CN117542174A - Fire detection alarm based on Internet of things environment sampling intelligent algorithm and system thereof - Google Patents
Fire detection alarm based on Internet of things environment sampling intelligent algorithm and system thereof Download PDFInfo
- Publication number
- CN117542174A CN117542174A CN202311426807.7A CN202311426807A CN117542174A CN 117542174 A CN117542174 A CN 117542174A CN 202311426807 A CN202311426807 A CN 202311426807A CN 117542174 A CN117542174 A CN 117542174A
- Authority
- CN
- China
- Prior art keywords
- fire
- early warning
- environment
- smoke
- neural network
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 45
- 238000004422 calculation algorithm Methods 0.000 title claims abstract description 36
- 238000005070 sampling Methods 0.000 title claims abstract description 23
- 239000000779 smoke Substances 0.000 claims abstract description 76
- 238000009434 installation Methods 0.000 claims abstract description 36
- KLARSDUHONHPRF-UHFFFAOYSA-N [Li].[Mn] Chemical compound [Li].[Mn] KLARSDUHONHPRF-UHFFFAOYSA-N 0.000 claims abstract description 13
- 230000007246 mechanism Effects 0.000 claims abstract description 7
- 238000012795 verification Methods 0.000 claims abstract description 6
- 230000006855 networking Effects 0.000 claims abstract description 4
- 238000013528 artificial neural network Methods 0.000 claims description 38
- 239000000428 dust Substances 0.000 claims description 19
- 239000002131 composite material Substances 0.000 claims description 15
- UGFAIRIUMAVXCW-UHFFFAOYSA-N Carbon monoxide Chemical compound [O+]#[C-] UGFAIRIUMAVXCW-UHFFFAOYSA-N 0.000 claims description 13
- 230000006870 function Effects 0.000 claims description 13
- 229910002091 carbon monoxide Inorganic materials 0.000 claims description 12
- 230000002068 genetic effect Effects 0.000 claims description 12
- 239000002245 particle Substances 0.000 claims description 12
- 230000030279 gene silencing Effects 0.000 claims description 11
- 238000000034 method Methods 0.000 claims description 11
- 230000004044 response Effects 0.000 claims description 10
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Chemical compound O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 10
- 230000008859 change Effects 0.000 claims description 9
- 239000011521 glass Substances 0.000 claims description 9
- 150000001875 compounds Chemical class 0.000 claims description 8
- 238000012544 monitoring process Methods 0.000 claims description 8
- 238000012549 training Methods 0.000 claims description 7
- 238000003379 elimination reaction Methods 0.000 claims description 5
- 230000007613 environmental effect Effects 0.000 claims description 5
- 230000035945 sensitivity Effects 0.000 claims description 4
- 206010052804 Drug tolerance Diseases 0.000 claims description 3
- 230000006978 adaptation Effects 0.000 claims description 3
- 230000006399 behavior Effects 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims description 3
- 238000007791 dehumidification Methods 0.000 claims description 3
- 238000001035 drying Methods 0.000 claims description 3
- 230000008030 elimination Effects 0.000 claims description 3
- 230000026781 habituation Effects 0.000 claims description 3
- 230000001769 paralizing effect Effects 0.000 claims description 3
- 239000000523 sample Substances 0.000 description 13
- 230000003287 optical effect Effects 0.000 description 7
- 238000004364 calculation method Methods 0.000 description 4
- 241000238631 Hexapoda Species 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 3
- 238000002485 combustion reaction Methods 0.000 description 3
- 239000012535 impurity Substances 0.000 description 3
- 230000002093 peripheral effect Effects 0.000 description 3
- 230000002265 prevention Effects 0.000 description 3
- 230000000630 rising effect Effects 0.000 description 3
- 238000009825 accumulation Methods 0.000 description 2
- 239000000443 aerosol Substances 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000012790 confirmation Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 230000006698 induction Effects 0.000 description 2
- 238000011897 real-time detection Methods 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 239000007921 spray Substances 0.000 description 2
- 230000001960 triggered effect Effects 0.000 description 2
- 241000256626 Pterygota <winged insects> Species 0.000 description 1
- 230000005856 abnormality Effects 0.000 description 1
- 238000010521 absorption reaction Methods 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- QVFWZNCVPCJQOP-UHFFFAOYSA-N chloralodol Chemical compound CC(O)(C)CC(C)OC(O)C(Cl)(Cl)Cl QVFWZNCVPCJQOP-UHFFFAOYSA-N 0.000 description 1
- 238000010411 cooking Methods 0.000 description 1
- 238000005202 decontamination Methods 0.000 description 1
- 230000003588 decontaminative effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 239000003546 flue gas Substances 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 238000003825 pressing Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B29/00—Checking or monitoring of signalling or alarm systems; Prevention or correction of operating errors, e.g. preventing unauthorised operation
- G08B29/18—Prevention or correction of operating errors
- G08B29/185—Signal analysis techniques for reducing or preventing false alarms or for enhancing the reliability of the system
- G08B29/186—Fuzzy logic; neural networks
