CN114582083A - Tunnel monitoring multi-sensor data fusion fire early warning method - Google Patents

Tunnel monitoring multi-sensor data fusion fire early warning method Download PDF

Info

Publication number
CN114582083A
CN114582083A CN202210041690.XA CN202210041690A CN114582083A CN 114582083 A CN114582083 A CN 114582083A CN 202210041690 A CN202210041690 A CN 202210041690A CN 114582083 A CN114582083 A CN 114582083A
Authority
CN
China
Prior art keywords
data
fire
stage
model
tunnel
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
CN202210041690.XA
Other languages
Chinese (zh)
Other versions
CN114582083B (en
Inventor
冷朝霞
刘庆丰
常晓军
张晓珂
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.)
Xian University of Technology
Original Assignee
Xian University of Technology
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 Xian University of Technology filed Critical Xian University of Technology
Priority to CN202210041690.XA priority Critical patent/CN114582083B/en
Publication of CN114582083A publication Critical patent/CN114582083A/en
Application granted granted Critical
Publication of CN114582083B publication Critical patent/CN114582083B/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
    • G08B17/00—Fire alarms; Alarms responsive to explosion
    • G08B17/06—Electric actuation of the alarm, e.g. using a thermally-operated switch
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • 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
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A50/00—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE in human health protection, e.g. against extreme weather
    • Y02A50/20—Air quality improvement or preservation, e.g. vehicle emission control or emission reduction by using catalytic converters

Abstract

The invention discloses a tunnel monitoring multi-sensor data fusion fire early warning method, which is characterized in that three historical data of three stages before and after a tunnel fire occurs are normalized, nine neural network data relation models are established, the temperature, the carbon monoxide concentration and the carbon dioxide concentration in the current tunnel are collected, after normalization, every two data are input as a group and are respectively input into the three models of each stage to obtain the output value of each model, the output value is compared with the third data corresponding to each group of input data, if the third data are greater than the output of the model, the probability of the progress of a fire event is set to be 1/3, otherwise, the probability of the progress of the fire event is set to be 0, the probability sum of the progress of the fire event of each stage is counted, if the probability sum of the event of the alarm stage or the early warning stage of the fire is greater than 1/3, fire alarm is performed, otherwise, no alarm is performed, and if the sum of the safety phase event probabilities is larger than 1/3, performing fire early warning, otherwise, not performing fire early warning.

