further judge whether a fire really occurs.
In order to achieve the purpose, the invention adopts the following technical scheme:
a fused fire-fighting Internet of things monitoring and early warning method comprises the following steps:
s1, acquiring a theoretical fire source position theta through a detector which sends out an alarm signal; wherein, the detector is used for collecting fire signals;
s2, forming actual time difference delta t by the time when any two detectors send out alarm signalsl' obtaining maximum likelihood position of theoretical fire source position theta
Figure BDA0003017800150000011
S3, maximum likelihood position
Figure BDA0003017800150000012
Setting the position of the fire source, and calculating the time difference T formed by the time when any two detectors in the step S2 send out alarm signals aiming at the position of the fire source;
s4, by Δ tl' the time difference T in step S3 is corrected by the maximum likelihood probability, and the corrected time difference is obtained
Figure BDA0003017800150000013
S5, calculating delta tl' and
Figure BDA0003017800150000014
relative error delta of formationkK is the calculated maximum likelihood position
Figure BDA0003017800150000015
The total number of iterations of (c);
s6, if δk≤δminIf the alarm signal is true, namely fire disaster occurs, carrying out fire disaster processing, otherwise, carrying out step S7; wherein, deltaminIs a constant;
s7, if Δ tl′≤Δt′lmaxAnd k is less than or equal to kmaxIf yes, the alarm signal is true, and fire disaster treatment is carried out, otherwise, fire disaster treatment is carried outStep S8, wherein kmaxAnd Δ t'lmaxAre all constants;
and S8, informing the fire handling unit through an alarm system for determining whether a fire is occurring.
Furthermore, the fire-fighting Internet of things alarm method also comprises the steps of dividing the severity of fire, wherein the fire probability levels comprise level I, level II, level III and level IV;
the fire-fighting Internet of things alarm method is used for building fire alarm, and a point type fire detector, a video fire detector, an electrical fire detector and a gas fire detector are mounted on a building;
if the two point type fire detectors detect a fire signal; or, a fire detector of the spot type detects a fire signal and a video fire detector also detects a fire signal; or, the two video fire detectors detect fire signals; or two gas fire detectors detect fire signals; the grade is I grade and corresponds to the highly questioned plausible alarm;
if a point type fire detector detects a fire signal and an electrical fire detector also detects a fire signal; or a video fire detector and an electrical fire detector also detect a fire signal; the grade is II grade and corresponds to moderate suspected plausibility alarm;
if a point type fire detector detects a fire signal and a gas fire detector also detects a fire signal; or, a video fire detector detects a fire signal and a gas fire detector also detects a fire signal; the grade is grade III and corresponds to a high-risk warning situation;
if one electrical fire detector detects a fire signal and one gas fire detector also detects a fire signal; or, if the two electric fire detectors detect fire signals, the grade is IV grade, and the fire detector corresponds to a medium-risk warning situation.
Further, the maximum likelihood position in step S2 is acquired
Figure BDA0003017800150000021
The method comprises the following specific steps:
s21, acquiring the theoretical fire source position theta through the position of the detector which sends the alarm signal;
s22, acquiring the building structure micro-element corresponding to the building through the building where the detector is located;
s23, assuming that detectors at any two positions in the building structure micro-element receive fire source information at a theoretical fire source position theta and send alarm signals, and recording the time interval of sending the alarm signals by the detectors at the two positions as delta t; the position of the detector in the building structure element corresponds to the position of the detector in the building;
s24, acquiring an edge probability density function of a fire probability density function corresponding to a theoretical fire source position theta in the building with respect to delta t through theta and delta t;
s25, establishing Bayes estimation of the edge probability density function;
s26, establishing a fire probability likelihood function through Bayesian estimation;
s27, the fire source position corresponding to the maximum value of the fire probability likelihood function iteration k times is the maximum likelihood position
Figure BDA0003017800150000031
Further preferably, the fire probability likelihood function y in step S26:
Figure BDA0003017800150000032
wherein Δ t 'is Δ t'lCorresponding parameters, dividing Δ t into n parts, Δ tiIs the time corresponding to the ith part of the n parts, Δ ti-1P (theta) is the prior probability of fire occurrence at the theoretical fire source position theta, f (delta t | theta) is the marginal probability density function of the fire probability density function f (delta t, theta) relative to delta t,
Figure BDA0003017800150000033
is taken for Δ t
Figure BDA0003017800150000034
An error distribution function of time;
projecting a theoretical fire source position theta to a rectangle formed by a horizontal plane of a building, wherein any one wide side of the rectangle is marked as L, and dividing grids at equal intervals along the length direction of the rectangle by taking the wide side L as a starting point, wherein theta isjIs the distance between the center of the jth cell and the broadside L, f (θ)j) Is the edge integral of f (Δ t, θ) with respect to Δ t, f (Δ t | θ)j) Take θ for θ in f (Δ t | θ)jA function of time.