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B25/00—Alarm systems in which the location of the alarm condition is signalled to a central station, e.g. fire or police telegraphic systems
- G08B25/008—Alarm setting and unsetting, i.e. arming or disarming of the security system
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B29/00—Checking or monitoring of signalling or alarm systems; Prevention or correction of operating errors, e.g. preventing unauthorised operation
- G08B29/02—Monitoring continuously signalling or alarm systems
- G08B29/04—Monitoring of the detection circuits
- G08B29/043—Monitoring of the detection circuits of fire detection circuits
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B29/00—Checking or monitoring of signalling or alarm systems; Prevention or correction of operating errors, e.g. preventing unauthorised operation
- G08B29/18—Prevention or correction of operating errors
- G08B29/20—Calibration, including self-calibrating arrangements
- G08B29/24—Self-calibration, e.g. compensating for environmental drift or ageing of components
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Security & Cryptography (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Business, Economics & Management (AREA)
- Emergency Management (AREA)
- Fire Alarms (AREA)
Abstract
The utility model relates to a fire detection alarm technical field discloses a fire detection alarm based on intelligent algorithm of thing networking environment sampling, including the installation chassis, installation chassis bottom fixedly connected with bottom, bottom internally mounted has lithium-manganese battery, lithium-manganese battery with install battery clamping spring between the bottom inner wall, battery clamping button is installed to bottom inside one side, PCB subassembly is installed to the bottom, PCB subassembly with install between the bottom and reset with the amortization button, installation chassis bottom edge fixedly connected with circular main part, circular main part bottom is installed and is responded to the knot, circular main part with go up to respond to and install anti-disassembly mechanism and switch between the knot. By establishing multiple sensing channels, acquiring detection values of the multiple channels, performing cross verification, establishing a model, further juxtaposing all the channels, and directly performing intelligent recognition on smoke detection conditions by collecting data.
Description
Technical Field
The invention relates to the technical field of fire detection alarms, in particular to a fire detection alarm based on an intelligent sampling algorithm of an environment of the Internet of things and a system thereof.
Background
Smoke is a fire smoke detection alarm or a free-standing fire smoke alarm, which is widely used in various fire alarm systems by monitoring the concentration of smoke to achieve fire prevention. Under normal conditions, the optical maze in smoke sense encounters smoke and only alarms after triggering the photoelectric sensor.
Under normal conditions, the optical maze in smoke sense encounters smoke and only alarms after triggering the photoelectric sensor. However, because the sensor is sensitive to very small smoke particles, dust accumulation exists in the optical labyrinth after the smoke is installed for a long time, dust can be blown up when airflow passes through the smoke, and the smoke is mistakenly considered as smoke to cause false alarm; in addition, aerosol like water vapor as small particles may also cause smoke false alarms. Therefore, the invention provides a fire detection alarm based on an intelligent sampling algorithm of the environment of the Internet of things and a system thereof.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides a fire detection alarm based on an intelligent sampling algorithm of the environment of the Internet of things and a system thereof, which solve the problem that the smoke is mistaken as smoke and false alarm is caused by the fact that the sensor is sensitive to tiny smoke particles, dust accumulation exists in an optical labyrinth after the smoke is installed for a long time, and dust can be blown up when air flows pass through the smoke.
In order to achieve the above purpose, the invention is realized by the following technical scheme: the utility model provides a fire detection alarm based on intelligent algorithm of thing networking environment sample, includes the installation chassis, installation chassis bottom fixedly connected with bottom, bottom internally mounted has lithium manganese battery, lithium manganese battery with install battery chucking spring between the bottom inner wall, battery chucking button is installed to bottom inside one side, the PCB subassembly is installed to the bottom, the PCB subassembly with install reset and amortization button between the bottom, installation chassis bottom edge fixedly connected with circular main part, the response is detained in the circular main part bottom installation, circular main part with install anti-disassembly mechanism and switch between the response is detained in the last response, circular main part internally mounted has compound maze, detachable connection has the detachable fly net in the outside of compound maze, buzzing piece is installed to compound maze bottom, maze lower cover is installed at compound maze top, the optical path lid is installed to maze lower cover top, the optical path lid is located PCB subassembly bottom.
