CN103116961B - A kind of confined space fire detection alarm system based on Electronic Nose Technology and method - Google Patents
A kind of confined space fire detection alarm system based on Electronic Nose Technology and method Download PDFInfo
- Publication number
- CN103116961B CN103116961B CN201310026143.5A CN201310026143A CN103116961B CN 103116961 B CN103116961 B CN 103116961B CN 201310026143 A CN201310026143 A CN 201310026143A CN 103116961 B CN103116961 B CN 103116961B
- Authority
- CN
- China
- Prior art keywords
- fire
- network model
- sensor
- pnn
- data
- 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.)
- Expired - Fee Related
Links
Abstract
Based on confined space fire detection alarm system and the method for Electronic Nose Technology, comprising: sampling pipe, filtration unit, three gas sensors, temperature sensor, sensor power unit, sensor control unit, signal gathering unit, signal condition unit, monitoring host computer, acoustic-optic alarm, flowmeter, vacuum pump, three-way solenoid valve and exhaust gas processing devices.The present invention can on-line analysis, study, intelligent decision, so that at pole early detection fire, and reduces wrong report, rate of failing to report, reduces the infringement of fire to confined space personnel and equipment to a greater extent.
Description
Technical field
The invention belongs to technical field of fire detection, be specifically related to the confined space fire detection alarm system based on Electronic Nose Technology and method.
Background technology
Confined space refers to that import and export are limited, and natural ventilation is bad, the unconventional finite space of isolating relative to the external world.Common confined space mainly comprises some power distribution cabinet, machine room, goods and materials freight house, aircraft hold and space capsule etc.This type of confined space is once breaking out of fire, and the smog of generation, poison gas and heat will certainly be assembled at short notice in a large number, cause great infringement to personnel and equipment.Therefore, early stage fire detecting and alarm seems particularly important.
At present, realize reporting to the police mainly for the change of smokescope visible during fire in confined space, produce wrong report than being easier to by the interference such as dust in air, steam, fail to report.Fire characteristic gas results from fire and occurs extremely early stage, prior to the appearance of visible smog.Further, CO and CO of natural fire generation
2very regular Deng characteristic gas concentration change, be beneficial to detection.But gas sensor has selectivity (namely for multiple gases cross sensitivity), use single sensor to carry out the warning of fire characteristic detection of gas and be easily subject to the impact of other gas or environmental factor and cause wrong report.And utilize Electronic Nose Technology to address this problem well.Electronic Nose Technology forms primarily of gas sensor array, Signal Pretreatment and pattern-recognition three part.The algorithm for pattern recognition that the Electronic Nose Technology being applied to detection field at present adopts mainly contains BP neural network, support vector machine (SVM) etc.This type of Algorithm for Training time is long, and complex structure is unfavorable for online updating network model, thus the generation causing wrong report, fail to report.
Summary of the invention
The technology of the present invention is dealt with problems: overcome the deficiencies in the prior art, provides a kind of confined space fire detection alarm system based on Electronic Nose Technology and method, extracts fire characteristic gas O
2, CO, CO
2changing Pattern, on-line analysis, study, intelligent decision, so that at pole early detection fire, and reduces wrong report, rate of failing to report, reduces the infringement of fire to confined space personnel and equipment to a greater extent.
The technology of the present invention solution: a kind of confined space fire detection alarm system based on Electronic Nose Technology, as shown in Figure 1, it is made up of sampling pipe, filtration unit, three gas sensors, temperature sensor, sensor power unit, sensor control unit, signal gathering unit, signal condition unit, monitoring host computer, acoustic-optic alarm, flowmeter, vacuum pump, three-way solenoid valve, exhaust gas processing devices etc. its structure.
Gas sensor gathers the O at monitored scene through sampling pipe
2, CO, CO
2deng signal, after data prediction, input the PNN network model trained, prediction fire probability U
i.If fire probability is greater than given threshold values U
d1, then start to record pre-warning time t (i), otherwise then data inputted PNN network model on-line training program.Carry out secondary judgement, if fire probability is more than or equal to given threshold values U
d2, then directly open acoustic-optic alarm and connect exhaust gas processing device; Otherwise, then judge whether pre-warning time t (i) is more than or equal to given threshold values t
d: be then open acoustic-optic alarm and connect exhaust gas processing device; No, then data are inputted PNN network model on-line training program.Fig. 2 is real-time Fire process flow diagram in the present invention.