Description

Tunnel monitoring multi-sensor data fusion fire early warning method
Technical Field
The invention belongs to the technical field of information fusion, and relates to a tunnel monitoring multi-sensor data fusion fire early warning method.
Background
Transportation is an important guarantee for the development of the current socioeconomic, and the tunnel plays a role in transportation. However, the tunnel is a semi-closed driving channel, the visibility is limited, a large amount of waste gas discharged by automobiles exists in the tunnel, the tunnel contains a lot of harmful gases, the air quality in the tunnel is seriously polluted, and the visibility in the tunnel is also seriously influenced, so the probability of accidents occurring in the tunnel is obviously higher than other road sections of urban roads, and secondary accidents easily occur due to the particularity of the tunnel environment, and the tunnel monitoring and processing device has important significance for timely monitoring and processing tunnel conditions.
The tunnel fire monitoring system generally monitors the temperature, the CO concentration and the smoke concentration in a tunnel, and in recent years, a multi-sensor data fusion technology is adopted in the monitoring system to fuse various data information acquired on site and finally generate comprehensive evaluation on a target to be measured.
At present, a tunnel fire monitoring system based on a multi-sensor data fusion technology is mainly researched for fire monitoring and alarming, monitoring data such as temperature, gas concentration, smoke concentration and the like are subjected to data fusion processing by adopting a data fusion method and an intelligent algorithm of some nonlinear systems, and whether a fire disaster happens or not is subjected to alarm processing according to a data processing result, so that the research is only helpful for improving the rescue speed of the fire disaster happening in a tunnel, but cannot early warn in advance and cannot guide effective measures to be taken for removing the possible fire hazard.
Disclosure of Invention
The invention aims to provide a tunnel monitoring multi-sensor data fusion fire early warning method, which solves the problems that the existing tunnel fire monitoring system cannot early warn hidden danger of tunnel fire in time, effectively removes the hidden danger and avoids fire.
The invention adopts the technical scheme that a tunnel monitoring multi-sensor data fusion fire early warning method comprises the following steps:
step 1, acquiring three kinds of historical data of three stages before and after a tunnel fire occurs, wherein the three stages are a safety stage, a fire early warning stage and a fire occurrence warning stage, and the three kinds of historical data are the temperature in the tunnel, the concentration of carbon monoxide and the concentration of carbon dioxide;
step 2, respectively carrying out normalization processing on three types of historical data of three stages, and establishing three neural network data relation models by using the three types of historical data after normalization processing of each stage, wherein each model represents the relation between any two types of data and the third type of data in the stage;
step 3, acquiring the temperature, the carbon monoxide concentration and the carbon dioxide concentration in the current tunnel, carrying out normalization processing on the acquired data, and inputting each two kinds of data serving as a group of input into three neural network data relation models of each stage respectively to obtain an output value of each model;
step 4, comparing the output value of each model with third data corresponding to each group of input data, if the third data is greater than the output of the model, the group of data does not meet the neural network data relation model, and setting the probability of the fire event process to be 1/3, otherwise, setting the probability of the fire event process to be 0;
and 5, counting the probability sum of the fire event processes tested by the three neural network data relation models at each stage, if the probability sum of the events at the alarm stage or the fire early warning stage of the fire is greater than 1/3, carrying out fire alarm, otherwise, not carrying out fire alarm, if the probability sum of the events at the safety stage is greater than 1/3, carrying out fire early warning, otherwise, not carrying out fire early warning.
In step 2, the method for normalizing the three historical data of the three stages comprises
Figure BDA0003470434750000031
Wherein x represents any history data in three stages, xminRepresenting the minimum value in the historical data category corresponding to the x corresponding stage, xmaxIndicating that x corresponds to the maximum value in the historical data categories for the phase.
The acquired temperature, carbon monoxide concentration and carbon dioxide concentration in the current tunnel are also normalized according to the formula (1), wherein x represents any acquired data, and xminAnd xmaxAnd is not changed.
Each neural network data relation model comprises an input layer, an output layer and a hidden layer, wherein the input layer is used for receiving two data inputs, and the output layer is used for outputting third data.
In the neural network data relation model, when two input data of the first layer of the model are temperature and carbon monoxide concentration, the third data output by the third layer of the model is carbon dioxide concentration, when two input data of the first layer of the model are carbon monoxide concentration and carbon dioxide concentration, the third data output by the third layer of the model is temperature, two input data of the first layer of the model are temperature and carbon dioxide concentration, and the third data output by the third layer of the model is carbon monoxide concentration.
Adopt temperature sensor to gather the interior temperature of tunnel now, adopt carbon monoxide sensor to gather the interior carbon monoxide concentration of tunnel now, adopt carbon dioxide sensor to gather the interior carbon dioxide concentration of tunnel now.
The method has the advantages that the relation models of the tunnel multi-sensor data in different stages of the fire process are established by utilizing the three-layer neural network, the relation judgment is carried out on the collected multi-sensor data through the established three-layer neural network data relation model, the probability statistics of the fire event process in different stages and the event probability statistics in the safety stage are carried out according to the judgment result, so that the tunnel monitoring system has the tunnel fire early warning function, and the accuracy of the tunnel monitoring system in tunnel fire warning is improved.