Preferably, the PCB assembly comprises a PCB, the PCB top is installed in the bottom of the bottom cover, the GSIM clamping groove seat is installed at the PCB top, two red and blue double light emitting tubes are installed on the middle side of the PCB bottom, a shielding cover is installed on one side of the PCB bottom, a humidity sensitive detector is installed on the other side of the PCB bottom, a first glass head NTC thermistor is installed on one side of the edge of the PCB, a second glass head NTC thermistor is installed on the other side of the edge of the PCB, and a carbon monoxide detector is installed on one side, close to the shielding cover, of the bottom of the PCB.
The fire detection alarm system method based on the Internet of things environment sampling intelligent algorithm comprises the following specific steps:
initializing smoke feeling to obtain an initialized sample data set A;
step two, setting a humidity sensor and a carbon monoxide sensor in cooperation with smoke feeling to obtain an initialized sample data set B;
step three, setting a temperature sensor in cooperation with smoke feeling to obtain an initialized sample data set C;
step four, constructing a BP neural network, taking the sets A and B as input samples, taking the set C as a verification set, and training the BP neural network;
fifthly, placing a humidity sensor in the smoke sense;
step six, acquiring data of the smoke sensor and the humidity sensor in real time, and outputting the data based on the trained BP neural network;
step seven, alarming or continuously monitoring based on the output result;
and step eight, when the installation application of the specific position is carried out, the detector can adapt to and adjust the corresponding early warning value according to the actual installation application environment.
Preferably, the temperature sensor in the third step is an NTC sensor.
Preferably, the BP neural network in the fourth step is a BP neural network based on a genetic algorithm.
Preferably, the constructing the BP neural network based on the genetic algorithm in the fourth step includes the following steps;
s1, initializing a population N;
s2, determining an adaptability function and control parameters;
s3, selecting, crossing and mutating operations are carried out, and a result is fed back to the BP neural network;
s4, if the optimal weight estimation is obtained, constructing a BP neural network based on a genetic algorithm, otherwise, returning to the step S3.
Preferably, the function in S2Wherein e is the maximum estimation error of the improved BP neural network, Y i For actual output, C i An output is desired.
Preferably, the error function of the BP neural network comprises a penalty term.
Preferably, in the seventh step, when the output result is that the water vapor is too much, the judgment is performed to cause the photoelectric early warning and the temperature change to be that the installation adaptation environment is that the water vapor triggers the alarm, and the judgment result is output, and the detector is reminded that the manual dehumidification drying treatment is needed.
Preferably, the self-adapting and self-adjusting the corresponding pre-warning value in the step eight includes the following steps:
i, through data sampling of actual product installation environment temperature, humidity and air particles, combining the specific building and space use position names determined during the installation of a specific front-end app system;
II, after the detector triggers early warning due to smoke or dust, the detector is judged to be not a fire alarm but an installation and application environment after being judged by other various conditions such as temperature, humidity and the like;
III, after the detector triggers the early warning in the step II for a plurality of times within a certain period of time, the system and the equipment can comprehensively evaluate the sensitivity threshold value of the environmental smoke detection and carry out special individual adjustment on the automatic early warning threshold value;
IV, comprehensively adjusting the punishment early warning threshold value of the application installation environment according to early warning times and time frequency to avoid excessive early warning consuming the electric quantity of the composite fire detector and causing frequent early warning to cause habituation of multiple fire warning behaviors of a user, so that paralytic consciousness is formed and the importance of fire warning is ignored;
v, after non-fire early warning judgment, equipment local photoelectric early warning is automatically carried out, and then non-fire alarm automatic elimination early warning sound is judged, so that fire alarm accurate timing linkage multiparty pushing is realized, after early warning, non-fire alarm timely automatic background silencing cancellation early warning can not continuously disturb a user, and other staff can not be disturbed, so that false alarm is formed.
The invention provides a fire detection alarm and a fire detection alarm system based on an intelligent sampling algorithm of an environment of the Internet of things.
The beneficial effects are as follows:
1. according to the invention, by establishing multiple sensing channels, acquiring detection values of the multiple channels, performing cross verification and establishing a model, and further juxtaposing all channels, the intelligent recognition is directly performed on detection conditions by collecting data.
2. The model disclosed by the invention has the advantages of high self-adaption degree, good robustness, high monitoring automation degree, small interference of factors forming false alarms of the detector and small false alarm probability.
