CN110046834A - Workplace pernicious gas health risk quantitative evaluating system and its appraisal procedure - Google Patents
Workplace pernicious gas health risk quantitative evaluating system and its appraisal procedure Download PDFInfo
- Publication number
- CN110046834A CN110046834A CN201910394208.9A CN201910394208A CN110046834A CN 110046834 A CN110046834 A CN 110046834A CN 201910394208 A CN201910394208 A CN 201910394208A CN 110046834 A CN110046834 A CN 110046834A
- Authority
- CN
- China
- Prior art keywords
- sensor
- information
- workplace
- communication module
- wireless communication
- 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
Links
- 238000000034 method Methods 0.000 title claims abstract description 42
- 230000036541 health Effects 0.000 title claims abstract description 28
- 230000000505 pernicious effect Effects 0.000 title claims abstract description 28
- 238000012544 monitoring process Methods 0.000 claims abstract description 63
- 230000006854 communication Effects 0.000 claims description 34
- 238000004891 communication Methods 0.000 claims description 33
- 238000013528 artificial neural network Methods 0.000 claims description 31
- VLKZOEOYAKHREP-UHFFFAOYSA-N n-Hexane Chemical compound CCCCCC VLKZOEOYAKHREP-UHFFFAOYSA-N 0.000 claims description 30
- 238000005070 sampling Methods 0.000 claims description 24
- 238000012360 testing method Methods 0.000 claims description 18
- 238000011478 gradient descent method Methods 0.000 claims description 17
- 230000005540 biological transmission Effects 0.000 claims description 12
- 238000012545 processing Methods 0.000 claims description 11
- RAHZWNYVWXNFOC-UHFFFAOYSA-N Sulphur dioxide Chemical compound O=S=O RAHZWNYVWXNFOC-UHFFFAOYSA-N 0.000 claims description 6
- 230000009471 action Effects 0.000 claims description 2
- 239000007789 gas Substances 0.000 description 75
- 238000004422 calculation algorithm Methods 0.000 description 23
- 239000000523 sample Substances 0.000 description 19
- 238000012549 training Methods 0.000 description 19
- LFQSCWFLJHTTHZ-UHFFFAOYSA-N Ethanol Chemical compound CCO LFQSCWFLJHTTHZ-UHFFFAOYSA-N 0.000 description 18
- 230000006870 function Effects 0.000 description 18
- 238000010586 diagram Methods 0.000 description 13
- 230000008569 process Effects 0.000 description 13
- 208000005374 Poisoning Diseases 0.000 description 10
- 238000003062 neural network model Methods 0.000 description 10
- 231100000572 poisoning Toxicity 0.000 description 10
- 230000000607 poisoning effect Effects 0.000 description 10
- 230000008717 functional decline Effects 0.000 description 8
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 7
- 208000028571 Occupational disease Diseases 0.000 description 7
- QGJOPFRUJISHPQ-UHFFFAOYSA-N Carbon disulfide Chemical compound S=C=S QGJOPFRUJISHPQ-UHFFFAOYSA-N 0.000 description 6
- 230000000694 effects Effects 0.000 description 5
- 230000000052 comparative effect Effects 0.000 description 4
- 238000013461 design Methods 0.000 description 4
- 238000004519 manufacturing process Methods 0.000 description 4
- 238000005457 optimization Methods 0.000 description 4
- 239000000126 substance Substances 0.000 description 4
- 241001269238 Data Species 0.000 description 3
- 230000004913 activation Effects 0.000 description 3
- 238000013480 data collection Methods 0.000 description 3
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 3
- 230000007613 environmental effect Effects 0.000 description 3
- 238000011156 evaluation Methods 0.000 description 3
- 238000004880 explosion Methods 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 240000007594 Oryza sativa Species 0.000 description 2
- 235000007164 Oryza sativa Nutrition 0.000 description 2
- 125000003158 alcohol group Chemical group 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000009792 diffusion process Methods 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 238000002407 reforming Methods 0.000 description 2
- 235000009566 rice Nutrition 0.000 description 2
- 238000012502 risk assessment Methods 0.000 description 2
- 238000004088 simulation Methods 0.000 description 2
- 244000283207 Indigofera tinctoria Species 0.000 description 1
- 230000001154 acute effect Effects 0.000 description 1
- 239000000853 adhesive Substances 0.000 description 1
- 230000001070 adhesive effect Effects 0.000 description 1
- 150000001335 aliphatic alkanes Chemical class 0.000 description 1
- 230000003321 amplification Effects 0.000 description 1
- 230000004888 barrier function Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000001684 chronic effect Effects 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000004868 gas analysis Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 230000035987 intoxication Effects 0.000 description 1
- 231100000566 intoxication Toxicity 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000005036 nerve Anatomy 0.000 description 1
- 210000004218 nerve net Anatomy 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000001473 noxious effect Effects 0.000 description 1
- 238000003199 nucleic acid amplification method Methods 0.000 description 1
- 239000003960 organic solvent Substances 0.000 description 1
- 238000011158 quantitative evaluation Methods 0.000 description 1
- 239000002904 solvent Substances 0.000 description 1
- 208000024891 symptom Diseases 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
- G01D21/00—Measuring or testing not otherwise provided for
- G01D21/02—Measuring two or more variables by means not covered by a single other subclass
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
- G01N33/0004—Gaseous mixtures, e.g. polluted air
- G01N33/0009—General constructional details of gas analysers, e.g. portable test equipment
- G01N33/0027—General constructional details of gas analysers, e.g. portable test equipment concerning the detector
- G01N33/0031—General constructional details of gas analysers, e.g. portable test equipment concerning the detector comprising two or more sensors, e.g. a sensor array
- G01N33/0034—General constructional details of gas analysers, e.g. portable test equipment concerning the detector comprising two or more sensors, e.g. a sensor array comprising neural networks or related mathematical techniques
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
- G01N33/0004—Gaseous mixtures, e.g. polluted air
- G01N33/0009—General constructional details of gas analysers, e.g. portable test equipment
- G01N33/0062—General constructional details of gas analysers, e.g. portable test equipment concerning the measuring method or the display, e.g. intermittent measurement or digital display
- G01N33/0068—General constructional details of gas analysers, e.g. portable test equipment concerning the measuring method or the display, e.g. intermittent measurement or digital display using a computer specifically programmed
-
- 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
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Chemical & Material Sciences (AREA)
- Human Resources & Organizations (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Theoretical Computer Science (AREA)
- Pathology (AREA)
- Medicinal Chemistry (AREA)
- Development Economics (AREA)
- Immunology (AREA)
- Educational Administration (AREA)
- Biochemistry (AREA)
- Strategic Management (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Mathematical Physics (AREA)
- Analytical Chemistry (AREA)
- Combustion & Propulsion (AREA)
- Entrepreneurship & Innovation (AREA)
- Economics (AREA)
- Food Science & Technology (AREA)
- Biomedical Technology (AREA)
- Computational Linguistics (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Operations Research (AREA)
- Tourism & Hospitality (AREA)
- Game Theory and Decision Science (AREA)
- Biophysics (AREA)
- Quality & Reliability (AREA)
- Data Mining & Analysis (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Arrangements For Transmission Of Measured Signals (AREA)
Abstract
The invention proposes a kind of workplace pernicious gas health risk quantitative evaluating system and its appraisal procedures, the system includes server, at least one workplace, each workplace is configured with, one host supervision system, several station monitoring subsystems, several stations monitoring subsystem respectively it is matched connect the station area first kind intelligent terminal and by received information be transmitted to by way of bluetooth its matching connection station monitoring subsystem, the station monitoring subsystem connects the host supervision system, received information is transmitted to host supervision system, host supervision system connects server, it is interacted with the server info.It is based on the monitoring system in this way, the data such as concentration and the operating time of pernicious gas are contacted according to staff in each workplace, that is, workshop, the hazard index of each staff is calculated, monitoring information then can be pushed to first kind intelligent terminal, the second class intelligent terminal.