The present invention is based on Electronic Nose Technology, adopt three to O
2, CO, CO
2the gas sensor of cross sensitivity and a temperature sensor composition sensor array image data, adopt probabilistic neural network (PNN) as its algorithm for pattern recognition.In order to reduce model complexity and training difficulty, adding filtration unit at sampling pipe end and adopting flowmeter control by the gas flow rate of sensor.Filtration unit has two objects: the smoke particle in (1) sampling by filtration gas, extends sensor life-time; (2) dry sample gas, reduces water vapor to the impact of gas sensor.When determining optimum gas velocity, configurable, organizing O more
2, CO, CO
2mixed gas sample, changes flow, and under obtaining multithread speed, gas sensor response curve judges.
The present invention adopts probabilistic neural network model (PNN) to carry out pattern-recognition.PNN, based on Bayes classification theory decision-making, utilizes Parzen window method to carry out estimated probability density function, has many excellent performances: (1) PNN training speed is fast, is beneficial to real-time application; (2) decision-making can realize Bayes Optimum; (3) fault-tolerant ability is strong; (4), when adding new training sample, re-training need not be carried out to network.
Need to carry out off-line training to it before employing PNN network model.Off-line training flow process as shown in Figure 3.First the on-the-spot disaster hidden-trouble of confined space is investigated, and carry out simulated fire experiment with this.Simulated fire field data is obtained by gas sensor and temperature sensor.Because the present invention is devoted to real-time monitoring, therefore gas sensor response S can only be chosen
i, gas sensor rate of change Δ S
iwith temperature-responsive value T
ipre-service is carried out as individual features.Data preprocessing method also directly affects the operating characteristic of system.Adopt array normalization can reach good effect, computing formula is:
Wherein, X
i,jrepresent jth eigenwert when measuring for i-th time,
value after representation transformation.By the data after normalization
carry out PCA principal component analysis (PCA), by dimensionality reduction, find N (N < j) individual orthogonal characteristic variable, i.e. major component, make it to reflect data
principal character, the scale of compression legacy data matrix.When top n principal component contributor rate reaches 90%, then think that this N number of major component can reflect
principal character, be used for the former data of matching.Utilize top n major component, i.e. characteristic parameter, form new sample set X={x
i,n| n=1,2 ..., N}, and the training being applied to newly-built PNN network model.When exporting expectation and meeting the demands, stop training, the PNN network model of formation is input in monitoring host computer.
The present invention is while real time and on line monitoring, and the data write PNN network model training sample that can will obtain, carries out retraining to it, thus optimize PNN network model.On-line training flow process as shown in Figure 4.Monitoring site data, after fire judges, send into online training program.For at sequence of threads, when system long-time stable, the data of write training sample are constantly tending towards identical, and waste storage space, reduce the generalized ability of model.Therefore, whether need be in stable state to system to judge.Setting threshold values θ, when meeting:
‖X(i+1)-X(i)‖＜θ
Then think that system enters stable state, thus stop on-line training, maintain PNN network model constant.When system is not in stable state, then using the desired output of the output of fire hazard monitoring master routine as this sample, the training sample set adding PNN network model carries out on-line training, until test sample book desired output satisfies condition.The new PNN network model obtained re-writes monitoring host computer, carries out follow-up monitoring.
When fire probability is near 0.5, directly judge whether that breaking out of fire is unreasonable.The present invention introduces pre-warning time and carries out intelligent decision to fire probability, increases warning fiduciary level.Pre-warning time is defined as follows:
U () representation unit step signal, i ∈ [0, T).T is cycle length, as i=T, resets pre-warning time.As fire probability U
ibe greater than U
d1time, start pre-warning time timing t (i).If U
ibe greater than U
d2, then directly start sound and light alarm, connect exhaust gas processing device and reset pre-warning time t (i); No, then need further judgement.Judge whether pre-warning time t (i) is more than or equal to given threshold values t
d, if so, then start warning, record and reset pre-warning time t (i); No, then data Input Online training program is reset pre-warning time t (i).
Based on a confined space fire detecting and alarm method for Electronic Nose Technology, it is characterized in that performing step is as follows:
(1) first fire accident investigation is carried out to confined space, find out its main fire risk, and carry out simulated experiment for this several principal risk.
(2) according to simulated experiment the data obtained, off-line training is carried out to PNN network model.After data normalization, carry out PCA principal component analysis (PCA), and be applied to the training of newly-built PNN network model.When exporting expectation and meeting the demands, stop training, the PNN network model of formation is input in monitoring host computer.