3. The fire detector system can perform self-learning, automatically adjust the smoke concentration early warning threshold value adapting to the environment, avoid excessive early warning to consume the electric quantity of the composite fire detector, and prolong the standby service time of the equipment by reducing the false alarm phenomenon.
Drawings
FIG. 1 is an exploded view of the present invention;
FIG. 2 is a schematic view of a composite maze structure of the present invention;
FIG. 3 is a schematic view of a PCB assembly according to the present invention;
FIG. 4 is a flow chart of the BP neural network construction of the present invention;
FIG. 5 is a flow chart of the operation of the fire detection alarm of the present invention
Fig. 6 is a flowchart of the fire detection alarm operation of the present invention.
Wherein 010, upper induction button; 020. a circular body; 030. a buzzer; 040. a composite maze; 050. a detachable insect-proof net; 060. a labyrinth lower cover; 070. a track cover; 080. a PCB assembly; 081. a PCB board; 082. a moisture sensitive detector; 083. a first glass-head NTC thermistor; 084. a glass head NTC thermistor II; 085. a red Lan Shuangguang emitter tube; 086. a carbon monoxide detector; 087. a shield; 088. 4GSIM card slot seat; 090. a bottom cover; 100. lithium-manganese battery; 110. installing a chassis; 120. a battery clamping button; 130. a battery clamping spring; 140. resetting the follow silencing key; 150. an anti-disassembly mechanism and a switch.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are only some, but not all embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Examples:
referring to fig. 1-5, the embodiment of the invention provides a fire detection alarm based on an intelligent sampling algorithm of an environment of the internet of things, which comprises a mounting chassis 110, wherein a bottom cover 090 is fixedly connected to the bottom of the mounting chassis 110, and the mounting chassis 110 is mounted on the top or a wall of a building for fixing the whole fire detector. The bottom cover 090 is internally provided with the lithium-manganese battery 100, and the lithium-manganese battery 100 provides necessary power support for the operation of the fire detector, so that the fire detector cannot work normally due to power failure. A battery clamping spring 130 is installed between the lithium-manganese battery 100 and the inner wall of the bottom cover 090, and the lithium-manganese battery 100 can be clamped in a battery groove inside the bottom cover 090 by the battery clamping spring 130. A battery clamping button 120 is installed at one side of the inside of the bottom cover 090, and the lithium-manganese battery 100 can be quickly and conveniently taken out of the groove in the bottom cover 090 through the battery clamping button 120 when the lithium-manganese battery 100 needs to be replaced. The voltage of the battery is lower than 2.7V, the alarm gives an audible and visual alarm (1 sound) every 50S, which indicates that the battery is under-voltage, and the battery is reminded to be replaced in time, otherwise, the normal operation of the alarm is influenced; when the battery voltage is lower than 2.6V, the alarm stops running and enters a deep sleep state.
A PCB assembly 080 is mounted to the bottom of the bottom cover 090, and the PCB assembly 080 is used to control the normal operation of the fire detector. A reset and silencing button 140 is arranged between the PCB component 080 and the bottom cover 090, and the probe can be controlled to reset and close the early warning alarm by pressing the reset and silencing button 140 through the PCB component 080 connected with the reset and silencing button. The short press self-checking/silencing function is used for self-checking alarm test of the alarm, silencing function under an alarm state and the like. When the self-checking/silencing key is pressed for 3 s-5 s for a long time, the fire alarm sound-light alarm is started, and alarm data are sent at the same time, and the duration is 300s. The circular main body 020 is fixedly connected to the bottom edge of the mounting chassis 110, and the circular main body 020 is an outer shell part of the detector, so that internal electronic components can be protected. The circular main part 020 bottom is installed and is detained 010, and external flue gas, dust etc. can detain 010 through last response and get into the detector inside, avoid impurity such as insect to get into in the detector simultaneously, lead to the detector to appear the false alarm. Install between circular main part 020 and the last response knot 010 and prevent tearing open mechanism and switch 150, can conveniently with last response knot 010 dismantlement through preventing tearing open mechanism and switch 150, after equipment installation debugging is fixed, if when dismantling not hard up action appears, prevent tearing open and close because not hard up or when removing, contact switch can form the loop response, send promptly and prevent tearing open the early warning to form a set of equipment through the thing networking communication and tear open and move early warning information, app has the position of corresponding combined type fire detection alarm and has removal early warning suggestion and record. Manual or background recording and reconfirmation of the geographical and spatial location of the device alarms is required. So as to prevent the position of the fire alarm from changing when the fire alarm occurs, and the actual specific position of the fire alarm notification alarm is wrong. The dust and impurities outside the upper induction button 010 are removed, the composite maze 040 is arranged in the circular main body 020, the detachable insect screen 050 is detachably connected to the outer side of the composite maze 040, the impurities such as dust can be further prevented from entering the inner part of the composite maze 040 by the detachable insect screen 050, the inner part of the composite maze 040 is kept smooth, and information about fire such as light and smoke is transmitted to the lower detection part of the PCB assembly 080. The buzzer 030 is installed to compound maze 040 bottom, and when PCB subassembly 080 detects data anomaly, just can start the buzzer 030 and send out the alarm. A labyrinth lower cover 060 is arranged at the top of the composite labyrinth 040, a light channel cover 070 is arranged at the top of the labyrinth lower cover 060, and the light channel cover 070 is positioned at the bottom of the PCB component 080.