Description
Technical field
The present invention relates to gas monitoring systems, are specifically related to field harmful gas monitoring system and its monitoring and evaluation side
Method.
Background technique
Industrial level steadily improves since with reform and opening-up, and domestic all trades and professions all achieve development at full speed,
But meanwhile the occupational hazards problem of workplace also results in the common concern of common people.Many chemical noxious substances are as good
Solvent or adhesive, be widely used in a variety of industries such as electronics, if such chemical substance is evaporate into air, for a long time
Sucking can cause occupational poisoning, seriously endanger the life and health of labourer and seriously affect the safety in production of enterprise.It is logical
It crosses and occupational disease morbidity data between 2001 to 2014 years of the announcement of national health State Family Planning Commission is counted, occupational disease morbidity number of cases is in
Wave ascendant trend reached peak 29972 in 2014, and wherein occupational poisoning disease incidence is minimum accounts for 4.27%, and highest accounts for
16.9%.By looking back 2006 to 2015 Jiangsu Province Nian Jian occupational poisoning cases, occupational poisoning case is 1527 total,
Middle acute occupational poisoning 547, accounting 35.82%;Chronic Occupational poisoning 980, accounting 64.18%, from slow poisoning gas
Analysis of Main Sources, leading to the predominant gas of cooupational intoxication is the carbon disulfide and n-hexane that volatilize in organic solvent.And
And since present profession or occupation health statistics coverage rate in China's is not high, true occupational poisoning number may be much higher than statistical data,
Actual conditions allow of no optimist.
Labourer is exposed to for a long time in the pernicious gas environment of low concentration, different degrees of symptom can be all occurred, be led to duty
Industry disease disease incidence can not have always been high any more, seriously threaten the healthy and safe of labourer.It can be seen that in workplace to production scene
Pernicious gas and environment carry out real-time monitoring, make occupational disease health risk assessment for each labourer just and seem especially heavy
It wants.By qualitative or quantitative assessment technology, carrying out occupational disease health risk assessment to the pernicious gas of workplace is labourer
Prevent effective behaves of occupational disease hazards, be that company managers' monitoring personnel health safety, monitoring shop safety are raw
The important measures of production.
Have at present for the major accidents such as chemical leakage monitoring system all, be not to ask for occupational disease health risk
Topic is monitored;Although many mechanisms assess occupational health risk, its assessment is stayed in substantially to theoretical model
Research and experiment, assessed with quantitative and semi-quantitative based on, do not build a set of monitoring system and quantitatively commented in real time to realize
Estimate.
Summary of the invention
Based on the above issues, it is an object of the present invention to propose that a kind of workplace pernicious gas health risk is quantitatively commented
Estimate system and its appraisal procedure, which causes the gas of occupational disease slow poisoning present in monitoring place
Body.The monitoring system assesses occupational health risk using quantitative evaluation method (i.e. appraisal procedure).
To achieve the above object, the present invention adopts the following technical scheme:
Workplace pernicious gas health risk quantitative evaluating system, which is characterized in that include server, at least one work
Make place, each workplace is configured with, a host supervision system, several station monitoring subsystems, several institutes
It states station monitoring subsystem to be configured in the default station area of the workplace, distinguishes the matched connection station area
Received first kind intelligent terminal information is transmitted to connected to it by interior first kind intelligent terminal, the station monitoring subsystem
Host supervision system, the host supervision system connects server, and interacts with the server info.
Preferably, which includes processor, power supply unit, the first wireless communication module, the second wireless communication
Module, interface circuit, first sensor, second sensor, 3rd sensor, display module, wherein power supply unit is electrically connected
Processor to provide driving electric energy, the first wireless communication module, the second wireless communication module be electrically connected processor its
Information transmission is carried out based on instruction, interface circuit is electrically connected to first sensor, second sensor, 3rd sensor, to
The sample information of its feedback is received respectively and is transmitted to the processor of electric connection, and display module is electrically connected interface circuit, uses
The information transmitted with video-stream processor.
Preferably, which is Temperature Humidity Sensor, to the temperature and humidity information of sampling location, second sensor
For gas sensor, to the harmful gas concentration information of sampling location,
3rd sensor is air velocity transducer, and to the wind speed information of sampling location, the first wireless communication module is 4G mould
Block or IOT module are electrically connected server, and the second wireless communication module is WiFi module or zigbee module, connect work
Position monitoring subsystem.
Preferably, which is gas sensor, the n-hexane information of sampling location or sampling location just oneself
The information of alkane and sulfur dioxide.
Preferably, the station monitoring subsystem include first processor, the first power supply unit, third wireless communication module,
4th wireless communication module, first interface circuit, the 4th sensor, the 5th sensor, wherein the first power supply unit is electrically connected
First processor is electrically connected first to provide driving electric energy, third wireless communication module, the 4th wireless communication module
It carries out information transmission based on instruction with processor, and first interface circuit is respectively and electrically connected to the 4th sensor, the 5th sensor
To receive the sample information of its feedback and transmit it to the first processor of electric connection.
Preferably, the 4th sensor is temperature/humidity sensor, temperature/humidity information to sampling operation place;It is described
5th sensor is air velocity transducer;
Third wireless communication module is bluetooth module, the first kind intelligent terminal in matching connection station area, and will
The first kind intelligent terminal information of connection feeds back to main monitoring system.
Preferably, the danger of staff (is assessed) in information of the server based on the main monitoring system transmission, estimation
Evil index, and the information of estimation (assessing) is pushed into preset first kind intelligent terminal, the second class intelligent terminal.
The embodiment of the invention also provides workplace pernicious gas health risk quantitative evaluating methods, and it includes above-mentioned
System, which is characterized in that the method comprises the following steps:
S11. the location information of workplace arbitrary point and the sample information of sensor are obtained;
S12. data processing model is constructed based on BP neural network;
S13. the gas concentration at any given position in workplace is calculated.
Preferably, in the S12, the mode input node layer has 6, is two-dimensional coordinate position data, 4 gases respectively
The concentration Value Data of sensor sample, output node layer have 1, are the sensor concentration estimated value at two-dimensional coordinate position;BP
The hidden layer of neural network is designed as two layers, and the node number of each hidden layer is 30.
The embodiment of the present invention also provides workplace pernicious gas health risk quantitative evaluating method, and it includes above-mentioned to be
System, which is characterized in that the method comprises the following steps:
S21. the data processing model based on BP neural network building;
S22. the gas concentration in prediction scene at gas sensor;
S23. the location information of position to be predicted is obtained;
S24. predictive information is generated, that is, predicts the harm quotient in staff's future at the position,
Wherein, test data is predicted using the model that self-adaption gradient descent method is practiced, the input layer of the model
Node has 10, is the information of wind speed, temperature and 8 preset fixed positions respectively;Model output node layer has 4, is 4
The concentration value of the gas sensor of preset fixed position;The hidden layer of BP neural network is designed as 2 layers, the section of each hidden layer
Point number is all 25.