(3) the PNN network model write monitoring host computer will trained.Systematic sampling pipe is placed in confined space to be monitored, and three-way solenoid valve connects air, and turn on sensor power supply unit and control module, open vacuum pump and start to bleed, and start monitoring host computer.Obtain fire probability from the data of monitoring site collection by PNN network model, carry out real-time fire identification in conjunction with pre-warning time judgment mechanism.
(4) the monitoring site real time data obtained is utilized to carry out on-line training to PNN network model.First whether be in stable state to system to judge.When system is in stable state, stop on-line training, maintain PNN network model constant; When system is not in stable state, then training sample set data being added PNN network model carries out on-line training, obtains new PNN network model, and re-writes monitoring host computer, to carry out follow-up monitoring.
The present invention's advantage is compared with prior art:
(1) the present invention utilizes the gas concentration change produced prior to smog during Electronic Nose Technology process fire, the conventional smoke detector of time of fire alarming use conventional prior to confined space and video smoke detector, and reduces wrong report, rate of failing to report.
(2) the present invention adopts the PNN network model that serious forgiveness is stronger, be more conducive to on-line training, and introduces stable state judgement, realizes confined space fire and monitors in real time and online updating.
(3) the present invention introduces pre-warning time judgment mechanism, makes the method more intelligent, increases the fiduciary level of reporting to the police.
Accompanying drawing explanation
Fig. 1 is the structural representation of the confined space fire detection alarm system that the present invention is based on Electronic Nose Technology;
Fig. 2 is real-time Fire process flow diagram in the present invention;
Fig. 3 is probabilistic neural network (PNN) off-line training process flow diagram in the present invention;
Fig. 4 is probabilistic neural network (PNN) on-line training process flow diagram in the present invention.
Embodiment
As shown in Figure 1, the confined space fire detection alarm system that the present invention is based on Electronic Nose Technology comprises; Sampling pipe, filtration unit, three gas sensors, temperature sensor, sensor power unit, sensor control unit, signal gathering unit, signal condition unit, monitoring host computer, acoustic-optic alarm, flowmeter, vacuum pump, three-way solenoid valve and exhaust gas processing devices; Add filtration unit at sampling pipe end and adopt flowmeter control by the gas flow rate of three gas sensors.
Sensor power unit provides power supply, its output terminal connecting sensor for three gas sensors and temperature sensor; Sensor control unit controls unlatching, the closedown of three gas sensors and temperature sensor, its output terminal connecting sensor; Three sensors are respectively O
2sensor, CO sensor and CO
2sensor; Signal gathering unit is used for the data of collection three gas sensors and temperature sensor, and its input end is connected with sensor, and output terminal is connected with signal condition unit; Signal condition unit can amplify the signal collected, filtering, and its input end is connected with signal gathering unit, and output terminal is connected with monitoring host computer; Monitoring host computer is provided with fire hazard monitoring master routine and PNN Neural Network Online training algorithm, can Real time identification fire provide warning, online updating PNN neural network and storage data, and its output terminal connects acoustic-optic alarm and three-way solenoid valve.
Sampling pipe gathers, conveying confined space gas on-site data, and its one end is as in confined space, and the other end connects filtration unit; Water vapor in filtration devices gas and smoke particle, reduce the impact on sensor, and its one end connects sampling pipe, and the other end connects gas analysis chamber by the road; Gas analysis chamber is used for placement three gas sensors and temperature sensor, and its other end is connection traffic meter by the road; Flow velocity in flowmeter pilot piping, its other end connects vacuum pump by the road; Vacuum pump is bled, and for systematic sampling provides power, its other end is connecting tee solenoid valve by the road; Three-way solenoid valve controls the flow direction of tail gas, and its other end connects exhaust gas processing device by the road; Exhaust gas processing device can process CO, CO in sample gas
2deng harmful gas, preventing pollution.
Under normal condition, three-way solenoid valve is communicated with air, and flowmeter controls the vacuum pump speed of evacuation, and the gas of confined space is by sampling pipe, and device enters gas analysis chamber after filtration.Place 3 gas sensors and a temperature sensor in gas analysis chamber, ensure that it normally works by power supply unit and control module.The signal of four sensor collections, after signal gathering unit and signal condition cell processing, is sent into monitoring host computer and is analyzed.During breaking out of fire, exhaust gas processing device connected by solenoid valve, and monitoring host computer controls acoustic-optic alarm and reports to the police.