It is further preferred that the first and second liquid crystal compositions,
Figure BDA0003017800150000035
in the formula, σ is a constant.
It is further preferred that the first and second liquid crystal compositions,
Figure BDA0003017800150000041
further, the number of alarm signals is recorded as N, N and kmaxThe following relation is satisfied:
kmax＝N-1。
further, δminThe value of (2) is 0.1.
Further preferably, the specific process of step S21 is as follows:
and marking the detector, wherein the mark of the detector corresponds to the position of the detector in the building, and the theoretical fire source position theta is obtained through the mark of the detector which sends an alarm signal.
Further, Δ t'lmaxValue range of [60s,600s ]]。
The invention has the following beneficial effects:
(1) the present invention obtains a time difference formed by the time when a detector installed at each position in a building detects a fire signal (fire smoke spreads to the detector) according to the propagation path of the fire smoke in the building when a fire occurs. Whether the fire really happens or not is judged through the plurality of detectors, and misjudgment of the fire can be reduced as much as possible.
(2) The invention adopts the theoretical time difference and the time difference obtained after correcting the theoretical time difference as the parameters for obtaining the relative error, and can improve the accuracy of the truth of the alarm signal obtained according to the relative error.
(3) The invention adopts the maximum likelihood position, not only can know whether the fire really happens, but also can conjecture the position of the fire source, thereby facilitating the subsequent rescue of fire fighters.
(4) In the invention, the theoretical alarm time difference in the relative error is obtained based on the maximum likelihood position, and the maximum possibility of generating alarm signal time difference by two detectors in theory is reflected from the statistical angle.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 is a process diagram of the present invention for implementing fire alarm from three levels of hardware, algorithm and pre-alarm mechanism.
Detailed Description
The technical scheme of the invention is clearly and completely described below by combining the embodiment and the attached drawings of the specification. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
A fused fire-fighting Internet of things monitoring and early warning method is shown in figure 1 and comprises the following steps:
s1, obtaining maximum likelihood position
Figure BDA0003017800150000051
Theoretical time difference and maximum likelihood position formed by corresponding time for sending out alarm signal
Figure BDA0003017800150000052
Being the maximum likelihood location of the theoretical fire source location theta,
Figure BDA0003017800150000053
by emitting an alarm signal with a time difference Δ tl' obtaining; the theoretical fire source position theta is obtained through an alarm signal; it is composed ofMiddle, Δ tl' is the actual time difference of the alarm signal acquired at the first time of the maximum likelihood iteration. The method comprises the following specific steps:
s11, obtaining the theoretical fire source position θ through the detector sending the alarm signal, i.e. marking the detector, where the mark of the detector corresponds to the position of the detector in the building, and obtaining the theoretical fire source position θ through the mark of the detector sending the alarm signal, for example: the detector marked A1 is located at the position A of the building, and when the detector marked A1 sends out an alarm signal, the position of the fire source at the position A in the building can be known.
S12, acquiring a building structure infinitesimal corresponding to the building through the building where the detector is located; the building infinitesimal is a model, and the building structure infinitesimal with the same structure as the structure for sending the alarm signal is established by referring to the building design fire protection code (GB 50016-2014).
S13, assuming a time difference delta t formed by the time when the detector sends the alarm signal in the building structure micro element, wherein the position of the detector in the building structure micro element corresponds to the position of the detector in the building; for example, a detector at a location S in a building emits an alarm signal, then the assumed detector is also S at the location of a building structure infinitesimal.
S14, acquiring an edge probability density function of a fire probability density function corresponding to a theoretical fire source position theta in the building with respect to delta t through theta and delta t;
s15, establishing Bayes estimation of the edge probability density function;
s16, establishing a fire probability likelihood function y through Bayesian estimation:
Figure BDA0003017800150000061
wherein
Figure BDA0003017800150000062
Wherein Δ t 'is Δ t'lCorresponding parameters, dividing Δ t into n parts, Δ tiIs the i-th part of n partsTime of day, Δ ti-1P (theta) is the prior probability of fire occurrence at the theoretical fire source position theta, f (delta t | theta) is an edge probability density function of a fire probability density function f (delta t, theta) relative to delta t at the time corresponding to the (i-1) th part of the n parts,