Scheme in compared with the existing technology, advantages of the present invention:
The workplace pernicious gas health risk quantitative evaluating system and its appraisal procedure that the application proposes, the monitoring system
System, which is realized, carries out the assessment of occupational health quantifying risk, accurate statistical staff (yard at the scene to field personnel
In) contact duration with pernicious gas, calculate the function of the exposure concentration of staff and pernicious gas.The system is based on BP
The concentration value of the pernicious gas of Neural Network model predictive scene any position predicts the harm quotient in staff's future.
Detailed description of the invention
The invention will be further described with reference to the accompanying drawings and embodiments:
Fig. 1 is the module diagram of the workplace pernicious gas health risk quantitative evaluating system of the embodiment of the present invention;
Fig. 2 a is the hardware block diagram of main monitoring system in Fig. 1;
Fig. 2 b is the hardware block diagram of station monitoring subsystem in Fig. 1;
Fig. 3 a is the layout and planning schematic diagram of the workplace of the embodiment of the present invention;
Fig. 3 b is the Bluetooth base. station configuration schematic diagram of the embodiment of the present invention;
Fig. 3 c is the flow diagram of the client location of the embodiment of the present invention;
Fig. 4 is the Multi-layer BP Neural Network topological structure of one embodiment of the invention;
Fig. 5 one embodiment of the invention calculates the process signal of workplace harmful gas concentration based on BP neural network
Figure;
Fig. 6 is that schematic diagram of the Sigmoid function as activation primitive is chosen in Fig. 4;
Fig. 7 is loss function decline curve schematic diagram;
Fig. 8 a is the true value and estimated value comparative result figure of the test sample of one embodiment of the invention;
Fig. 8 b is the true value and estimated value comparative result figure that the test sample of preceding 200 points is taken in 8a;
Fig. 9 is the true value and estimated value comparative result figure of preceding 100 test samples of one embodiment of the invention;
Figure 10 is that one embodiment of the invention is shown based on BP neural network come the process of prediction work place harmful gas concentration
It is intended to;
Figure 11 a is batch gradient descent method loss function decline curve figure;Figure 11 b is batch gradient descent method loss function
Decline curve enlarged drawing;Figure 11 c is self-adaption gradient descent method loss function decline curve figure, and Figure 11 d is under self-adaption gradient
Drop method loss function decline curve enlarged drawing;
Figure 12 is the true value and predicted value comparison diagram of test sample;
Figure 13 is the true value and predicted value comparison diagram of test sample;
Figure 14 is the test point distribution map that the concentration of alcohol gas at the scene is emulated using Fluent.
Specific embodiment
Above scheme is described further below in conjunction with specific embodiment.It should be understood that these embodiments are for illustrating
The present invention and be not limited to limit the scope of the invention.Implementation condition used in the examples can be done such as the condition of specific producer into
One successive step, the implementation condition being not specified are usually the condition in routine experiment.
The module for the workplace pernicious gas health risk quantitative evaluating system of the embodiment of the present application is shown as shown in Figure 1
It is intended to, concentration of the system to the pernicious gas in monitoring place (workshop), it includes server, at least one yards
Institute, each workplace are configured with, a host supervision system, several station monitoring subsystems, several stations
Monitoring subsystem distinguishes the matched first kind intelligent terminal for connecting the station area and by received information by way of bluetooth
It is transmitted to the station monitoring subsystem of its matching connection, which connects the host supervision system, will be received
Information is transmitted to host supervision system, and host supervision system connects server, interacts with the server info.It is based on the prison in this way
Examining system contacts the data such as concentration and the operating time of pernicious gas according to staff in each workplace (workshop),
The hazard index of each staff is calculated, monitoring information then can be pushed to first kind intelligent terminal, the second class intelligence is eventually
It holds (being pushed to staff's mobile phone terminal, administrative staff's mobile phone terminal).In present embodiment, illustrate a workplace,
In other implementations, there can be multiple workplaces, workplace each in this way is configured with, a host supervision system,
Several station monitoring subsystems, several stations monitoring subsystem distinguish the matched first kind intelligence for connecting the station area
Terminal and the station monitoring subsystem that received information is transmitted to its matching connection by way of bluetooth, station monitoring
System connects the host supervision system, and received information is transmitted to host supervision system, and host supervision system connects server,
It is interacted with the server info.In one embodiment, the activity of the staff at certain station at the scene is detected (to produce
Journey) in unsoundness risk when, corresponding first kind intelligent terminal provides warning note.In one embodiment, host supervision
System is connected with display module (e.g., display), and the information monitored in this way is shown in real time on the display module of its connection.?
In one embodiment, server is connected with display interface, and such administrative staff remotely (e.g., can pass through remote web page WEB in real time
End) check monitoring information.The system contacts the concentration and work of pernicious gas (e.g., n-hexane) to monitoring field staff
Make the data such as duration, calculate the hazard index of each staff, monitoring information is then pushed to staff's mobile phone terminal
(first kind intelligent terminal) and administrative staff's mobile phone terminal (the second class intelligent terminal), the information can also pass through webpage WEB terminal shape
Formula is shown, when staff's unsoundness risk in process of production, first/second class intelligent terminal provides warning note.This
The field harmful gas that sample application embodiment proposes monitors system, to the work in the workplace containing pernicious gas
Personnel carry out the assessment of occupational health quantifying risk.
Next Fig. 2 a is combined, 2b describes host supervision system, the station monitoring subsystem of the application embodiment,
The hardware block diagram of monitoring system, the main monitoring system include based on as shown in Figure 2 a,
Processor, power supply unit, the first wireless communication module, the second wireless communication module, interface circuit, the first sensing
Device, second sensor, 3rd sensor, display module, wherein power supply unit electrically connected processing device is to provide driving electricity
Can, processor is electrically connected in the first wireless communication module/second wireless communication module, and it carries out information transmission based on instruction,
Interface circuit is electrically connected to first sensor, second sensor, 3rd sensor to receive the sample information of its feedback and incite somebody to action
Its processor for being transmitted to electric connection, display module are electrically connected interface circuit, the information to video-stream processor transmission.This
Implement in which, first sensor is Temperature Humidity Sensor, the temperature and humidity information to sampling operation place.Second sensor
Gas (preferably, sampling n-hexane, or sampling n-hexane and sulfur dioxide) for gas sensor, to sampling operation place
Information, 3rd sensor are wind speed information of the air velocity transducer to sampling operation place, and the first wireless communication module is 4G mould
Block, IOT module, to connect server, the second wireless communication module is WiFi module or zigbee module, to connect station
Monitoring subsystem.Power circuit and single-chip minimum system.Interface circuit uses RS-485 circuit.Host supervision system in this way
System is communicated by RS-485 bus with gas sensor, Temperature Humidity Sensor, air velocity transducer, is acquired each gas in workshop and is passed
Pernicious gas (preferably, be n-hexane) concentration data at sensor and temperature, humidity, wind speed these environmental datas, furthermore
The data for summarizing each station monitoring subsystem in workshop by WiFi module, will own finally by 4G module or IOT module
Data are sent to the server of specified IP address.In present embodiment, processor uses STMF10X Series MCU processor.In this way
The gas concentration value acquired in connected workshop at each n-hexane sensor probe, Yi Jiwen are realized by host supervision system
The environmental parameters such as degree, humidity, wind speed.Several station monitoring subsystems acquisition in workshop is received by network (such as WiFi module)
Data and summarize.Give monitoring data hair (being based on 4G module) to monitoring central server.It (e.g., is shown by display interface
Screen), show the concentration value of the gas (e.g., n-hexane sensor probe collects) of sampling in real time in workshop.