PCB assembly 080 includes a PCB board 081, which is the main circuit board of the fire detector, that contains a number of important electronic components for controlling and monitoring various aspects of the fire detector. The top of the PCB 081 is arranged at the bottom of the bottom cover 090, and the top of the PCB 081 is provided with a 4GSIM card slot 088 for keeping communication with the outside, so as to realize data transmission. Two red and blue double light emitting tubes 085 are arranged on the middle side of the bottom of the PCB 081, and light sources emitted by the two red and blue double light emitting tubes 085 enter the composite labyrinth 040 through a light channel cover 070.
The double-light double-emission single-receiving maze photoelectric detector emits red light and blue light simultaneously, and the light beams are scattered and absorbed when passing through smoke and dust of a maze. The reflected light signals are received by the detector, and the smoke and dust conditions can be known by analyzing the intensity and wavelength of the light signals.
Screening particle size: the intensity of the reflected red and blue light is affected by the particle size. Generally, the commonly used infrared photoelectric smoke detectors have lower particle sensitivity below 300nm, but have strong response to interference particles larger than 1 mu m (such as dust and small particles), the dust particle size distribution range of 0.97-176 mu m has stronger scattering effect on red light, the combustion smoke aerosol is mostly in submicron order, and the absorption effect of light concentrated and distributed at 100-500 nm is stronger. Therefore, by comparing the reflected intensities of the red light and the blue light, it is possible to discriminate smoke and dust.
Specifically, if the red light reflection intensity is greater than the blue light reflection intensity, it can be judged that smoke is present; if the blue light reflection intensity is greater than the red light reflection intensity, it can be judged that dust is present.
In conclusion, the double-light double-emission single-receiving labyrinth photoelectric detector can be used for screening dust and smoke. In addition, a double-light double-emission single-receiving hardware maze and a circuit structure are used as a primary early warning mechanism, and then specific smoke concentration value ratios are detected and output to perform continuous detection for a plurality of times so as to provide training for a system and provide real-time detection basis after early warning for a bp neural algorithm.
A shielding cover 087 is installed at one side of the bottom of the PCB 081, and the shielding cover 087 is used for shielding external information interference and ensuring that the detector can stably operate. The humidity-sensitive detector 082 is installed to the opposite side of PCB board 081 bottom for detect the humidity in control place, and glass head NTC thermistor one 083 is installed to PCB board 081 edge one side, and glass head NTC thermistor two 084 is installed to PCB board 081 edge opposite side, and glass head NTC thermistor one 083 and glass head NTC thermistor two 084 carry out real-time detection to the temperature in control place. A carbon monoxide detector 086 is installed at one side of the bottom of the PCB 081, which is close to the shielding case 087, and carbon monoxide generated by combustion is detected by the carbon monoxide detector 086, so that whether a fire disaster occurs in a monitoring place is further confirmed. The presence of the combustion products can be directly confirmed by the carbon monoxide detector 086. When the fire detector triggers an alarm, early confirmation of the fire should be made within one minute. Such as when the kitchen is cooking, it is very easy to trigger an alarm due to carbon monoxide. Meanwhile, in offices, living rooms and other places, false alarms caused by dust and water vapor are also easily caused, and the primary confirmation of necessary fires is also indispensable. After early warning and judgment, the equipment local photoelectric early warning and the non-fire alarm automatic elimination early warning and silencing are automatically carried out, so that the fire alarm accurate timing linkage multiparty pushing is realized, the early warning and the non-fire alarm automatic background silencing and cancelling early warning can not continuously disturb a user, and other staff can not be disturbed, so that false alarm is formed.