Figure BDA0003017800150000063
taken for Δ t
Figure BDA0003017800150000064
An error distribution function of time;
projecting the theoretical fire source position theta to a rectangle formed by a building on a horizontal plane where the building is located, wherein the length of the rectangle is the length of an area where the theoretical fire source position theta is located, the width of the rectangle is the width of the area where the theoretical fire source position theta is located, any one wide side of the rectangle is marked as L, grids are divided at equal intervals along the length direction of the rectangle by taking the wide side L as a starting point, and theta is equal to the width of the area where the theoretical fire source position theta is locatedjIs the distance between the center of the jth cell and the broadside L, f (θ)j) Is the edge integral of f (Δ t, θ) with respect to Δ t, f (Δ t | θ)j) Take θ for θ in f (Δ t | θ)jA function of time.
n is equal fraction of Δ t, Δ tiAt a time corresponding to the end of the i-th part of the n parts, Δ ti-1The time corresponding to the end of the (i-1) th part of n parts is divided into equal parts by delta t, for example, the theoretical time difference delta t corresponding to the alarm signals sent by the two detectors is 10 seconds and is divided into equal parts by 5, the value of n is 5, and the delta t isiIs the time corresponding to the end of the 4 th second, Δ ti-1P (theta) is the prior probability of fire occurrence at the theoretical fire source position theta at the moment corresponding to the end of 6 seconds, f (delta t | theta) is an edge probability density function of a fire probability density function f (delta t, theta) relative to delta t,
Figure BDA0003017800150000065
for a time difference Δ t
Figure BDA0003017800150000066
Error distribution function of time, thetajF (θ) is the position where the j-th part after θ is equally dividedj) Is the edge integral of f (Δ t, θ) with respect to Δ t, f (Δ t | θ)j) Is composed ofWhen theta in f (delta t | theta) is taken as thetajFunction of time, Δ t' being Δ tl' corresponding parameter, Δ tl' As a specific value, all Δ tlThe specific values represented by can be collectively expressed by Δ t'.
S17, the maximum value of the fire probability likelihood function iterates for k times is the maximum likelihood position
Figure BDA0003017800150000071
S2, by Δ tl' correcting the theoretical time difference by using the maximum likelihood probability to obtain the corrected time difference
Figure BDA0003017800150000072
S3，Δtl' and
Figure BDA0003017800150000073
relative error delta of formationkK is the calculated maximum likelihood position
Figure BDA0003017800150000074
The number of iterations of (c):
Figure BDA0003017800150000075
s4, if δk≤δminIf the alarm signal is true, namely fire disaster occurs, carrying out fire disaster processing, otherwise, carrying out step S5; wherein, deltaminIs a constant, δminThe value of (A) is 0.1.
S5, if Δ tl′≤Δt′lmaxAnd k is less than or equal to kmaxIf yes, the alarm signal is true, and fire processing is performed, otherwise, step S6 is performed, wherein kmaxAnd Δ t'lmaxAre all constants; the number of alarm signals is recorded as N, N and kmaxSatisfies the following relation:
kmax＝N-1
and S6, informing the fire handling unit through the alarm system for determining whether a fire is occurring.
The fire-fighting alarm method of the Internet of things can also be used for dividing the grade of the severity of the building into I grade, II grade, III grade and IV grade, and the severity of the fire corresponding to the I grade, II grade, III grade and IV grade is reduced in sequence;
the fire-fighting Internet of things alarm method is used for building fire alarm, and a point type fire detector, a video fire detector, an electrical fire detector and a gas fire detector are mounted on a building;
if the two point type fire detectors detect fire signals; or, a fire detector of the spot type detects a fire signal and a video fire detector also detects a fire signal; or, the two video fire detectors detect fire signals; or two gas fire detectors detect fire signals; the grade is I grade;
if a point fire detector detects a fire signal and an electrical fire detector also detects a fire signal; or a video fire detector and an electrical fire detector also detect the fire signal; the grade is II;
if a point type fire detector detects a fire signal and a gas fire detector also detects a fire signal; or, a video fire detector detects a fire signal and a gas fire detector also detects a fire signal; the grade is grade III;
if one electrical fire detector detects a fire signal and one gas fire detector also detects a fire signal; or, if two electric fire detectors detect fire signals, the fire level is IV level.
The point type fire detector is JTY-GD-G3/G3T of gulf; the gas fire detector is JB-KR-GSTN004 of bay, the video fire detector is TT-H-FANT6091 of French fire, and the electrical fire detector is GST-DH9000 of bay.
The following specifically illustrates the principles of the present invention:
aiming at the defect of high false alarm rate of the traditional fire alarm system (detector), the invention adopts a multi-point detector linkage method to know whether a fire really occurs in a building (namely, whether the fire really occurs is known through the time difference of alarm signals sent by a plurality of detectors), and through reasonable design of a hardware layer, an algorithm layer and an actual working layer, early fire alarm is really realized and the false alarm rate is reduced. Based on the thought, the invention comprises three major parts, namely a hardware design (each detector), a principle and an algorithm (the alarm method of the invention) and a pre-alarm mechanism, and is characterized in that the invention is shown in figure 2 and specifically comprises the following contents:
the hardware design part mainly completes the detection of fire signals, the transmission, the processing and the storage of fire parameters on a hardware level, and guides the subsequent fire prevention and control work according to the processing result of the fire parameters, and the series of processes are completed through a big data platform. According to the thought, the hardware design part is divided into a fire protection Internet of things big data platform, a fire detector, data transmission equipment, data processing equipment and data storage equipment, the fire detector is divided into four types of a point type fire detector, a gas fire detector, a video fire detector and an electrical fire detector according to the fire signal type and the requirement of subsequent fire prevention and control work, the installation of the point type fire detector, the gas fire detector, the video fire detector and the electrical fire detector accords with the design Specification for an automatic fire alarm system (GB 50116 plus 2013), and the point type fire detector, the gas fire detector, the video fire detector and the electrical fire detector and other equipment of the fire monitoring and early warning system form a fire monitoring and early warning hardware system based on the fire protection Internet of things, as shown in figure 2.
The principle and algorithm part mainly provides a method for processing fire signals of different scenes and a method for judging fire conditions from the principle and algorithm level. In order to achieve the purpose, firstly, the concept of building structure infinitesimal is provided, the transmission modes of fire signals under different building structures are respectively described, and a fire pre-alarm model is constructed according to the transmission characteristics of the fire signals of each building structure.
When the fire detector monitors an actual fire signal, an actual fire alarm time difference delta t 'is obtained, and Bayesian estimation is carried out on a fire probability density function according to the actual fire alarm time difference delta t':
Figure BDA0003017800150000091
in the formula:
f (delta t' | delta t) — reflecting the error distribution rule between the theoretical value and the observed value of the alarm time difference, approximately following normal distribution:
Figure BDA0003017800150000092
in order to facilitate the work of the data processing equipment, the alarm time parameter and the fire source position are discretized, and the alarm time difference is divided into regions
Figure BDA0003017800150000093
Divided into n parts, each interval expressed as:
Ui＝(Δti-1，Δti]，i＝1，2，...，n
linearly processing each interval probability density function:
Figure BDA0003017800150000094
when the spatial parameters are subjected to discretization processing, the building structure micro-elements are divided into m subspaces, and when a fire disaster occurs but the position is not determined, the prior probability that the fire source occurs in each subspace is considered to be equal, namely:
Figure BDA0003017800150000095
accordingly, a Bayesian estimation of the probability density of each time interval with respect to the alarm time difference observation value delta t' is established:
Figure BDA0003017800150000096
in order to find the most probable fire source position under the observed value delta t' of the alarm time difference
Figure BDA0003017800150000097
Establishing a fire probability switchAt the position p of the fire sourcex，yLikelihood function of (c):
Figure BDA0003017800150000098
in the formula:
p(θ|Ui) -representing in a time interval UiAnd the probability of the fire source appearing at theta is expressed as:
Figure BDA0003017800150000101
under the condition of considering time equal division and space equal division, obtaining a maximum likelihood point of the fire probability relative to the fire source position:
Figure BDA0003017800150000102
in practice, there may be more than two sensors that detect fire signals, and each additional alarm signal will update a series of parameters by iteration, assuming that there are a total of k +1 alarm signals, the updated parameters include:
Δt′＝(Δt1′，Δt2′，...，Δtk′)
f(Δt|Δt′)＝f(Δt|Δt1′，Δt2′，...，Δtk′)
fΔt(Ui|Δt′)＝fΔt(Ui|Δt1′，Δt2′，...，Δtk′)
p(θ|Δt′)＝p(θ|Δt1′，Δt2′，...，Δtk′)
Figure BDA0003017800150000103
along with the increase of the number of the alarm signals, the fire probability theoretical value is closer to the true value, and the fire source theoretical position is also closer to the actual situation. However, an increase in the number of alarm signals also means that the fire is closer to the time node of the disaster, which is disadvantageous for early prevention and control of the fire. The actual fire protection work needs to be effectively combined with other auxiliary means.
And then, judging the authenticity of the alarm signal through an early warning discrimination model, and providing a basis for subsequent work. The early warning discrimination model comprises three parts of fire source position determination, theoretical alarm time difference calculation and true and false alarm judgment. In the early warning discrimination model, determining the fire probability maximum likelihood position obtained by the fire probability calculation model as the position of a suspected fire source:
Figure BDA0003017800150000104
and calculating the theoretical alarm time difference under the condition that the fire source is generated at the position by combining fire scene models with different structural infinitesimals:
Figure BDA0003017800150000111
in the formula:
p is structural infinitesimal type, number 1, 2, 3, 4, 5;
Δtpan empirical formula is calculated based on the alarm time difference of five common structural micro elements.
Judging the truth of the alarm signal according to the error between the alarm time difference estimation value and the observation value, and constructing a judgment formula as follows:
Figure BDA0003017800150000112
when deltakWhen the signal is less than or equal to 0.1, judging that the alarm signal is true, and simultaneously starting a pre-alarm mechanism;
when deltakAnd when the fire disaster alarm signal is greater than 0.1, judging that the alarm signal is temporary false, accessing the next alarm signal, calculating by using the fire disaster pre-alarm model again, and further judging according to the calculation result.
Considering the urgency of actual fire prevention and control work, the fire condition can be judged as accurately as possible by combining a scientific fire pre-alarm mechanism on the basis of a fire pre-alarm model according to fire pre-alarm signals as few as possible. This requires the analysis of the alarm time error delta in addition tokBesides, the single alarm time difference delta t 'should be considered'lUpper limit of (3) and maximum number of iterations k of the fire pre-alarm modelmaxThey are affected by the fire scenario and the detector operation. Therefore, a set of fire pre-alarm mechanism needs to be established to realize the combination of the scientificity of the fire principle and algorithm and the urgency of the actual fire prevention and control work.
The pre-alarm mechanism part mainly combines the fire prediction alarm method provided by the hardware part, the principle and the algorithm part and the actual requirements of the fire prevention and control work to formulate the pre-alarm flow and the graded pre-alarm response rule related by the invention: the pre-alarm flow analyzes the detector alarm signal access from a hardware level, an algorithm level and a pre-alarm mechanism level → the fire pre-alarm model calculation → the implementation process of pre-alarm information disposal. The pre-alarm information processing comprises two types of processing modes, namely single-detector intelligent voice alarm checking, multi-detector linkage alarm checking and hierarchical response. And a graded pre-alarm response rule part, wherein a multi-detector linkage pre-alarm rule is formulated by combining the actual types and the arrangement of the fire detectors, and a 16 linkage alarm rule is established in a mode of combining two fire detectors which are actually used by a point type fire detector, a gas fire detector, a video fire detector and an electrical fire detector. And pre-alarm grades are divided according to the linkage types of the detectors, namely high-doubtful plausibility alarms, medium-doubtful plausibility alarms, high-risk alarms and medium-risk alarms, and the linkage rules of each detector correspond to the four pre-alarm grades (I grade, II grade, III grade and IV grade) one by one. And finally, reasonably setting each structural infinitesimal element and a pre-alarm parameter set under each alarm linkage rule by combining a fire probability calculation model and field models under different structural infinitesimal elements and considering the urgency of actual fire prevention and control work, wherein the pre-alarm parameter set specifically comprises a single alarm time difference upper limit delta t'lmaxAnd fire pre-alarm modelMaximum number of iterations kmaxIn total, the pre-alarm parameter set thresholds in 80 cases are listed. Thus forming an upper limit delta t 'considering the single alarm time difference'lmaxAnd maximum number of iterations kmaxFire pre-alarm rules under the circumstances when the pre-alarm mechanism is activated, i.e. deltakWhen the temperature is less than or equal to 0.1, delta t'lOr when k exceeds the upper limit, the fire pre-alarm model is not calculated any more, and the truth of the alarm is checked by adopting an intelligent voice alarm checking mode, wherein the method specifically comprises the following conditions:
Figure BDA0003017800150000121
if the signal is true, the signal issues pre-alarm information and takes corresponding fire prevention and control measures;
Figure BDA0003017800150000122
when the signal is false, the intelligent voice alarm is checked, and the next step is taken according to the alarm checking result;
Figure BDA0003017800150000123
when the signal is false, the intelligent voice alarm is checked, and the next step is taken according to the alarm checking result;
Figure BDA0003017800150000124
and meanwhile, temporarily false signals are generated, the intelligent voice is used for checking the alarm, and the next step of measures are taken according to the result of the alarm checking.
And performing alarm handling according to the result of the pre-alarm information verification, wherein the alarm handling comprises two handling modes of alarm grading handling and alarm relieving. Thus completing the design of the pre-alarm mechanism part.
Examples
The method comprises a hardware part, a principle and algorithm part and a pre-alarm mechanism part, and based on the design, the fire pre-alarm is realized from three layers of hardware, algorithm and pre-alarm mechanism, and the false alarm is reduced. The following is specifically set forth:
a fire pre-alarm system is established on the basis of a fire-fighting Internet of things big data platform. The platform can be accessed to fire signals received by various fire detectors, monitored fire parameters are uploaded to the big data platform in real time through the data transmission equipment, the fire signal processing is completed through the data processing equipment arranged in the platform, the processing result is fed back to relevant departments through the data transmission equipment on the one hand, and meanwhile, the data storage equipment is used for storing necessary data, so that data support is provided for improvement of a fire pre-alarm algorithm and daily fire prevention work.
When a fire disaster occurs or occurs nearby, a fire detector arranged at the front end senses a fire signal, and specifically comprises four basic types of fire detectors, namely a point type fire detector, a gas fire detector, a video fire detector and an electrical fire detector. And entering a disposal stage immediately after receiving the detector alarm signal. Because the fire detector has position marks when being installed, the building structure infinitesimal where the alarm detector is located can be immediately inquired, and the building structure infinitesimal specifically comprises six basic structure infinitesimal types including a cavity, a transverse long and narrow structure, a vertical long and narrow structure, a cavity-cavity, a cavity-transverse long and narrow structure and a cavity-vertical long and narrow structure. After the structural infinitesimal is determined, a fire pre-alarm model can be started, the alarm time of a fire detector and the theoretical value of the alarm time difference are calculated by constructing a fire field model, and five basic field models, specifically, a cavity fire field model, a transverse long and narrow structure fire field model, a vertical long and narrow structure fire field model, a cavity-cavity fire field model, a cavity-vertical long and narrow structure fire field model and a cavity-transverse long and narrow structure fire field model are included according to the transmission mode of plumes in fire. In order to ensure the accuracy of the alarm time difference and the fire source position, a fire probability calculation model is adopted to correct the alarm time and the position, the relation between the fire probability and the alarm time difference and the space position is established through a fire probability distribution function, after a plurality of alarm signals are generated, Bayesian estimation is carried out on the fire probability density according to the observation value of the alarm time difference, the actual working efficiency of data processing equipment is considered, a discretization data processing mode is adopted to carry out discretization processing on the alarm time difference and the space, a likelihood function of the fire occurrence probability about the space position is established, maximum likelihood estimation is carried out, and a maximum likelihood point is returned. Considering the urgency of actual fire prevention and control work, the fire probability calculation model only completes limited iterations, the last returned maximum likelihood point is considered to be the estimated value of the fire source position, the fire source position is determined by the early warning judgment model, the alarm time estimated value based on the fire source estimated position is calculated by combining with the fire field model, and the alarm time estimated value is compared with the alarm time observation value to judge the truth of the alarm signal.
When deltakWhen the signal is less than or equal to 0.1, judging that the alarm signal is true, and simultaneously starting a pre-alarm mechanism; when deltakAnd when the fire disaster alarm signal is greater than 0.1, judging that the alarm signal is temporary false, accessing the next alarm signal, calculating by using the fire disaster pre-alarm model again, and further judging according to the calculation result.
After the pre-alarm mechanism is started, determining pre-alarm levels according to detector linkage alarm rules determined by the types of a plurality of alarm detectors, wherein the pre-alarm levels comprise four pre-alarm levels of high-doubt plausibility alarm, medium-doubt plausibility alarm, high-risk alarm and medium-risk alarm, and the actual alarm time difference delta t is calculatedlComparing a parameter set consisting of iteration times k of the model with a pre-alarm parameter set threshold value under the same scene, and determining the authenticity of an alarm signal according to the following rules:
Figure BDA0003017800150000141
if the signal is true, issuing pre-alarm information and taking corresponding fire prevention and control measures;
Figure BDA0003017800150000142
temporarily falsifying the signal, performing intelligent voice alarm checking, and taking the next step of measures according to an alarm checking result;
Figure BDA0003017800150000143
temporarily false signals, performing intelligent voice alarm checking, and taking the next step according to the alarm checking result;
Figure BDA0003017800150000144
and (3) temporarily false signals, performing intelligent voice alarm checking, and taking next step measures according to the alarm checking result.
And taking alarm condition grading treatment or alarm condition treatment measures for releasing the alarm according to the result of the pre-alarm information verification.