It is as shown in Figure 2 b the hardware block diagram of station monitoring subsystem, it includes first processor, the first power supply unit, the
Three wireless communication modules, the 4th wireless communication module, first interface circuit, the 4th sensor, the 5th sensor, wherein first
Power supply unit is electrically connected first processor and is electrically connected to provide driving electric energy, third/the 4th wireless communication module
It carries out information transmission to first processor based on instruction, and first interface circuit is electrically connected to the 4th/the 5th sensor to receive
Its sample information fed back and the first processor for transmitting it to electric connection.In this implementation which, the 4th sensor is
Temperature Humidity Sensor, the temperature and humidity information to sampling operation place.5th sensor is air velocity transducer.Third wireless communication
Module is bluetooth module, the location information by the module to determine staff in the region, the 4th wireless communication module
For WiFi module.Temperature, humidity, wind at station locating for staff of the station monitoring subsystem to its configuring area
Fast data, then host supervision system is sent for the data of acquisition by WiFi module.In addition, station monitoring subsystem passes through indigo plant
Tooth module sends positional distance information to staff's mobile phone terminal in real time, while realizing mobile phone positioning, also acts as system
Count the operating time of staff.It can configure multiple station monitoring subsystems in a workplace.The monitoring subsystem in this way
The environmental data in workshop at staff's station, including temperature, humidity, wind speed can be acquired.By WiFi module by acquisition
Data are sent to host supervision system.Location information is sent by staff mobile phone terminal of the bluetooth module on station.
First kind intelligent terminal energy in station monitoring subsystem region, the position that receiving station monitoring subsystem is sent
Confidence breath, and bluetooth positioning is carried out, staff is obtained in shop locations and the operating time of determining staff.Receive clothes
The staff that business device end sends over refers in workplace with the exposure concentration of n-hexane gas and health risk harm
Number, and can real-time display.The health risk hazard index for detecting the staff in certain post is more than certain numerical value, gives its intelligence
Terminal sends warning note.
In one embodiment, which also includes the second class intelligent terminal, acceptable server transmission
The health risk hazard index for showing each staff in workshop, facilitates management statistics, to adjust in time, takes pre-
Anti- measure.Show the workload of each staff in daily workshop.
In above-mentioned embodiment, need that the staff of workplace is positioned and counted its operating time, this
Apply for that embodiment uses the low-power consumption bluetooth module of built-in iBeacon agreement, the station prison in each location of personnel
Positional distance letter can be sent to staff's mobile phone terminal (i.e. first kind intelligent terminal) by broadcast mode by surveying subsystem
Breath.In order to guarantee that the first kind intelligent terminal at the station can receive the broadcast message of bluetooth, bluetooth module is using timing
Cold reset.In present embodiment using the Indoor Position Techniques Based on Location Fingerprint (LF, Location Fingerprint) based on RSSI into
Row positioning, constructs location information using multipath transmisstion, due to being influenced in signal communication process by landform and barrier, thus
Multipath shows very strong position particularity, for each position, the multidiameter configuration of the channel be it is unique, it
Multipath characteristics are considered the fingerprint of the position.Location fingerprint positioning is divided into two stages: offline fingerprint data collection rank
Section and online real-time positioning stage.Next the two stages are described in detail,
1) the offline fingerprint data collection stage,
The fingerprint data collection stage is mainly acquired finger print data, and regional location each in workplace is collected
Base station signal strength value handle and record according to certain condition as finger print data, finger print data and each region position
One-to-one correspondence is set, final building belongs to the fingerprint database of the workplace, is stored in the APP of (the first/bis- class) intelligent terminal
Client is prepared for next step positioning stage.It determines place, number of sampling points is set, establishes position coordinates, acquires each area
The signal strength indication in domain builds fingerprint database etc., and method is as follows
(1) it determines workplace, determine number of sampling points and workplace is divided by region and coordinate is set
In workplace, sampled point spacing, the number of Bluetooth base. station are to influence the principal element of positioning accuracy.Some researchs
The result shows that spacing is smaller between sampled point, positioning can't be more accurate, and sampled point spacing takes 2m or so.According to interior
The research conclusion of signal propagation model, Bluetooth base. station press 3~6 meters of arranged for interval, can arrange 4/8/16/ according to interior space size
32 Bluetooth base. stations.If each point requires to be acquired work in addition, there are many number of sampled point, this is just inevitable
The workload of the workload for constructing fingerprint database and later maintenance fingerprint database is caused to greatly increase, it is online fixed in real time to cause
The increase of position stage computation complexity.In the application embodiment, the experimental place area of selection is 80 square meters, 10 meters long, wide by 8
Rice, number of sampling points take 12, and Bluetooth base. station number takes 4.
Workplace is equably divided into 12 identical regions according to homalographic, each sampled point all in this 12
The center in a region, and be that experimental place establishes two-dimensional coordinate system, coordinate in whole coordinate system according to actual measurement distance
Unit be rice, as shown in Figure 3a, the lower left corner is experimental place doorway, and coordinate herein is set as (0,0), the upper right corner in laboratory
Coordinate is set as (10,8), and the center point coordinate in 12 regions can be calculated, this 12 groups of coordinate datas are stored in intelligent terminal
In, reference position information is provided for the tuning on-line stage of next step.
As shown in Figure 3b be this Bluetooth base. station configuration schematic diagram, deploy 4 Bluetooth base. stations in this implementation altogether, and by they
It is separately mounted to the top of workplace four sides wall, non-line-of-sight propagation is reduced with this.Since indoor environment is very complicated, bluetooth
Signal indoors spatial when, non line of sight effect can be generated, therefore Bluetooth base. station is preferably deployed in the interior spaces such as ceiling
Relatively high, spacious position.Due to the transmission power of each Bluetooth base. station be it is different, each Bluetooth base. station is being pacified
It requires to be configured before dress, AT instruction is sent to bluetooth module by processor (and single-chip microcontroller), by the hair of bluetooth module
Penetrate power grade be arranged to it is identical.In order to facilitate this four bluetooth modules are distinguished, instructed by AT to each bluetooth module
It is configured, they work in broadcast mode, and UUID, Major, Minor of each bluetooth module have respectively in broadcast message
From number, the APP client software of such intelligent terminal can identify when reading Bluetooth broadcast signal.