Method of the present invention is applied to some equipment machine room, and step is as follows:
(1) first fire accident investigation is carried out to such confined space, find that main fire risk is the overheated or short circuits of nonmetallic materials such as cable, circuit board wherein.Therefore, simulated experiment can be carried out for this several situation.
(2) to PNN network model off-line training, as shown in Figure 3.Simulated fire field data is obtained by gas sensor and temperature sensor.Choose gas sensor response S
i, gas sensor rate of change Δ S
iwith temperature-responsive value T
ipre-service is carried out as individual features.Adopt array normalization can reach good effect, computing formula is:
Wherein, X
i,jrepresent jth eigenwert when measuring for i-th time,
value after representation transformation.By the data after normalization
carry out PCA principal component analysis (PCA).Front 3 principal component contributor rates reach 90%, can be used for the former data of matching.Utilize front 3 major components, i.e. characteristic parameters, form new sample set X={x
i,n| n=1,2,3}, and the training being applied to newly-built PNN network model.When exporting expectation and meeting the demands, stop training, the PNN network model of formation is input in monitoring host computer.
(3) real-time Fire in the present invention, as shown in Figure 2, gas sensor gathers the O at monitored scene through sampling pipe
2, CO, CO
2deng signal, after data prediction, input the PNN network model trained, prediction fire probability U
i.If fire probability is greater than given threshold values U
d1=0.4, then start to record pre-warning time t (i), otherwise then data are inputted PNN network model on-line training program.Carry out secondary judgement, if fire probability is more than or equal to given threshold values U
d2=0.6, then directly open acoustic-optic alarm and connect exhaust gas processing device; Otherwise, then judge whether pre-warning time t (i) is more than or equal to given threshold values t
d=15: be, then open acoustic-optic alarm, connect exhaust gas processing device, and data are inputted PNN network model on-line training program; No, then data are inputted PNN network model on-line training program.
Pre-warning time is defined as follows:
U () representation unit step signal, i ∈ [0, T).T=20 is cycle length, as i=T, resets pre-warning time.Get p=5, if namely represent continuous 4 U
iall be less than or equal to U
d1, then pre-warning time t (i) is reset.
(4) to PNN network model on-line training, as shown in Figure 4.Monitoring site data, after fire hazard monitoring master routine judges, send into online training program.To system whether be in stable state judge time, setting threshold values θ=0.01, when meet:
‖X(i+1)-X(i)‖＜θ
Think that system enters stable state, thus stop on-line training, maintain PNN network model constant.When system is not in stable state, then using the desired output of the output of fire hazard monitoring master routine as this sample, the training sample set adding PNN network model carries out on-line training, until test sample book desired output satisfies condition.The new PNN network model obtained re-writes monitoring host computer, carries out follow-up monitoring.
In a word, gas concentration change when the present invention utilizes Electronic Nose Technology process fire, time of fire alarming is shorter, and reduces wrong report, rate of failing to report; Adopt PNN network model, and introduce stable state judgement, realize monitoring and online updating in real time; Introduce pre-warning time judgment mechanism, more intelligent, reliable.The present invention can be applicable to the fire detecting and alarm of the confined spaces such as power distribution cabinet, machine room, goods and materials freight house, aircraft hold and space capsule.
Instructions of the present invention does not elaborate the known technology that part belongs to those skilled in the art.
The above; be only the embodiment in the present invention; but protection scope of the present invention is not limited thereto; any people being familiar with this technology is in the technical scope disclosed by the present invention; the conversion or replacement expected can be understood; all should be encompassed in and of the present inventionly comprise within scope, therefore, protection scope of the present invention should be as the criterion with the protection domain of claims.