Drawings
FIG. 1 is a graph showing temperature changes in three stages before and after a tunnel fire in an embodiment of the present invention;
FIG. 2 is a graph showing the variation of CO concentration in three stages before and after a tunnel fire occurs in the embodiment of the present invention;
FIG. 3 is a graph showing the change in concentration of carbon dioxide in three stages before and after a tunnel fire in the embodiment of the present invention;
FIG. 4 is a graph illustrating three data variations obtained after normalization of each data in the security phase according to an embodiment of the present invention;
FIG. 5 is a graph illustrating three data variations obtained after normalization processing of data in a fire early warning stage according to an embodiment of the present invention;
FIG. 6 is a graph showing three data changes obtained after normalization processing of data in the alarm stage in the case of fire occurrence according to an embodiment of the present invention;
FIG. 7 is a block diagram of a neural network data relationship model in the present invention.
Detailed Description
The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.
The invention discloses a tunnel monitoring multi-sensor data fusion fire early warning method, which comprises the following steps:
step 1, acquiring three types of historical data of three stages before and after a tunnel fire occurs, wherein the three stages are a safety stage, a fire early warning stage and a fire occurrence warning stage, the three types of historical data are the temperature in the tunnel, the concentration of carbon monoxide and the concentration of carbon dioxide, and the three types of historical data are shown in figures 1-3;
step 2, selecting three kinds of historical data corresponding to the same time point in each stage in fig. 1-3, and performing normalization processing on the three kinds of selected historical data of the three stages, wherein the data normalization processing method comprises the following steps:
Figure BDA0003470434750000051
wherein x represents any history data in three stages, xminRepresents the minimum value in the historical data category corresponding to the x corresponding stage, xmaxIndicating that x corresponds to the maximum value in the historical data categories for the phase.
Fig. 4 is a graph showing the variation of three kinds of data in the security phase obtained after normalization processing is performed on 10 data selected at the same time point in the security phase of fig. 1 to 3, fig. 5 is a graph showing the variation of three kinds of data in the fire early warning phase obtained after normalization processing is performed on 10 data selected at the same time point in the fire early warning phase of fig. 1 to 3, and fig. 6 is a graph showing the variation of three kinds of data in the fire alarm phase obtained after normalization processing is performed on 10 data selected at the same time point in the fire alarm phase of fig. 1 to 3;
three neural network data relation models are established by utilizing three types of historical data after normalization processing of each stage, each neural network data relation model comprises an input layer, an output layer and a hidden layer (namely each neural network data relation model is a three-layer neural network data relation model), the input layer is used for receiving two data inputs, the output layer is used for outputting third data, and each model represents the relation between any two types of data and the third data in the stage, and the reference is made to fig. 7.
In the neural network data relation model, when two input data of the first layer of the model are temperature and carbon monoxide concentration, the third data output by the third layer of the model is carbon dioxide concentration, when two input data of the first layer of the model are carbon monoxide concentration and carbon dioxide concentration, the third data output by the third layer of the model is temperature, two input data of the first layer of the model are temperature and carbon dioxide concentration, and the third data output by the third layer of the model is carbon monoxide concentration.
Namely, the data in fig. 4, 5 and 6 are respectively used as the training data of the three neural networks in each stage, and three neural network data relationship models in the security stage, three neural network data relationship models in the fire early warning stage and three neural network data relationship models in the fire occurrence warning stage can be respectively established, that is, nine neural network data relationship models can be established in total.
Step 3, adopting a temperature sensor to collect the temperature in the current tunnel, adopting a carbon monoxide sensor to collect the concentration of carbon monoxide in the current tunnel, and adopting a carbon dioxide sensor to collect the concentration of carbon monoxide in the current tunnelNormalizing the collected data according to the formula (1) when the concentration of carbon dioxide in the channel is in the channel, wherein x in the formula (1) represents any collected data, and xminAnd xmaxKeeping unchanged, then inputting every two kinds of data as a group of input into three neural network data relation models of each stage respectively to obtain the output value of each model, and finally obtaining the output values of nine models;
step 4, comparing the output value of each model with third data corresponding to each group of input data, if the third data is greater than the output of the model, the group of data does not meet the neural network data relation model, and setting the probability of the fire event process to be 1/3, otherwise, setting the probability of the fire event process to be 0;
and 5, counting the probability sum of the fire event processes tested by the three neural network data relation models in each stage, if the probability sum of the events in the fire alarm stage or the fire early warning stage is greater than 1/3, carrying out fire alarm, otherwise, not carrying out fire alarm, if the probability sum of the events in the safety stage is greater than 1/3, carrying out fire early warning, otherwise, not carrying out fire early warning.