It is as shown in Figure 3b the APP client software process of intelligent terminal, firstly, (staff's) intelligent terminal APP
Client software calling interface function obtains all Bluetooth information in periphery, according to iBeacon agreement read fixed UUID,
The Bluetooth broadcast signal of Major and Minor.Then, it during reading Bluetooth signal, needs to carry out collected RSSI
Filtering, the positioning accuracy of entire positioning system are heavily dependent on the precision of location fingerprint data acquisition, intelligent terminal APP
The every signal strength indication for acquiring some 50 point of Bluetooth base. station of client software, just maximum and the smallest 10 numbers in this 50 points
It is rejected according to value, remaining data is carried out mean filter, obtained final result is as finger print data.Then according to the method described above
Successively in the signal strength indication of 12 station acquisitions, 4 Bluetooth base. stations, need to keep height and the court of mobile phone during acquisition
To being consistent, finally 12 groups of finger print datas are recorded and saved in intelligent terminal APP client.
2) online real-time positioning stage
The tuning on-line stage mainly passes through the signal that each base station that terminal real-time reception is distributed in workplace issues,
Then under online record each base station signal strength indication, and be compared with the data in the fingerprint database constructed offline,
One group of most like finger print data of signal strength indication is searched out from finger print data, finally by the regional location of this group of finger print data
Result as positioning.The mobile phone that the specific tuning on-line stage is in unknown position obtains different signal strength indications,
And be compared these signal strength indications with the signal strength indication in the fingerprint database for being stored in mobile phone terminal, pass through some
Positioning coordinate is obtained with algorithm.The performance of matching algorithm directly determines the precision of positioning, and determining matching algorithm has at present
Nearest neighbor algorithm (NN), K nearest neighbor algorithm (KNN) and K weighting nearest neighbor algorithm (WKNN).
Nearest neighbor algorithm (NN) is a kind of neighborhood matching algorithm on basis, its basic principle is will to receive at physical location
Each Bluetooth base. station signal strength indication and the position at the finger print data that collects offline calculated, obtain them it
Between Euclidean distance, the smallest one group of finger print data of Euclidean distance is then looked for out, finally by the coordinate of this group of finger print data
(xi, yi) it is used as final positioning result.
The calculation expression of Euclidean distance is as shown in following formula a in nearest neighbor algorithm (NN).
In formula, DiIndicate the signal strength indication RSSI for each Bluetooth base. station that the i-th bit place of setting receivesjAt the position
Finger print data RSSIijBetween Euclidean distance.
It is more single that nearest neighbor algorithm (NN) can only access closest one group of finger print data and position coordinates, positioning result
One, therefore in actual use, locating effect is poor vulnerable to interference, stability, causes positioning accuracy not high.
K nearest neighbor algorithm (KNN) is compared to nearest neighbor algorithm (NN) and proposes improvement, it is not to search out Euclidean distance
The smallest one group of finger print data, but lifting sequence is carried out after every group of Euclidean distance is all calculated, then find out Euclidean away from
From the smallest K group finger print data, the average value (x, y) of this K group finger print data position coordinates is finally taken to be used as final positioning result,
Shown in the following formula b of expression.
In formula, (x, y) is the two-dimensional coordinate of positioning result;K is finally to be used to average coordinate points number, under normal circumstances
K value is more than or equal to 2.
K weighting nearest neighbor algorithm (WKNN) is made that further Optimal improvements on the basis of K nearest neighbor algorithm (KNN).K
It is by the average value (x, y) of K group finger print data position coordinates in nearest neighbor algorithm (KNN) as final positioning result, but this
Obviously be not inconsistent in a practical situation it is reasonable, theoretically from actual measured value close to a certain group of finger print data, this group of fingerprint
The value of the coordinate value reference of data is bigger, so K weighting nearest neighbor algorithm (WKNN) proposes to introduce weight system on this basis
The position coordinates of neighbour are multiplied by a bigger weight coefficient η by several methodsi, smaller coordinate position weight coefficient
Smaller, it is 1 that all weight coefficients, which add up, finally calculates final coordinate value (X, Y) according to the weight proportion distributed
As positioning result, shown in the following formula c of expression.
In formula, (X, Y) is to introduce weighting coefficient ηiThe two-dimensional coordinate being calculated afterwards;ηiFor weight coefficient, calculation method
As shown in formula (d).
In formula, DiIndicate the signal strength indication RSSI for each Bluetooth base. station that the i-th bit place of setting receivesjAt the position
Finger print data RSSIijBetween Euclidean distance;ε is the constant of a very little, in order to guarantee that divisor is not zero.
In present embodiment, the matching algorithm positioned using K weighting nearest neighbor algorithm (WKNN) as location fingerprint is in intelligence
It is run in the APP client software of terminal.
Next description is combined, the workplace pernicious gas health risk quantitative evaluating method that the application proposes, this is commented
Estimate method and estimate based on BP neural network and (estimate) concentration such as Fig. 5 institute of the pernicious gas (such as n-hexane) of workplace
Show.This method comprises the following steps:
S11. the location information of workplace arbitrary point and the sample information of sensor are obtained;
S12. data processing model is constructed based on BP neural network;
S13. the gas concentration at any given position in workplace is calculated.
Based on the gas concentration value at each n-hexane sensor acquisition in gas sensor acquisition workshop, base in the S11
In Temperature Humidity Sensor temperature collection, humidity information, wind speed information is acquired based on air velocity transducer.Position is obtained based on bluetooth module
Confidence breath.
In the S12 based on BP neural network model processing module design in, activation primitive be selected from have Sigmoid function,
Tanh function, Relu function.In present embodiment, the concentration value of gas is indicated with the full scale percentage of sensor, number
It is function curve that value range selects Sigmoid function as activation primitive between 0% to 100%, therefore in present embodiment
As shown in Figure 6.
In processing module design in the S12 based on BP neural network model, structure is as shown in figure 4, present implementation
The input layer of middle BP neural network model has 6, is two-dimensional coordinate position data and 4 gas sensor concentration values respectively
Data, output node layer only have 1, are the sensor concentration estimated value at two-dimensional coordinate position.The hidden layer of BP neural network
It is designed as two layers, the node number of each hidden layer is designed to 30;Optimized learning algorithm uses self-adaption gradient descent method
(Adam), in the training process of BP neural network, select mean absolute error (MAE) as loss function.In BP nerve net
In network training, using the alcohol gas concentration cloud charts under 50 kinds of different conditions of Fluent software emulation, then utilize
CFD-Post processing software reads the alcohol gas concentration value from 200 positions at random from every width cloud atlas, and alcohol gas is dense
Angle value needs to be converted into volume fraction from mass fraction by calculating, and therefrom chooses volume fraction 3.3% hereinafter, i.e.
100%LEL (lower explosion limit) totally 6532 points below are as data sample.In 6532 group data sets, 70% data are chosen
As the training set of BP neural network, test set of 30% data as BP neural network.Because alcohol reaches in air
The volume fraction of lower explosion limit is 3.3%, i.e. 100%LEL, so exporting herein using LEL as BP neural network estimated value single
Position, full scale error (%FS) calculate the evaluation criterion of output result as BP neural network.Carrying out BP neural network training
Before need to pre-process 6532 groups of data.The concentration value at the two-dimensional coordinate of input, 4 sensors is done at normalization first
It manages, the alcohol gas concentration numerical value at arbitrary point is between 0 to 100%, so not doing normalized.It is obtained in training process
Loss function decline curve it is as shown in Figure 7.With the increase of frequency of training, loss function curve is gradually restrained, when training time
When number reaches 2,000,000 times, obtained mean absolute error value is 0.0076.Survey after having trained neural network model, to 30%
Examination data are tested.The fit solution of test data and reckoning result is as shown in Figure 8 a, and the reckoning result at each point is with this
Truthful data at point compares, and 1960 groups of data compare altogether, and solid line is true value, and dotted line is estimated value.Scheming
In 8a, abscissa is reference numeral a little, and ordinate is alcohol gas volume fraction.Estimated value and true value is finally calculated
Mean absolute error be 0.0086, calculate mean accuracy reach 0.86%FS.