Claims (4)
1. based on a confined space fire detection alarm system for Electronic Nose Technology, it is characterized in that comprising: sampling pipe, filtration unit, three gas sensors, temperature sensor, sensor power unit, sensor control unit, signal gathering unit, signal condition unit, monitoring host computer, acoustic-optic alarm, flowmeter, vacuum pump, three-way solenoid valve and exhaust gas processing devices;
Sensor power unit provides power supply, its output terminal connecting sensor for three gas sensors and temperature sensor; Sensor control unit controls unlatching, the closedown of three gas sensors and temperature sensor, its output terminal connecting sensor; Three sensors are respectively O
2sensor, CO sensor and CO
2sensor; Signal gathering unit is used for the data of collection three gas sensors and temperature sensor, and its input end is connected with sensor, and output terminal is connected with signal condition unit; Signal condition unit amplifies the signal collected, filtering, and its input end is connected with signal gathering unit, and output terminal is connected with monitoring host computer; Monitoring host computer is provided with fire hazard monitoring master routine and PNN Neural Network Online training algorithm, can Real time identification fire provide warning, online updating PNN neural network and storage data, and its output terminal connects acoustic-optic alarm and three-way solenoid valve;
Sampling pipe gathers, conveying confined space gas on-site data, and its one end is as in confined space, and the other end connects filtration unit; Water vapor in filtration devices gas and smoke particle, reduce the impact on sensor, and its one end connects sampling pipe, and the other end connects gas analysis chamber by the road; Gas analysis chamber is used for placement three gas sensors and temperature sensor, and its other end is connection traffic meter by the road; Flow velocity in flowmeter pilot piping, its other end connects vacuum pump by the road; Vacuum pump is bled, and for systematic sampling provides power, its other end is connecting tee solenoid valve by the road; Three-way solenoid valve controls the flow direction of tail gas, and its other end connects exhaust gas processing device by the road; Exhaust gas processing device can process CO, CO in sample gas
2harmful gas, preventing pollution;
Monitoring host computer implementation procedure: the O at monitored scene
2, CO, CO
2and temperature signal, the PNN neural network model trained is inputted after data prediction, prediction fire probability also carries out pre-warning time judgement, if judgement breaking out of fire, then open acoustic-optic alarm, connect exhaust gas processing device, and by PNN network model on-line training program that data input has trained; If judge non-breaking out of fire, then continue monitoring, and data are inputted PNN network model on-line training program; Before carrying out on-line training, carry out stable state judgement, if judge to be in stable state, then return fire hazard monitoring master routine; If not, then sample is added the training of PNN neural network model, obtain new PNN network model, and write fire hazard monitoring master routine;
The described PNN network model trained adopts off-line training mode, and off-line training process is: first investigate the on-the-spot disaster hidden-trouble of confined space, and carry out simulated fire experiment with this; Obtain simulated fire field data by gas sensor and temperature sensor, choose gas sensor response S
i, gas sensor rate of change △ S
iwith temperature-responsive value T
icarry out pre-service as individual features, data preprocessing method adopts array method for normalizing, and computing formula is:
Wherein, X
i,jrepresent jth eigenwert when measuring for i-th time, X
i'
, jvalue after representation transformation;
By the data X after normalization
i'
, jcarry out PCA principal component analysis (PCA), by dimensionality reduction, find N (N<j) individual orthogonal characteristic variable, i.e. major component, make it to reflect data X
i'
, jprincipal character, the scale of compression legacy data matrix; When top n principal component contributor rate reaches 90%, then think that this N number of major component can reflect X
i'
, jprincipal character, be used for the former data of matching; Utilize top n major component, i.e. characteristic parameter, form new sample set X={x
i,n| n=1,2 ..., N}, and the training being applied to newly-built PNN network model; When exporting expectation and meeting the demands, stop training, by the PNN network model write monitoring host computer formed.
2. the confined space fire detection alarm system based on Electronic Nose Technology according to claim 1, is characterized in that: described PNN network model on-line training program is embodied as: whether be in stable state to system and judge; Setting threshold values θ, when meeting:
||X(i+1)-X(i)||<θ
Think that system enters stable state, thus stop on-line training, maintain PNN network model constant; When system is not in stable state, then using the desired output of the output of fire hazard monitoring master routine as this sample, the training sample set adding PNN network model carries out on-line training, until test sample book desired output satisfies condition; The new PNN network model obtained re-writes monitoring host computer, carries out follow-up monitoring.
3. the confined space fire detection alarm system based on Electronic Nose Technology according to claim 1, is characterized in that: described pre-warning time t (i) is defined as follows:
U () representation unit step signal, and i ∈ [0, T), T is cycle length, as i=T, resets pre-warning time t (i), U
ifor fire probability.