Claims (10)

1. A fused fire-fighting Internet of things monitoring and early warning method is characterized by comprising the following steps:
s1, acquiring a theoretical fire source position theta through a detector which sends out an alarm signal; wherein, the detector is used for collecting fire signals;
s2, forming actual time difference delta t by the time when any two detectors send out alarm signalsl' obtaining maximum likelihood position of theoretical fire source position theta
Figure FDA0003538141410000011
S3, maximum likelihood position
Figure FDA0003538141410000012
Setting the position of the fire source, and calculating the time difference T formed by the time when any two detectors in the step S2 send out alarm signals aiming at the position of the fire source;
s4, by Δ tl' the time difference T in step S3 is corrected by the maximum likelihood probability, and the corrected time difference is obtained
Figure FDA0003538141410000013
S5, calculating delta tl' and
Figure FDA0003538141410000014
relative error delta of formationkK is the calculated maximum likelihood position
Figure FDA0003538141410000015
Of (a) stackTotal number of generations;
s6, if δk≤δminIf yes, the alarm signal is true, namely fire disaster occurs, and fire disaster processing is carried out, otherwise, the step S7 is carried out; wherein, deltaminIs a constant;
s7, if Δ tl′≤Δt′lmaxAnd k is less than or equal to kmaxIf yes, the alarm signal is true, and fire processing is performed, otherwise, the process proceeds to step S8, where kmaxAnd Δ t'lmaxAre all constants;
and S8, informing the fire handling unit through the alarm system for determining whether a fire is occurring.
2. The fusion type monitoring and early warning method for the fire-fighting internet of things as claimed in claim 1, wherein the alarm method for the fire-fighting internet of things further comprises the steps of dividing the severity degree of the fire, wherein the fire probability levels comprise level I, level II, level III and level IV;
the fire-fighting Internet of things alarm method is used for building fire alarm, and a point type fire detector, a video fire detector, an electrical fire detector and a gas fire detector are mounted on a building;
if the two point type fire detectors detect fire signals; or, a fire detector of the spot type detects a fire signal and a video fire detector also detects a fire signal; or, the two video fire detectors detect fire signals; or two gas fire detectors detect fire signals; the grade is I grade and corresponds to the highly questioned plausible alarm;
if a point type fire detector detects a fire signal and an electrical fire detector also detects a fire signal; or a video fire detector and an electrical fire detector also detect the fire signal; the grade is II grade, corresponding to moderate suspicion police;
if a point type fire detector detects a fire signal and a gas fire detector also detects a fire signal; or, a video fire detector detects a fire signal and a gas fire detector also detects a fire signal; the grade is grade III, and the high risk warning condition is corresponded;
if one electrical fire detector detects a fire signal and one gas fire detector also detects a fire signal; or if the two electrical fire detectors detect fire signals, the grade is IV grade, and the fire detectors correspond to medium-risk warning situations.
3. The fused fire-fighting internet of things monitoring and early warning method according to claim 1, characterized in that: the maximum likelihood position in step S2 is acquired
Figure FDA0003538141410000023
The method comprises the following specific steps:
s21, acquiring the theoretical fire source position theta through the position of the detector which sends the alarm signal;
s22, acquiring a building structure infinitesimal corresponding to the building through the building where the detector is located;
s23, assuming that detectors at any two positions in the building structure micro-element receive fire source information at a theoretical fire source position theta and send alarm signals, and recording the time interval of sending the alarm signals by the detectors at the two positions as delta t; the position of the detector in the building structure element corresponds to the position of the detector in the building;
s24, acquiring an edge probability density function of a fire probability density function corresponding to a theoretical fire source position theta in the building relative to delta t through theta and delta t;
s25, establishing Bayesian estimation of an edge probability density function;
s26, establishing a fire probability likelihood function through Bayesian estimation;
s27, the fire source position corresponding to the maximum value of the fire probability likelihood function iteration k times is the maximum likelihood position
Figure FDA0003538141410000021
4. The fused fire-fighting internet of things monitoring and early warning method as claimed in claim 3, wherein the fire probability likelihood function y in the step S26 is:
Figure FDA0003538141410000022
wherein Δ t 'is Δ t'lCorresponding parameters, dividing Δ t into n parts, Δ tiAt the time corresponding to the ith of the n copies,. DELTA.ti-1P (theta) is the prior probability of fire occurrence at the theoretical fire source position theta, f (delta t | theta) is the marginal probability density function of the fire probability density function f (delta t, theta) relative to delta t,
Figure FDA0003538141410000031
is taken for Δ t
Figure FDA0003538141410000032
An error distribution function of time;
projecting a theoretical fire source position theta to a rectangle formed by a horizontal plane of a building, wherein any one wide side of the rectangle is marked as L, and dividing grids at equal intervals along the length direction of the rectangle by taking the wide side L as a starting point, wherein theta isjIs the distance between the center of the jth cell and the broadside L, f (θ)j) Is the edge integral of f (Δ t, θ) with respect to Δ t, f (Δ t | θ)j) Take θ for θ in f (Δ t | θ)jA function of time.
5. The fused fire-fighting internet of things monitoring and early warning method according to claim 4, characterized in that:
Figure FDA0003538141410000033
in the formula, σ is a constant.
6. The fused fire-fighting Internet of things monitoring and early warning method as claimed in any one of claims 1 to 5,
Figure FDA0003538141410000034
7. the fusion type fire-fighting Internet of things monitoring and early warning method as claimed in claim 1, wherein the number of the warning signals is recorded as N, N and kmaxThe following relation is satisfied:
kmax＝N-1。
8. the fused monitoring and early warning method for the fire-fighting internet of things as claimed in claim 1, characterized in that: deltaminThe value of (A) is 0.1.
9. The fused fire-fighting internet of things monitoring and early warning method as claimed in claim 3, wherein the specific process of the step S21 is as follows:
and marking the detector, wherein the mark of the detector corresponds to the position of the detector in the building, and the theoretical fire source position theta is obtained through the mark of the detector which sends the alarm signal.
10. The fused monitoring and early warning method for the fire-fighting internet of things as claimed in claim 1, characterized in that: delta t'lmaxValue range of [60s,600s ]]。
CN202110393906.4A 2021-04-13 2021-04-13 Fused fire-fighting Internet of things monitoring and early warning method Active CN113205669B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110393906.4A CN113205669B (en) 2021-04-13 2021-04-13 Fused fire-fighting Internet of things monitoring and early warning method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110393906.4A CN113205669B (en) 2021-04-13 2021-04-13 Fused fire-fighting Internet of things monitoring and early warning method