In one embodiment, in order to more intuitively understand the practical mistake between each sample true value and estimated value
Difference, by the partial fitting curve magnification in Fig. 8 a.Fig. 8 b is the true value and estimated value that the test sample of preceding 200 points is taken in 8a
Comparative result figure.
True value and the reckoning result of preceding 100 test samples are as shown in figure 9, on the whole estimated value and true value are (real
Line) curve matching it is relatively good, it is all that relatively, a few point tolerance is larger, always that this 100 groups of data values are most of
Design is achieved the purpose that on body.
It is based on BP neural network model in the S13, calculates the harmful gas concentration of any position in workplace.In this way
To assess the working environment of staff at the position.Specifically, according to arbitrary point position coordinates and several gas in workshop
Concentration value at body sensor calculates the gas concentration value at the position using BP neural network model.
In one embodiment, which can be for according to the location information of staff and configuration on site several
The information of gas sensor sampling, the harmful gas concentration of any position in corresponding workshop is predicted based on BP neural network
Value.Its process is as shown in Figure 10, and this method comprises the following steps:
S21. the data processing model based on BP neural network building;
S22. the gas concentration in prediction scene at gas sensor;
S23. the location information of position to be predicted is obtained;
S24. predictive information is generated, that is, predicts the harm quotient in staff's future at the position.
This method utilizes BP neural network model, according to different mean wind speeds, different mean temperatures and different workers
Workload, the harmful gas concentration at sensor be fixedly mounted to several are predicted.The workload of staff passes through
Whether setting staff simulates on post, and the rate of volatilization for passing through setting fixed point sources in Fluent software is 0
Or whether some fixed value simulation staff is on post.In present embodiment, the input layer section of BP neural network model
Point has 10, is the alcohol point source rate of volatilization of wind speed, temperature and 8 fixed positions (station) respectively.BP neural network model
Output node layer has 4, is the gas sensor concentration value of 4 fixed positions.The hidden layer of BP neural network is designed as 2 layers,
The node number of each hidden layer is designed to 25.In BP neural network training process, select mean absolute error (MAE)
As loss function, intuitively reflect the decline process of average full scale error in training process in this way.In present embodiment, make
With self-adaption gradient descent method (Adam) train come model test data is predicted.In present embodiment, pass through
The simulation of Fluent software sensor installation place has under conditions of different wind speed, different temperatures, different operating person works amount
Evil gas concentration.Concentration distribution of the alcohol gas that Fluent is emulated in workshop, as shown in figure 14 in a workshop 10,
Alcohol gas volatilization point source 11 configured with 8 fixed positions and 4 fixed sensor positions for recording alcohol concentration value
12, wind speed, temperature and the 8 respective rate of volatilization of diffusion point source this 10 data in workshop are recorded as BP nerve
The alcohol concentration value of four fixed positions in workshop is recorded and is sensed as 4 alcohol gas by the input data of network
The collected real history data of device.
It is emulated by Fluent software in different condition and obtains 220 groups of data, concentration numbers at 4 sensors are provided
According to wherein 70% training set as BP neural network, 30% test set as BP neural network.Four in this 220 groups of data
Alcohol concentration value at a sensor needs to be converted into volume fraction from mass fraction by calculating, and volume fraction is both less than
3.3%, but because the volume fraction that alcohol reaches lower explosion limit in air is 3.3%, i.e. 100%LEL, so LEL is made
For the unit of BP neural network budget output valve, full scale error (%FS) predicts the evaluation of output result as BP neural network
Standard.
It needs to pre-process 220 groups of data before carrying out BP neural network training.First to the wind speed of input, temperature, 8
The respective rate of volatilization of a diffusion point source and the percentage of alcohol this 11 data normalizeds, 4 fixed positions
Alcohol gas concentration numerical value is between 0 to 100%, so not doing normalized.
In one embodiment, in the S21, batch gradient descent method (BGD) or self-adaption gradient descent method are selected
(Adam) as study optimization algorithm.
When batch gradient descent method (BGD) is as study optimization algorithm, obtained loss function decline curve such as Figure 11 a institute
Show.In fig. 11 a, with the increase of frequency of training, loss function just almost no longer declines.Frequency of training reaches 1,000,000 times
When, obtained mean absolute error value is 0.0076.Loss function curve in Figure 11 a is dropped to 0.0076 process from 0.025
Amplification it can be seen that loss function curve reforming phenomena it is obvious that as shown in figure 11b, with the increase loss function of frequency of training
It gradually tends to be steady, but still there is reforming phenomena always.
When self-adaption gradient descent method (Adam) is as study optimization algorithm, obtained loss function decline curve is as schemed
Shown in 11c.Loss function curve declines more steady with the increase of frequency of training.When frequency of training reaches 1,000,000 times,
Obtained mean absolute error value is 0.0037.By loss function curve from 0.25 drop to 0.0037 Product management model, such as scheme
Shown in 11d.It can be seen that loss function no longer declines when frequency of training reaches 700,000 times, it is in convergence state, at this time
Mean absolute error value is 0.0037.By comparison it can clearly be seen that study optimization when with BP neural network training pattern
Algorithms selection self-adaption gradient descent method (Adam) is more preferable than batch gradient descent method (BGD) effect.
It trains the model come using batch gradient descent method (BGD) to predict test data, in order to more intuitive
The fit solution of each test sample and prediction result is observed on ground, now by the comparison knot of the true value of test sample and prediction result
Fruit is provided with Figure 12.Be respectively in Figure 12 4 fixed position alcohol gas volume fractions true value and predicted value fitting and
At 4 curves, mark the true value (it goes out generally within the spike of line, mark be only illustrated as marking one by one) of a, other are pre-
Measured value.Abscissa is reference numeral a little in the figure, and ordinate is alcohol gas volume fraction.Predicted value is finally calculated
Mean absolute error with true value is 0.00583, and being equivalent to precision of prediction is 0.583%FS.
It trains the model come using self-adaption gradient descent method (Adam) to predict test data, such as Figure 13 institute
Show, mark the true value of a1 (it goes out generally within the spike of line, and mark is only illustrated as marking one by one).Prediction is finally calculated
Value and the mean absolute error of true value are 0.00097, and being equivalent to precision of prediction is 0.097%FS.
It can be concluded that, alcohol gas concentration value is predicted using BP neural network by a series of above comparisons
When, either training pattern or to model measurement selects self-adaption gradient descent method (Adam) all than batch gradient descent method
(BGD) effect is more preferable, and precision can reach 0.097%FS.
Workplace is also referred to as scene or workshop in above embodiment.Appraisal procedure sometimes can table in above embodiment
It states to calculate the gas concentration at any given position in workplace, can also be to be arbitrarily designated position in prediction work place
Set the gas concentration at place.
In above embodiment, in the design of workplace monitoring harmful gases system, in occupation work place
The main pernicious gas for leading to labourer's occupation slow poisoning is n-hexane or n-hexane and carbon disulfide.
