CN105067768A - Multi-component mixed gas quantitative recognition system for dangerous chemical detection - Google Patents
Multi-component mixed gas quantitative recognition system for dangerous chemical detection Download PDFInfo
- Publication number
- CN105067768A CN105067768A CN201510457724.3A CN201510457724A CN105067768A CN 105067768 A CN105067768 A CN 105067768A CN 201510457724 A CN201510457724 A CN 201510457724A CN 105067768 A CN105067768 A CN 105067768A
- Authority
- CN
- China
- Prior art keywords
- layer
- data
- gas
- mixed gas
- 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.)
- Pending
Links
Landscapes
- Investigating Or Analyzing Materials By The Use Of Electric Means (AREA)
- Investigating Or Analyzing Materials By The Use Of Fluid Adsorption Or Reactions (AREA)
Abstract
The invention discloses a multi-component mixed gas quantitative recognition system for dangerous chemical detection, and relates to the technical field of detection and examination. The system is characterized by comprising a data layer, a characteristic layer, a classification layer, and a decision making layer; wherein the data layer, the characteristic layer, the classification layer, and the decision making layer are connected in sequence. The system utilizes the cross sensitivity of gas sensors to generate a high-dimensional corresponding mode to mixed gas; and then processes the obtained data by combining a corresponding mode recognition method so as to recognize the mixed gas.
Description
Technical field
The present invention relates to inspection technical field, be specifically related to a kind of quantitative judge system of the Diversity gas for harmful influence detection.
Background technology
Along with the continuous expansion of industrial and agricultural production scale; the demand of global threat chemicals (abbreviation harmful influence) constantly increases; the disaster accident that the links such as the production of harmful influence, transport, storage and use occur also constantly increases; the generation of harmful influence accident can cause huge disaster usually, and this is determined by following characteristics:
The first, harmful influence accident has in all contingent feature of links, and special analysis was done by relevant department, harmful influence is managed from producing to, transport, store, use, 6 discarded links none can escape by luck from accident.The accident of transit link is maximum, has accounted for 41% of total number of accident.
The second, harmful influence accident has volatile feature under hot and humid condition.
3rd, harmful influence accident has the feature of linksystem, and the harmful influence overwhelming majority is inflammable and explosive, and harmful influence accident often causes the consequence of series connection mutually between damaged in collision, leakage, burning, blast, pollution.
4th, harmful influence accident has rescue, gives treatment to the feature that difficulty is large and consequence is serious, and harmful influence generally all has special physics, chemical property, and the chemicals performance varied adds the difficulty of rescue, treatment work.Therefore, harmful influence is rescued is usually fire extinguishing, rescue the multiple means such as people, sterilization, evacuating personnel, detection carries out simultaneously.
According to statistics, in the production scale of harmful influence, Chinese Enterprises can produce 45000 kinds of chemicals in 2007, wherein had the kind that 3823 kinds belong to government and strictly supervise, and also had 335 kinds to be put into severe poisonous chemicals catalogue and implemented management.Along with China's rapid development of economy, the safety problem of harmful influence and the research of regulation technique also more urgent.
In international and domestic bibliographical information up to now, not yet find around hazardous chemical safety problem based on the relevant signals process of sensor network and the research of information fusion technology aspect.
Summary of the invention
The object of this invention is to provide a kind of quantitative judge system of the Diversity gas for harmful influence detection; it utilizes the cross-sensitivity of gas sensor to produce the response modes of higher-dimension to mixed gas; and in conjunction with corresponding mode identification method, the data obtained are processed, thus realize the identification to mixed gas.
In order to solve the problem existing for background technology, the present invention is by the following technical solutions: it comprises data Layer, characteristic layer, classification layer, decision-making level, and data Layer, characteristic layer, classification layer, decision-making level are connected successively.