Publications (2)

Publication Number Publication Date
CN113205669A CN113205669A (en) 2021-08-03
CN113205669B true CN113205669B (en) 2022-07-19

Family

ID=77026714

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110393906.4A Active CN113205669B (en) 2021-04-13 2021-04-13 Fused fire-fighting Internet of things monitoring and early warning method

Country Status (1)

Country Link
CN (1) CN113205669B (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113741258B (en) * 2021-08-25 2023-01-13 浙江省交通投资集团有限公司智慧交通研究分公司 Rail transit station fire monitoring system based on Internet of things and optimization method thereof
CN114333225A (en) * 2021-12-08 2022-04-12 江苏昂内斯电力科技股份有限公司 Household electric, gas and fire alarm linkage control system and method

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2013257847A (en) * 2012-06-14 2013-12-26 New Cosmos Electric Corp Method for transmitting alarm tone
CN105741474A (en) * 2016-04-11 2016-07-06 泉州师范学院 Fire early-warning method based on multiple sensors
CN105956664A (en) * 2016-04-27 2016-09-21 浙江大学 Tracing method for sudden river point source pollution
CN110097727A (en) * 2019-04-30 2019-08-06 暨南大学 Forest Fire Alarm method and system based on fuzzy Bayesian network
CN112258362A (en) * 2020-09-27 2021-01-22 汉威科技集团股份有限公司 Danger source identification method, system and readable storage medium
CN112347208A (en) * 2020-10-20 2021-02-09 燕山大学 Multi-target detection and tracking method based on distributed sensor network

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107798811B (en) * 2017-10-26 2019-07-16 上海腾盛智能安全科技股份有限公司 A kind of tunnel fire monitoring device, monitoring system and monitoring method

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2013257847A (en) * 2012-06-14 2013-12-26 New Cosmos Electric Corp Method for transmitting alarm tone
CN105741474A (en) * 2016-04-11 2016-07-06 泉州师范学院 Fire early-warning method based on multiple sensors
CN105956664A (en) * 2016-04-27 2016-09-21 浙江大学 Tracing method for sudden river point source pollution
CN110097727A (en) * 2019-04-30 2019-08-06 暨南大学 Forest Fire Alarm method and system based on fuzzy Bayesian network
CN112258362A (en) * 2020-09-27 2021-01-22 汉威科技集团股份有限公司 Danger source identification method, system and readable storage medium
CN112347208A (en) * 2020-10-20 2021-02-09 燕山大学 Multi-target detection and tracking method based on distributed sensor network

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
基于贝叶斯估计的多探测器火警判定方法研究;刘罡等;《中国安全生产科学技术》;20210131;第17卷(第1期);第12-18页 *

Also Published As

Publication number Publication date
CN113205669A (en) 2021-08-03

Similar Documents

Publication Publication Date Title
CN113205669B (en) Fused fire-fighting Internet of things monitoring and early warning method
Zervas et al. Multisensor data fusion for fire detection
US20150009031A1 (en) Multilayer perimeter instrusion detection system for multi-processor sensing
CN113990018B (en) Safety risk prediction system
KR102322427B1 (en) Bigdata based building fire prevention response system and method
CN111311869B (en) Fire safety monitoring method and system based on area alarm model and cloud platform
CN113129569B (en) Fusion type fire-fighting Internet of things monitoring and early warning signal identification method
US20160258766A1 (en) Vehicle localization and transmission method and system using a plurality of communication methods
US10271016B2 (en) Integrated monitoring CCTV, abnormality detection apparatus, and method for operating the apparatus
CN108650139A (en) A kind of powerline network monitoring system
CN111369761B (en) Early warning range determining method, device, equipment and system
CN107331090A (en) A kind of indoor fire alarm evacuation method
CN103033648B (en) A kind of wind sensor output data validity detection method
CN111986436B (en) Comprehensive flame detection method based on ultraviolet and deep neural networks
CN113393665A (en) Planning method for dangerous goods transportation path under uncertain time-varying road network
CN113516820A (en) Fire early warning method and fire early warning system
Sekkas et al. A multi-level data fusion approach for early fire detection
CN114435172B (en) Automatically-managed intelligent charging pile and intelligent charging method for new energy automobile
CN107063168B (en) Building deformation monitoring and early warning system that collapses
CN113222221A (en) Public safety risk early warning system and method
JP2001144669A (en) Sound source position detection system
Yoshida et al. Incident alert by an anomaly indicator of probe trajectories
CN112672299B (en) Sensor data reliability evaluation method based on multi-source heterogeneous information fusion
CN114708701A (en) Travel safety information monitoring method and system based on intelligent well lid
GB2608639A (en) Threat assessment system

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