The technical concepts and features of above embodiment only to illustrate the invention, its object is to allow be familiar with technique
People is to can understand the content of the present invention and implement it accordingly, and it is not intended to limit the scope of the present invention.All such as present invention
The equivalent transformation or modification that Spirit Essence is done, should be covered by the protection scope of the present invention.
Claims (10)
1. workplace pernicious gas health risk quantitative evaluating system, which is characterized in that it include server, at least one work
Place, each workplace are configured with, a host supervision system, several station monitoring subsystems,
Several described station monitoring subsystems are configured in the default station area of the workplace, distinguish matched connection
First kind intelligent terminal in the station area, the station monitoring subsystem transmit received first kind intelligent terminal information
To host supervision system connected to it, the host supervision system connects server, and interacts with the server info.
2. the system as claimed in claim 1, which is characterized in that the main monitoring system includes processor, power supply unit,
One wireless communication module, interface circuit, first sensor, second sensor, 3rd sensor, is shown the second wireless communication module
Show module, wherein power supply unit electrically connected processing device is to provide driving electric energy, and the first wireless communication module, second are wirelessly
Communication module is electrically connected processor and carries out information transmission based on instruction, and interface circuit is respectively and electrically connected to the first biography
Sensor, second sensor, 3rd sensor, to receive the sample information and the processing for being transmitted to electric connection that it feeds back respectively
Device,
Display module is electrically connected interface circuit, to show the information of the processor transmission.
3. system as claimed in claim 2, which is characterized in that
The first sensor is Temperature Humidity Sensor, to temperature/humidity information of sampling location,
Second sensor is gas sensor, to the harmful gas concentration information of sampling location,
3rd sensor is air velocity transducer, to the wind speed information of sampling location,
First wireless communication module is 4G module or IOT module, is electrically connected server,
Second wireless communication module be WiFi module or zigbee module, connect station monitoring subsystem, to the work
Position monitoring subsystem information exchange.
4. system as claimed in claim 3, which is characterized in that the second sensor is gas sensor, sampling location
The concentration information of the concentration information of n-hexane or the n-hexane of sampling location and sulfur dioxide.
5. the system as claimed in claim 1, which is characterized in that the station monitoring subsystem includes first processor, first
Power supply unit, third wireless communication module, the 4th wireless communication module, first interface circuit, the 4th sensor, the 5th sensing
Device, wherein
First power supply unit be electrically connected first processor to provide driving electric energy,
First processor is electrically connected in third wireless communication module, the 4th wireless communication module, and it carries out information based on instruction
Transmission,
First interface circuit is respectively and electrically connected to the 4th sensor, the 5th sensor to receive the sample information of its feedback and incite somebody to action
Its first processor for being transmitted to electric connection.
6. system as claimed in claim 5, which is characterized in that
4th sensor is temperature/humidity sensor, temperature/humidity information to sampling operation place;
5th sensor is air velocity transducer;
Third wireless communication module is bluetooth module, the first kind intelligent terminal in matching connection station area, and will connection
The first kind intelligent terminal information feed back to main monitoring system.
7. the system as claimed in claim 1, which is characterized in that letter of the server based on the main monitoring system transmission
Breath estimates the hazard index of staff, and the information of estimation is pushed to preset first kind intelligent terminal, the second class intelligence
Terminal.
8. workplace pernicious gas health risk quantitative evaluating method, which is characterized in that comprising any in such as claim 1-7
System described in, the method comprises the following steps:
S11. the location information of workplace arbitrary point and the sample information of sensor are obtained;
S12. data processing model is constructed based on BP neural network;
S13. the gas concentration at any given position in workplace is calculated.
9. method according to claim 8, which is characterized in that in the S12, the mode input node layer has 6, respectively
It is two-dimensional coordinate position data, the concentration Value Data of 4 gas sensors sampling, it is two-dimensional coordinate position that output node layer, which has 1,
Set the sensor concentration estimated value at place;The hidden layer of BP neural network is designed as two layers, and the node number of each hidden layer is 30
It is a.
10. workplace pernicious gas health risk quantitative evaluating method, which is characterized in that comprising as appointed in claim 1-7
System described in one, the method comprises the following steps:
S21. the data processing model based on BP neural network building;
S22. the gas concentration in prediction scene at gas sensor;
S23. the location information of position to be predicted is obtained;
S24. predictive information is generated, that is, predicts the harm quotient in staff's future at the position,
Wherein, test data is predicted based on self-adaption gradient descent method trained model, the input layer section of the model
Point has 10, is the information of wind speed, temperature and 8 preset fixed positions respectively;Model output node layer has 4, is 4 pre-
If fixation position gas sensor concentration value;The hidden layer of BP neural network is designed as 2 layers, the node of each hidden layer
Number is 25.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910394208.9A CN110046834B (en) | 2019-05-13 | 2019-05-13 | Workplace harmful gas health risk quantitative evaluation system and evaluation method thereof |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910394208.9A CN110046834B (en) | 2019-05-13 | 2019-05-13 | Workplace harmful gas health risk quantitative evaluation system and evaluation method thereof |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110046834A true CN110046834A (en) | 2019-07-23 |
CN110046834B CN110046834B (en) | 2022-06-24 |
Family
ID=67281567
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910394208.9A Active CN110046834B (en) | 2019-05-13 | 2019-05-13 | Workplace harmful gas health risk quantitative evaluation system and evaluation method thereof |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110046834B (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111047474A (en) * | 2019-10-21 | 2020-04-21 | 贝壳技术有限公司 | Indoor harmful substance volatilization time estimation method and device and storage medium |
CN111080976A (en) * | 2019-12-24 | 2020-04-28 | 北京优航机电技术有限公司 | Method and device for monitoring natural gas leakage in real time under temperature change scene |
CN111131239A (en) * | 2019-12-23 | 2020-05-08 | 杭州安恒信息技术股份有限公司 | Network security device, method, equipment and medium |
CN112763550A (en) * | 2020-12-29 | 2021-05-07 | 中国科学技术大学 | Integrated gas detection system with odor recognition function |
CN113611418A (en) * | 2021-08-05 | 2021-11-05 | 河北鑫牧思网络科技有限公司 | Occupational health automatic classification and grading management system, perception module and process |
CN113780690A (en) * | 2021-11-12 | 2021-12-10 | 南京亚流航空科技有限公司 | Environmental risk assessment system and method based on CFD software technology |