The fire detection alarm system method based on the Internet of things environment sampling intelligent algorithm comprises the following specific steps:
step one, initializing smoke feeling and obtaining an initialized sample data set A.
And secondly, setting a humidity sensor and a carbon monoxide sensor in cooperation with smoke feeling to obtain an initialized sample data set B.
And thirdly, setting a temperature sensor in cooperation with smoke feeling to obtain an initialized sample data set C.
In the third step, the temperature sensor is an NTC sensor.
The initialization sample data set in the first to third steps may collect data of smoke feeling under various use conditions, or obtain a group of data with a brand new smoke feeling, obtain a group of data with a smoke feeling of three months, and the like, and finally aggregate all the groups of data together to become the initialization sample data set A, B, C.
To maintain the accuracy of the operation of the network for a single product, the elements in the sample data set A, B, C are collected from the same or a class of smoke-sensing products in a continuous tracking collection.
And fourthly, constructing a BP neural network, taking the sets A and B as input samples, taking the set C as a verification set, and training the BP neural network.
And step four, the BP neural network is based on a genetic algorithm.
In the fourth step, constructing the BP neural network based on the genetic algorithm comprises the following steps:
s1, initializing and grouping N;
s2, determining an adaptability function and control parameters;
in step S2, the functionWherein e is the improvement of BP god
Maximum estimation error via network, Y i For actual output, C i An output is desired.
S3, selecting, crossing and mutating operations are carried out, and a result is fed back to the BP neural network;
s4, if the optimal weight estimation is obtained, constructing a BP neural network based on a genetic algorithm, otherwise, returning to the step S3.
The error function of the BP neural network comprises a penalty term.
The BP neural network is used for prediction operation, is a multi-layer feedforward neural network, and is characterized in that signals are transmitted forward and errors are transmitted backward, and the weight relation between input and output is found out by utilizing the existing data so as to perform simulation.
N in S1 is the power N of 2.
Further, the BP neural network is optimized by a genetic algorithm, the weight and the value of the BP neural network can be optimized, each individual in the population comprises a network ownership weight and a threshold value, the individual fitness value is calculated by the individual through a fitness function, the genetic algorithm finds the individual corresponding to the optimal fitness value through selection, intersection and mutation operation, so that the BP neural network obtains the optimal individual by the genetic algorithm, assigns a value to the initial weight and the running value of the network, and the detection accuracy is higher.
In order to avoid overfitting and further increase the robustness of the network, penalty items are added in the loss function, so that constraint items are added in the loss function compared with a common BP neural network; the penalty term is set as known in the art, and can be set by one skilled in the art according to the requirements.
Fifthly, a humidity sensor is arranged in the smoke sense.
And step six, acquiring the data of the smoke sensor and the humidity sensor in real time, and outputting the data based on the trained BP neural network.
After the network construction is completed, a humidity sensor and a temperature sensor are arranged in a new smoke-sensing two-stage conjoined labyrinth darkroom, and the humidity sensor and a normal one-stage optical labyrinth are used for jointly detecting and outputting a network judgment result, so as to further confirm whether to trigger an alarm.
And step seven, alarming or continuously monitoring based on the output result.
And step seven, when the output result is that the water vapor is too much, judging to lead to photoelectric early warning and temperature change, wherein the installation adaptation environment is water vapor triggering alarm, and the judgment result is output to remind the detector that manual dehumidification and drying treatment are needed.
When the output result includes the oversized sensing value of the humidity sensor, the device abnormality prompt should be output first for manual inspection treatment.
And step eight, when the installation application of the specific position is carried out, the detector can adapt to and adjust the corresponding early warning value according to the actual installation application environment.
The front-end installation position environment data is determined by the data sampling of the actual product installation environment temperature, humidity and air particles and combining with the specific building and space use position names determined during the installation of a specific front-end app system.
The front end is triggered and early-warned due to smoke or dust, and is judged by various other conditions such as temperature, humidity and the like, so that the front end is not a fire alarm, and belongs to the installation and application environment problem. For example, after smoke early warning is triggered twice in a week in the actual installation environment of a kitchen or a chess and card room, the system and the equipment can comprehensively evaluate the sensitivity threshold value of the environment smoke detection and perform special individual adjustment on the automatic early warning threshold value.