Claims (5)

1. A tunnel monitoring multi-sensor data fusion fire early warning method is characterized by comprising the following steps:
step 1, acquiring three kinds of historical data of three stages before and after a tunnel fire happens, wherein the three stages comprise a safety stage, a fire early warning stage and a fire occurrence warning stage, and the three kinds of historical data comprise the temperature in the tunnel, the concentration of carbon monoxide and the concentration of carbon dioxide;
step 2, respectively carrying out normalization processing on three types of historical data of three stages, and establishing three neural network data relation models by using the three types of historical data after normalization processing of each stage, wherein each model represents the relation between any two types of data and the third type of data in the stage;
step 3, acquiring the temperature, the carbon monoxide concentration and the carbon dioxide concentration in the current tunnel, carrying out normalization processing on the acquired data, and inputting each two kinds of data serving as a group of input into three neural network data relation models of each stage respectively to obtain an output value of each model;
step 4, comparing the output value of each model with third data corresponding to each group of input data, if the third data is greater than the output of the model, the group of data does not meet the neural network data relation model, and setting the probability of the fire event process to be 1/3, otherwise, setting the probability of the fire event process to be 0;
and 5, counting the probability sum of the fire event processes tested by the three neural network data relation models at each stage, if the probability sum of the events at the alarm stage or the fire early warning stage of the fire is greater than 1/3, carrying out fire alarm, otherwise, not carrying out fire alarm, if the probability sum of the events at the safety stage is greater than 1/3, carrying out fire early warning, otherwise, not carrying out fire early warning.
2. The method for fire early warning through data fusion of tunnel monitoring and multiple sensors as claimed in claim 1, wherein in the step 2, the three historical data of the three stages are normalized by
Figure FDA0003470434740000021
Wherein x represents any history data in three stages, xminRepresenting the minimum value in the historical data category corresponding to the x corresponding stage, xmaxIndicating that x corresponds to the maximum value in the historical data categories for the phase.
3. The method of claim 1, wherein each neural network data relationship model comprises an input layer, an output layer and a hidden layer, the input layer is used for receiving two data inputs, and the output layer is used for outputting a third data.
4. The method as claimed in claim 3, wherein in the neural network data relationship model, when the two input data of the first layer of the model are temperature and carbon monoxide concentration, the third data outputted by the model is carbon dioxide concentration, when the two input data of the first layer of the model are carbon monoxide concentration and carbon dioxide concentration, the third data outputted by the model is temperature, the two input data of the first layer of the model are temperature and carbon dioxide concentration, and the third data outputted by the model is carbon monoxide concentration.
5. The method for fire early warning through data fusion of multiple sensors in tunnel monitoring according to claim 1, wherein a temperature sensor is used for collecting the temperature in the current tunnel, a carbon monoxide sensor is used for collecting the concentration of carbon monoxide in the current tunnel, and a carbon dioxide sensor is used for collecting the concentration of carbon dioxide in the current tunnel.
CN202210041690.XA 2022-01-14 2022-01-14 Tunnel monitoring multi-sensor data fusion fire disaster early warning method Active CN114582083B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210041690.XA CN114582083B (en) 2022-01-14 2022-01-14 Tunnel monitoring multi-sensor data fusion fire disaster early warning method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210041690.XA CN114582083B (en) 2022-01-14 2022-01-14 Tunnel monitoring multi-sensor data fusion fire disaster early warning method

Publications (2)

Publication Number Publication Date
CN114582083A true CN114582083A (en) 2022-06-03
CN114582083B CN114582083B (en) 2023-06-30