CN115577968A (en) * | 2022-11-02 | 2023-01-06 | 中国安全生产科学研究院 | Integral system and method based on enterprise violation management and automatic violation judgment |
Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20110063098A1 (en) * | 2009-09-16 | 2011-03-17 | Fischer John J | Standard mobile communication device distraction prevention and safety protocols |
CN102200456A (en) * | 2011-05-25 | 2011-09-28 | 杭州电子科技大学 | System for monitoring toxic and harmful gases in chemical industry park |
CN102855737A (en) * | 2012-09-06 | 2013-01-02 | 江苏科技大学 | All-dimensional harmful gas detection alarm device in welding field |
CN102890754A (en) * | 2012-10-31 | 2013-01-23 | 中国科学院自动化研究所 | Danger source monitoring system for mine |
CN103914622A (en) * | 2014-04-04 | 2014-07-09 | 清华大学 | Quick chemical leakage predicating and warning emergency response decision-making method |
CN205384049U (en) * | 2016-02-18 | 2016-07-13 | 深圳市职业病防治院 | Occupational disease harm risk monitor terminal and system |
CN105989243A (en) * | 2016-03-29 | 2016-10-05 | 深圳市职业病防治院 | Occupational disease hazardous risk assessment method based on real-time information big data of working environment |
CA2932204A1 (en) * | 2015-06-25 | 2016-12-25 | Alaya Care Inc. | Method for predicting adverse events for home healthcare of remotely monitored patients |
CN107358388A (en) * | 2016-11-03 | 2017-11-17 | 厦门嵘拓物联科技有限公司 | A kind of WMS based on Internet of Things and the storage quality risk appraisal procedure based on the system |
CN108506043A (en) * | 2018-04-08 | 2018-09-07 | 安徽宏昌机电装备制造有限公司 | A kind of downhole intelligent InterWorking Equipment monitoring system and its personnel positioning method based on NB-IOT honeycomb technology of Internet of things |
CN108650648A (en) * | 2018-04-04 | 2018-10-12 | 云南民族大学 | A kind of indoor environment monitoring system |
-
2019
- 2019-05-13 CN CN201910394208.9A patent/CN110046834B/en active Active
Patent Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20110063098A1 (en) * | 2009-09-16 | 2011-03-17 | Fischer John J | Standard mobile communication device distraction prevention and safety protocols |
CN102200456A (en) * | 2011-05-25 | 2011-09-28 | 杭州电子科技大学 | System for monitoring toxic and harmful gases in chemical industry park |
CN102855737A (en) * | 2012-09-06 | 2013-01-02 | 江苏科技大学 | All-dimensional harmful gas detection alarm device in welding field |
CN102890754A (en) * | 2012-10-31 | 2013-01-23 | 中国科学院自动化研究所 | Danger source monitoring system for mine |
CN103914622A (en) * | 2014-04-04 | 2014-07-09 | 清华大学 | Quick chemical leakage predicating and warning emergency response decision-making method |
CA2932204A1 (en) * | 2015-06-25 | 2016-12-25 | Alaya Care Inc. | Method for predicting adverse events for home healthcare of remotely monitored patients |
CN205384049U (en) * | 2016-02-18 | 2016-07-13 | 深圳市职业病防治院 | Occupational disease harm risk monitor terminal and system |
CN105989243A (en) * | 2016-03-29 | 2016-10-05 | 深圳市职业病防治院 | Occupational disease hazardous risk assessment method based on real-time information big data of working environment |
CN107358388A (en) * | 2016-11-03 | 2017-11-17 | 厦门嵘拓物联科技有限公司 | A kind of WMS based on Internet of Things and the storage quality risk appraisal procedure based on the system |
CN108650648A (en) * | 2018-04-04 | 2018-10-12 | 云南民族大学 | A kind of indoor environment monitoring system |
CN108506043A (en) * | 2018-04-08 | 2018-09-07 | 安徽宏昌机电装备制造有限公司 | A kind of downhole intelligent InterWorking Equipment monitoring system and its personnel positioning method based on NB-IOT honeycomb technology of Internet of things |
Non-Patent Citations (1)
Title |
---|
李麟等: "《智慧银行 未来银行服务新模式》", 30 November 2017 * |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111047474A (en) * | 2019-10-21 | 2020-04-21 | 贝壳技术有限公司 | Indoor harmful substance volatilization time estimation method and device and storage medium |
CN111047474B (en) * | 2019-10-21 | 2024-03-22 | 贝壳技术有限公司 | Indoor harmful substance volatilization time estimation method, device and storage medium |
CN111131239A (en) * | 2019-12-23 | 2020-05-08 | 杭州安恒信息技术股份有限公司 | Network security device, method, equipment and medium |
CN111131239B (en) * | 2019-12-23 | 2022-03-22 | 杭州安恒信息技术股份有限公司 | Network security device, method, equipment and medium |
CN111080976A (en) * | 2019-12-24 | 2020-04-28 | 北京优航机电技术有限公司 | Method and device for monitoring natural gas leakage in real time under temperature change scene |
CN112763550A (en) * | 2020-12-29 | 2021-05-07 | 中国科学技术大学 | Integrated gas detection system with odor recognition function |
CN113611418A (en) * | 2021-08-05 | 2021-11-05 | 河北鑫牧思网络科技有限公司 | Occupational health automatic classification and grading management system, perception module and process |
CN113611418B (en) * | 2021-08-05 | 2024-05-14 | 河北鑫牧思网络科技有限公司 | Professional health automatic classification hierarchical management system, perception module and process |
CN113780690A (en) * | 2021-11-12 | 2021-12-10 | 南京亚流航空科技有限公司 | Environmental risk assessment system and method based on CFD software technology |
CN113780690B (en) * | 2021-11-12 | 2022-02-15 | 南京亚流航空科技有限公司 | Environmental risk assessment system and method based on CFD software technology |
CN115577968A (en) * | 2022-11-02 | 2023-01-06 | 中国安全生产科学研究院 | Integral system and method based on enterprise violation management and automatic violation judgment |
Also Published As
Publication number | Publication date |
---|---|
CN110046834B (en) | 2022-06-24 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110046834A (en) | Workplace pernicious gas health risk quantitative evaluating system and its appraisal procedure | |
CN109444813B (en) | RFID indoor positioning method based on BP and DNN double neural networks | |
CN201594088U (en) | Novel indoor wireless air quality monitor | |
CN107328437A (en) | Towards the wearable device of electric power safety inspection operation | |
CN106338705A (en) | Error verifying method and device for three-phase electric energy metering device | |
CN111601270A (en) | Internet of things power monitoring system and method of wireless sensor network topological structure | |
CN202486118U (en) | Gas environment emergency detection system | |
CN106647514A (en) | Cement enterprise carbon emission real-time on-line monitoring management system | |
CN117610779B (en) | Mapping geographic information quality supervision method and system based on CORS big data | |
CN109580268A (en) | A kind of product abnormal sound, abnormal sound intelligent detecting method | |
CN116626238A (en) | Dual-channel detection compass system for sensing air flow and air combination, air leakage detection method, data fusion and tracking method | |
CN116308958A (en) | Carbon emission online detection and early warning system and method based on mobile terminal | |
CN112702408A (en) | Internet of things system and method based on multi-sensing function | |
WO2019224545A2 (en) | A system for detecting air pollution | |
CN112326583A (en) | Intelligent toxic gas detection system and method based on Internet of things | |
CN113155196A (en) | Bridge operation real-time monitoring system based on AIoT and monitoring method thereof | |
KR20210085772A (en) | Multi-sensor based air quality status mobile notification system and method | |
CN106770532B (en) | A kind of forecasting of Gas Concentration device and gas concentration real-time detection and forecasting system | |
CN202093016U (en) | Multi-parameter comprehensive COD (Chemical Oxygen Demand) water quality monitoring system | |
CN108810843A (en) | Urban air-quality intelligent monitor system | |
Andria et al. | Model characterization in measurements of environmental pollutants via data correlation of sensor outputs | |
CN115936680A (en) | Intelligent order dispatching method and system for equipment operation and maintenance | |
CN106530175A (en) | Urban ecosystem health diagnosis system | |
CN112261718B (en) | Indoor positioning method | |
CN112150008B (en) | System structure evaluation method and system based on design data and experimental data |
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 |