After twice or three frequent early warning triggers and detecting equipment and a system, the system can make targeted improvement adjustment on the smoke concentration early warning threshold value of the composite fire detector according to the characteristic, so that the situation that the electric quantity of the composite fire detector is consumed due to the fact that the installation environment is frequently awakened and the electric quantity of the composite fire detector is prevented from being excessively consumed due to the fact that frequent early warning is caused, and the user generates habituation to fire warning behaviors, so that paralytic consciousness is formed and fire warning importance is ignored.
After the non-fire early warning judgment, the equipment local photoelectric early warning is automatically carried out, and then the non-fire alarm is judged to automatically eliminate the early warning sound, so that the fire alarm is precisely timed and linked to be pushed in multiple ways, the early warning is carried out, and the non-fire alarm is judged to be timely and automatically background to cancel the early warning, so that the user is not continuously disturbed, and other staff are not disturbed, so that the false alarm is formed.
By respectively initializing smoke feeling and cooperatively arranging a humidity sensor, a carbon monoxide sensor and a temperature sensor, an initialized sample data set A, B, C is obtained based on the three; and constructing a BP neural network, taking the sets A and B as input samples, taking the set C as a verification set, training the BP neural network, after training, putting a humidity-sensitive sensor in the smoke sensor, integrally setting the two sensors, acquiring the data of the smoke sensor and the humidity-sensitive sensor in real time, and outputting the BP neural network after the training, and alarming or continuously monitoring based on an output result.
Referring to fig. 5-6, by recording initialization data at the time of installation of the smoke sensing device, storing the initialization data into a system database, wherein the initialization sample data comprises address data of an installation position of the device and environment data of the installation position;
the position data comprises detailed information of province, city, district, street, detailed address, building, floor unit, installation position and the like which are accurate to the installation direction of the room number, and the position data provides help for data comparison and analysis of peripheral equipment when the equipment alarms; and automatically positioning longitude and latitude data information of equipment through an installed position system.
The environmental data includes: smoke concentration, temperature sensing, air humidity;
when the equipment automatically triggers an alarm, reporting environmental data for 6 times continuously, and starting from the triggering of the alarm, comparing the environmental data with a threshold value for triggering a fire condition through differencing comparison analysis of three groups of data of a smoke sensor, a temperature sensor and a humidity sensor for 6 times continuously for 5S/times, and judging whether the alarm of the equipment belongs to false alarm or not; the data principle is as follows:
data 1: comparing the 6 smoke concentration change data to analyze whether the smoke concentration change data continuously rises or not, and judging the possibility of fire disaster;
data 2: comparing the temperature data uploaded for 6 times, and changing whether continuous temperature rise exists or not; if the temperature is kept unchanged, false alarm caused by shielding of the sensor due to dust, winged insects and the like is possible;
data 3: comparing the change of the air humidity in the 6 times of data, and judging the possibility of fire disaster by calculating the decrease data of the humidity;
according to the method, through big data summarization, initial data installed by all fire smoke detection alarms in a region of longitude and latitude are summarized to an Anxiao integrated intelligent cabinet system database for storage and archiving; after the fire smoke detection alarm is installed and activated, heartbeat packet data detection uploading is carried out every 24 hours in a standby state, and the data comprises: the smoke concentration, the detection temperature and the air humidity can be used for synchronously comparing data with the peripheral equipment to judge whether calculation belongs to the false alarm condition when the smoke sensing equipment alarms.
When the 4G network is used for deployment connection of the fire detector, when the composite intelligent fire detection alarm detects smoke rising to trigger early warning, the smoke concentration, air humidity, environment temperature and temperature change condition in a maze are continuously detected for 6 times and 5S/times in 30S, data are transmitted to an Internet of things platform through 4G signals, the platform calculates and deduces 4 groups of data through an algorithm, and output results are synchronously pushed to a fire management platform, a user mobile phone end and a control-elimination duty room through wireless transmission; the intelligent security and fire-extinguishing integrated guard cabinet screen displays fire alarm popup windows and transmits alarm information to the fire-extinguishing control host, and the fire-extinguishing control host transmits related dynamic instruction orders such as fire alarm linkage starting for emergency evacuation, smoke prevention and fire control water supply pressurization, spray fire extinguishing and the like.
When the wired local area network is used for deployment, when the compound intelligent fire detection alarm detects smoke rising triggering early warning, the smoke rising triggering early warning is continuously performed for 6 times in 30S, the smoke concentration, the air humidity, the ambient temperature and the temperature change condition in a maze are continuously detected, data are transmitted to an Internet of things platform through a LORA host local area network, the platform calculates and deduces 4 groups of data through an algorithm, and an output result is synchronously pushed to a fire management platform, a user mobile phone end and a control-elimination duty room through wireless transmission; the intelligent security and fire-extinguishing integrated guard cabinet screen displays fire alarm popup windows and transmits alarm information to the fire-extinguishing control host, and the fire-extinguishing control host transmits related dynamic instructions such as fire alarm linkage starting, emergency evacuation, smoke prevention, fire water supply pressurization, spray fire extinguishing and the like.