4., based on a confined space fire detecting and alarm method for Electronic Nose Technology, it is characterized in that performing step is as follows:
(1) first fire accident investigation is carried out to confined space, find out its main fire risk, and carry out simulated experiment for this several principal risk;
(2) according to simulated experiment the data obtained, off-line training is carried out to PNN network model;
After data normalization, carry out PCA principal component analysis (PCA), and be applied to the training of newly-built PNN network model; When exporting expectation and meeting the demands, stop training, the PNN network model of formation is input in monitoring host computer;
(3) the PNN network model write monitoring host computer will trained;
Systematic sampling pipe is placed in confined space to be monitored, three-way solenoid valve connects air, turn on sensor power supply unit and control module, open vacuum pump to start to bleed, and start monitoring host computer, obtain fire probability from the data of monitoring site collection by PNN network model, carry out real-time fire identification in conjunction with pre-warning time judgment mechanism;
(4) the monitoring site real time data obtained is utilized to carry out on-line training to PNN network model;
First whether be in stable state to system to judge; When system is in stable state, stop on-line training, maintain PNN network model constant; When system is not in stable state, then training sample set data being added PNN network model carries out on-line training, obtains new PNN network model, and re-writes monitoring host computer, to carry out follow-up monitoring.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310026143.5A CN103116961B (en) | 2013-01-21 | 2013-01-21 | A kind of confined space fire detection alarm system based on Electronic Nose Technology and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310026143.5A CN103116961B (en) | 2013-01-21 | 2013-01-21 | A kind of confined space fire detection alarm system based on Electronic Nose Technology and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103116961A CN103116961A (en) | 2013-05-22 |
| CN103116961B true CN103116961B (en) | 2015-08-12 |
Family
ID=48415323
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310026143.5A Expired - Fee Related CN103116961B (en) | 2013-01-21 | 2013-01-21 | A kind of confined space fire detection alarm system based on Electronic Nose Technology and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103116961B (en) |
Families Citing this family (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103366495B (en) * | 2013-07-11 | 2015-08-05 | 合肥工业大学 | A kind of air suction type high sensitivity smoke particle detector and application thereof |
| CN103389230B (en) * | 2013-07-22 | 2016-08-10 | 中国人民解放军后勤工程学院 | A kind of Novel open space oil plant fire contamination sampling monitoring system |
| CN103646490B (en) * | 2013-12-20 | 2016-03-30 | 中国科学技术大学 | A kind of air suction type aircraft hold fire detecting arrangement based on complex probe technology |
| CN103996263B (en) * | 2014-05-11 | 2016-08-17 | 中国科学技术大学 | A kind of air suction type aircraft hold fire detector using aerosol gases to sense |
| CN105185022A (en) * | 2015-10-21 | 2015-12-23 | 国家电网公司 | Transformer substation fire detection system based on multi-sensor information combination and detection information combination method |
| CN107085079A (en) * | 2017-05-12 | 2017-08-22 | 盐城工学院 | A kind of Artificial Olfactory device and organic pollution control device |
| CN107331102A (en) * | 2017-08-30 | 2017-11-07 | 国网安徽省电力公司电力科学研究院 | Intelligent concentration suction-type cable fire pole early warning device and method |
| CN108648401A (en) * | 2018-07-18 | 2018-10-12 | 国家电网有限公司 | Comprehensive survey device for fire |
| CN109035714A (en) * | 2018-10-12 | 2018-12-18 | 南京同安智能科技有限公司 | A kind of Working in confined space safe preservation system |
| CN109615814A (en) * | 2018-12-03 | 2019-04-12 | 中国船舶重工集团公司第七〇九研究所 | Radiation resistance combined type fire detecting arrangement |
| CN111311869B (en) * | 2020-02-14 | 2021-04-27 | 清华大学合肥公共安全研究院 | Fire safety monitoring method and system based on area alarm model and cloud platform |
| CN111443173A (en) * | 2020-03-03 | 2020-07-24 | 南京农业大学 | Method for detecting adulterated illegal cooking oil in edible peanut oil based on electronic nose method |
| CN112036619B (en) * | 2020-08-17 | 2023-07-11 | 中国标准化研究院 | Method for judging whether roast duck exceeds goods shelf end point by combining electronic nose with Bayes algorithm |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6166647A (en) * | 2000-01-18 | 2000-12-26 | Jaesent Inc. | Fire detector |