The present invention comprises following job step:
In pretreatment process, metering circuit is converted to various gas sensor signal the voltage signal being convenient to acquisition and processing, the raw voltage signals of collection is converted to the gas response signal that physical significance is clearer and more definite;
At data preprocessing phase, adopt the technology such as normalization, denoising and signal compensation to a certain degree can remove noise and the baseline wander of compensation sensor;
In signal compensation, adopt mark ratioing technigue X=(S
gas-s
o)/S
o, have effect to the drift of additivity and product noise;
Feature extraction comprises two parts task, one is compress information, two is reduce the relevance between data, principal component analysis (PCA) (PCA) is a kind of typical feature extracting method, feature little for variance contribution ratio is removed by it, reach dimensionality reduction object, and eliminate the correlativity between sample data, make the neural computing of next stage more simple efficient; In addition Independent Component Analysis can be adopted to be used for feature extraction, and its correlativity can removed between data can remove again the higher order dependencies between data;
Pattern-recognition layer adopts reverse transfers neural network (BP) to realize the estimation of gas composition and concentration value, BP neural network full name is error backward propagation method, it is a kind of nonparametric learning method, zmodem, Nonlinear Processing ability is strong, is suitable for Gas Sensor Array Signals process;
The output of neural network and warning threshold value at different levels contrast and realize alarm level and adjudicate by last Situation Assessment.
Accompanying drawing illustrates:
Fig. 1 is structural representation of the present invention.
Embodiment
With reference to Fig. 1, for the quantitative judge system of the Diversity gas that harmful influence detects, it comprises data Layer, characteristic layer, classification layer, decision-making level, and data Layer, characteristic layer, classification layer, decision-making level are connected successively, and comprise following job step:
In pretreatment process, metering circuit is converted to various gas sensor signal the voltage signal being convenient to acquisition and processing, the raw voltage signals of collection is converted to the gas response signal that physical significance is clearer and more definite;
At data preprocessing phase, adopt the technology such as normalization, denoising and signal compensation to a certain degree can remove noise and the baseline wander of compensation sensor;
In signal compensation, adopt mark ratioing technigue X=(S
gas-s
o)/S
o, have effect to the drift of additivity and product noise;
Feature extraction comprises two parts task, one is compress information, two is reduce the relevance between data, principal component analysis (PCA) (PCA) is a kind of typical feature extracting method, feature little for variance contribution ratio is removed by it, reach dimensionality reduction object, and eliminate the correlativity between sample data, make the neural computing of next stage more simple efficient; In addition Independent Component Analysis can be adopted to be used for feature extraction, and its correlativity can removed between data can remove again the higher order dependencies between data;
Pattern-recognition layer adopts reverse transfers neural network (BP) to realize the estimation of gas composition and concentration value, BP neural network full name is error backward propagation method, it is a kind of nonparametric learning method, zmodem, Nonlinear Processing ability is strong, is suitable for Gas Sensor Array Signals process;
The output of neural network and warning threshold value at different levels contrast and realize alarm level and adjudicate by last Situation Assessment.
Claims (2)
1., for the quantitative judge system of the Diversity gas of harmful influence detection, it is characterized in that comprising data Layer, characteristic layer, classification layer, decision-making level, data Layer, characteristic layer, classification layer, decision-making level are connected successively.
2. the quantitative judge system of the Diversity gas for harmful influence detection according to claim 1, is characterized in that comprising following job step:
In pretreatment process, metering circuit is converted to various gas sensor signal the voltage signal being convenient to acquisition and processing, the raw voltage signals of collection is converted to the gas response signal that physical significance is clearer and more definite;
At data preprocessing phase, adopt the technology such as normalization, denoising and signal compensation to a certain degree can remove noise and the baseline wander of compensation sensor;
In signal compensation, adopt mark ratioing technigue X=(S
gas-s
o)/S
o, have effect to the drift of additivity and product noise;
Feature extraction comprises two parts task, one is compress information, two is reduce the relevance between data, principal component analysis (PCA) (PCA) is a kind of typical feature extracting method, feature little for variance contribution ratio is removed by it, reach dimensionality reduction object, and eliminate the correlativity between sample data, make the neural computing of next stage more simple efficient; In addition Independent Component Analysis can be adopted to be used for feature extraction, and its correlativity can removed between data can remove again the higher order dependencies between data;