Family

ID=81771198

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210041690.XA Active CN114582083B (en) 2022-01-14 2022-01-14 Tunnel monitoring multi-sensor data fusion fire disaster early warning method

Country Status (1)

Country Link
CN (1) CN114582083B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115248873A (en) * 2022-09-22 2022-10-28 国网山西省电力公司太原供电公司 Cable tunnel safety monitoring method and system based on data fusion

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103136893A (en) * 2013-01-24 2013-06-05 浙江工业大学 Tunnel fire early-warning controlling method based on multi-sensor data fusion technology and system using the same
US20190371147A1 (en) * 2018-05-31 2019-12-05 Boe Technology Group Co., Ltd. Fire alarming method and device
CN111627181A (en) * 2020-06-28 2020-09-04 四川旷谷信息工程有限公司 Comprehensive pipe rack fire early warning method fusing multi-source parameters and gradient information thereof
CN112002095A (en) * 2020-07-14 2020-11-27 中国人民解放军63653部队 Fire early warning method in mine tunnel

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103136893A (en) * 2013-01-24 2013-06-05 浙江工业大学 Tunnel fire early-warning controlling method based on multi-sensor data fusion technology and system using the same
US20190371147A1 (en) * 2018-05-31 2019-12-05 Boe Technology Group Co., Ltd. Fire alarming method and device
CN111627181A (en) * 2020-06-28 2020-09-04 四川旷谷信息工程有限公司 Comprehensive pipe rack fire early warning method fusing multi-source parameters and gradient information thereof
CN112002095A (en) * 2020-07-14 2020-11-27 中国人民解放军63653部队 Fire early warning method in mine tunnel

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
吴会琴;: "基于信息融合技术的隧道火灾监控系统", 电气时代, no. 02 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115248873A (en) * 2022-09-22 2022-10-28 国网山西省电力公司太原供电公司 Cable tunnel safety monitoring method and system based on data fusion

Also Published As

Publication number Publication date
CN114582083B (en) 2023-06-30

Similar Documents

Publication Publication Date Title
CN107798876B (en) Road traffic abnormal jam judging method based on event
JP4242422B2 (en) Sudden event recording and analysis system
CN110217238B (en) Driving risk grade judgment and optimization method
CN103150900A (en) Traffic jam event automatic detecting method based on videos
CN106406287A (en) Method and system for vehicle safety state monitoring and early warning
CN116105802B (en) Underground facility safety monitoring and early warning method based on Internet of things
CN116975378B (en) Equipment environment monitoring method and system based on big data
CN113190653B (en) Traffic data monitoring system based on block chain
CN115077627B (en) Multi-fusion environmental data supervision method and supervision system
CN104408578A (en) Track-point-based quantitative assessment system and method for mechanical operation
CN114863675A (en) Method for predicting position of key vehicle and alarming abnormity based on road police data fusion
CN115018343A (en) System and method for recognizing and processing abnormity of mass mine gas monitoring data
CN114582083A (en) Tunnel monitoring multi-sensor data fusion fire early warning method
CN113761728A (en) Airport electric special vehicle fault early warning method based on Internet of vehicles platform
CN115358647A (en) Hydrogen energy industry chain risk monitoring system and monitoring method based on big data
CN116881749A (en) Pollution site construction monitoring method and system
CN110060370B (en) Equivalent statistical method for times of rapid acceleration and rapid deceleration of vehicle
CN116758762A (en) Control method based on big data
CN113901043B (en) Pollution source intelligent supervision and data fusion analysis method and system
CN113888866B (en) Road vehicle management system with multistage early warning function
CN115565373A (en) Real-time risk prediction method, device, equipment and medium for highway tunnel accident
CN111780809B (en) Rail vehicle part temperature and vibration monitoring and early warning method and system thereof
CN112989069B (en) Traffic violation analysis method based on knowledge graph and block chain
JP3164100B2 (en) Traffic sound source type identification device
CN111798091B (en) Expressway lane change scoring model building method based on lane change duration

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