| CN101241019A (en) * | 2008-02-28 | 2008-08-13 | 河北工业大学 | Robot bionic olfactory system |
| CN101769889A (en) * | 2010-01-06 | 2010-07-07 | 浙江清华长三角研究院 | Electronic nose system for detecting quality of agricultural products |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7034701B1 (en) * | 2000-06-16 | 2006-04-25 | The United States Of America As Represented By The Secretary Of The Navy | Identification of fire signatures for shipboard multi-criteria fire detection systems |
| US7171312B2 (en) * | 2002-07-19 | 2007-01-30 | Smiths Detection, Inc. | Chemical and biological agent sensor array detectors |
| KR20040080664A (en) * | 2003-03-12 | 2004-09-20 | 신화전자주식회사 | Fire-warning facilities with electronic nose |
-
2013
- 2013-01-21 CN CN201310026143.5A patent/CN103116961B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6166647A (en) * | 2000-01-18 | 2000-12-26 | Jaesent Inc. | Fire detector |
| CN101241019A (en) * | 2008-02-28 | 2008-08-13 | 河北工业大学 | Robot bionic olfactory system |
| CN101769889A (en) * | 2010-01-06 | 2010-07-07 | 浙江清华长三角研究院 | Electronic nose system for detecting quality of agricultural products |
Non-Patent Citations (6)
| Title |
|---|
| Classifying Smokes Using an Electronic Nose and Neural Networks;Bancha Charumpom etc.;《SlCE 2002》;20020807;2661-2665 * |
| Intelligent Electronic Nose Systems for Fire Detection Systems Based on Neural Networks;Toru Fujinaka etc.;《The Second International Conference on Advanced Engineering Computing and Applications in Sciences》;20081231;73-76 * |
| 人工神经网络对电子鼻性能的影响;秦树基等;《同济大学学报(自然科学版)》;20050630;第33卷(第6期);804-808 * |
| 基于人工神经网络的电子鼻对混合气体检测研究;耿志广等;《现代计算机》;20101231(第5期);45-48 * |
| 基于仿生嗅觉的电子鼻在煤矿火灾预报识别的应用研究;吕振等;《计算机测量与控制》;20121231;第20卷(第5期);1369-1371 * |
| 基于电子鼻技术的电气火灾预警***研究;方向生;《中国博士学位论文全文数据库》;20080728;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103116961A (en) | 2013-05-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103116961B (en) | A kind of confined space fire detection alarm system based on Electronic Nose Technology and method | |
| CN103325204B (en) | A kind of fire hazard judgment method of environment parameter perception | |
| Clotet et al. | A real-time anomaly-based IDS for cyber-attack detection at the industrial process level of critical infrastructures | |
| Dong et al. | The gas leak detection based on a wireless monitoring system | |
| CN206524538U (en) | A kind of high-tension switch cabinet of automatic failure alarm | |
| CN107807860B (en) | Power failure analysis method and system based on matrix decomposition | |
| CN104199441B (en) | Blast furnace multi-state fault separating method based on sparse contribution plot and system | |
| CN108803465B (en) | A kind of distributed plug and play fault monitoring method based on closed-loop data driving | |
| CN105975991B (en) | A kind of improved extreme learning machine fire kind recognition methods | |
| CN108415810A (en) | Hard disk state monitoring method and device | |
| CN103488800A (en) | SVM (Support Vector Machine)-based power consumption abnormality detection method | |
| CN202119467U (en) | Self-adaptive wavelet neural network categorizing system of anomaly detection and fault diagnosis | |
| CN105403705A (en) | Continuous blood glucose monitoring equipment comprising blood glucose classification function fault detection module | |
| CN103529337B (en) | The recognition methods of nonlinear correlation relation between equipment failure and electric quantity information | |
| CN114785573A (en) | Intelligent substation process layer network abnormal flow detection method based on deep learning | |
| CN108875276B (en) | A kind of closed-loop system stability monitoring method of data-driven | |
| CN107607342A (en) | The healthy efficiency detection method of Air Conditioning Facilities device cluster | |
| CN106970301A (en) | A kind of detecting system based on smart home power supply circuit | |
| CN106195247A (en) | A kind of control system of speed variator based on big Db Management Model | |
| CN205334181U (en) | Intelligent water conservancy on -line monitoring monitored control system towards secure transmission net | |
| CN106292610A (en) | Intelligent home control system | |
| Yang et al. | Multi-view broad learning system for electricity theft detection | |
| CN106379323A (en) | Vehicle control system | |
| Hong et al. | A novel CNN-TCN-TAM classification model based method for fault diagnosis of chiller sensors | |
| CN205451492U (en) | Intelligent transportation system based on three -dimensional scene of computer |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20150812 Termination date: 20210121 |