Pattern-recognition layer adopts reverse transfers neural network (BP) to realize the estimation of gas composition and concentration value, BP neural network full name is error backward propagation method, it is a kind of nonparametric learning method, zmodem, Nonlinear Processing ability is strong, is suitable for Gas Sensor Array Signals process;
The output of neural network and warning threshold value at different levels contrast and realize alarm level and adjudicate by last Situation Assessment.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510457724.3A CN105067768A (en) | 2015-07-30 | 2015-07-30 | Multi-component mixed gas quantitative recognition system for dangerous chemical detection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510457724.3A CN105067768A (en) | 2015-07-30 | 2015-07-30 | Multi-component mixed gas quantitative recognition system for dangerous chemical detection |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105067768A true CN105067768A (en) | 2015-11-18 |
Family
ID=54497189
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510457724.3A Pending CN105067768A (en) | 2015-07-30 | 2015-07-30 | Multi-component mixed gas quantitative recognition system for dangerous chemical detection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105067768A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107132315A (en) * | 2017-05-12 | 2017-09-05 | 盐城工学院 | Signal recognition method, device and volatile organic matter detection device |
| CN109116042A (en) * | 2018-09-04 | 2019-01-01 | 中国农业大学 | A kind of network communication electronic nose detection system and detection method |
| CN114399883A (en) * | 2021-11-30 | 2022-04-26 | 宏大爆破工程集团有限责任公司 | Blasting equipment transportation monitoring and early warning system and method |
-
2015
- 2015-07-30 CN CN201510457724.3A patent/CN105067768A/en active Pending
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107132315A (en) * | 2017-05-12 | 2017-09-05 | 盐城工学院 | Signal recognition method, device and volatile organic matter detection device |
| CN109116042A (en) * | 2018-09-04 | 2019-01-01 | 中国农业大学 | A kind of network communication electronic nose detection system and detection method |
| CN114399883A (en) * | 2021-11-30 | 2022-04-26 | 宏大爆破工程集团有限责任公司 | Blasting equipment transportation monitoring and early warning system and method |
| CN114399883B (en) * | 2021-11-30 | 2023-09-26 | 宏大爆破工程集团有限责任公司 | Blasting equipment transportation monitoring and early warning system and method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| McKnight et al. | microDecon: A highly accurate read‐subtraction tool for the post‐sequencing removal of contamination in metabarcoding studies | |
| Arad et al. | A dynamic thresholds scheme for contaminant event detection in water distribution systems | |
| CN106888205B (en) | Non-invasive PLC anomaly detection method based on power consumption analysis | |
| Staples et al. | A risk assessment of selected phthalate esters in North American and Western European surface waters | |
| Zhao et al. | How sulfate-rich mine drainage affected aquatic ecosystem degradation in northeastern China, and potential ecological risk | |
| CN104793604B (en) | Principal component tracking based industrial fault monitoring method and application thereof | |
| Nallathambi et al. | Prediction of influencing atmospheric conditions for explosion Avoidance in fireworks manufacturing Industry-A network approach | |
| Liu et al. | A time–scaling property of air pollution indices: a case study of Shanghai, China | |
| CN103412557A (en) | Industrial fault detection and diagnostic method suitable for nonlinear process on-line monitoring | |
| CN105067768A (en) | Multi-component mixed gas quantitative recognition system for dangerous chemical detection | |
| Li et al. | Dynamic intelligent risk assessment of hazardous chemical warehouse fire based on electrostatic discharge method and improved support vector machine | |
| CN106209829A (en) | A kind of network security management system based on warning strategies | |
| Liu et al. | Why conventional detection methods fail in identifying the existence of contamination events | |
| CN105806400A (en) | Intelligent method and system for monitoring hydrant's safety state | |
| CN112861770A (en) | Water pollution early warning method and device | |
| Liu et al. | Performance evaluation for three pollution detection methods using data from a real contamination accident | |
| Chang et al. | Investigating the long-term change of injury pattern on severity, accident types and sources of injury in Taiwan’s manufacturing sector between 1996 and 2012 | |
| Chelani | Nonlinear dynamical analysis of ground level ozone concentrations at different temporal scales | |
| CN117975372A (en) | Construction site safety detection system and method based on YOLOv and transducer encoder | |
| CN103150484A (en) | Improved fault tree analysis method for safety evaluation | |
| CN103995985A (en) | Fault detection method based on Daubechies wavelet transform and elastic network | |
| CN111080139A (en) | Hypergraph-based public security event scenario representation method | |
| CN104966019B (en) | A kind of heuristic document threat detection method and system | |
| Liu et al. | Time series prediction of the chemical components of PM2. 5 based on a deep learning model | |
| CN105389475B (en) | A kind of electric power factory equipment fault detection method based on WAVELET PACKET DECOMPOSITION |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20151118 |