The principle of the steps is as follows:
step 1: the detection equipment gives early warning, and three groups of data including smoke concentration, temperature and humidity are uploaded;
step 2: comparing the data difference between three groups of data of the early warning device and the data uploaded by the peripheral devices at the same position (the position comparison sequence comprises the same region, the same room number, the same layer and the same span, and the comparison is carried out from thin to wide) and judging whether the data reported by the early warning device exceeds a fire alarm threshold value or not through calculation, outputting a primary judgment result, and judging that the fire is suspected if the data exceeding the fire alarm threshold value is more;
step 3: continuously uploading data, triggering an alarm by the equipment, continuously reporting the data of fog concentration, temperature and humidity for 5 times every 5 seconds within 30 seconds, and calculating whether the continuously uploaded data value of the alarm equipment continuously rises;
step 4: outputting a calculation result 1, wherein various data have no obvious change, and the system judges that a larger probability belongs to equipment false alarm caused by external uncertain factors, and only performs early warning push/false alarm push on a user terminal mobile phone APP, does not make telephone and short message notification and does not push to a management terminal user;
step 5: the calculation result 2 is output, the fluctuation of each item of data is obvious, and each item of data index continuously rises, the system judges that the larger probability is fire alarm, and the system pushes fire alarm information to each management terminal: cell-phone APP (user side, management end), integrative intelligent guard cabinet, intelligent miniature rescue station of safety and decontamination, fire alarm information contains: alarm device ID, alarm time, alarm location, device contact phone.
Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made therein without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.
Claims (10)
1. The utility model provides a fire detection alarm based on thing networking environment sampling intelligent algorithm, includes mounting base (110), its characterized in that, mounting base (110) bottom fixedly connected with bottom (090), bottom (090) internally mounted has lithium manganese battery (100), lithium manganese battery (100) with install battery clamping spring (130) between bottom (090) inner wall, battery clamping button (120) are installed to bottom (090) inside one side, PCB subassembly (080) are installed to bottom (090) bottom, PCB subassembly (080) with install reset and amortization button (140) between bottom (090), mounting base (110) bottom edge fixedly connected with circular main part (020), circular main part (020) bottom is installed and is detained (010) in the response, circular main part (020) with install anti-fake mechanism and switch (150) between last response knot (010), circular main part (040) internally mounted has compound maze (040), compound maze (040) outside detachably connects with reset and amortization button (040), install the top (060) and tear down, install the top (060) maze (060), the track cover (070) is located at the bottom of the PCB assembly (080).
2. The fire detection alarm based on the internet of things environment sampling intelligent algorithm according to claim 1, wherein the PCB assembly (080) comprises a PCB board (081), the top of the PCB board (081) is installed at the bottom of the bottom cover (090), a 4g sim card slot seat (088) is installed at the top of the PCB board (081), two red and blue double light emitting tubes (085) are installed at the middle side of the bottom of the PCB board (081), a shielding cover (087) is installed at one side of the bottom of the PCB board (081), a humidity sensitive detector (082) is installed at the other side of the bottom of the PCB board (081), a glass head NTC thermistor one (083) is installed at one side of the edge of the PCB board (081), a glass head NTC thermistor two (084) is installed at the other side of the edge of the PCB board (081), and a carbon monoxide detector (086) is installed at one side of the bottom of the PCB board (081) close to the shielding cover (087).
3. The fire detection alarm system method based on the intelligent sampling algorithm of the environment of the Internet of things is characterized by comprising the following specific steps of:
initializing smoke feeling to obtain an initialized sample data set A;
step two, setting a humidity sensor and a carbon monoxide sensor in cooperation with smoke feeling to obtain an initialized sample data set B;
step three, setting a temperature sensor in cooperation with smoke feeling to obtain an initialized sample data set C;
step four, constructing a BP neural network, taking the sets A and B as input samples, taking the set C as a verification set, and training the BP neural network;
fifthly, placing a humidity sensor in the smoke sense;
step six, acquiring data of the smoke sensor and the humidity sensor in real time, and outputting the data based on the trained BP neural network;
step seven, alarming or continuously monitoring based on the output result;
and step eight, when the installation application of the specific position is carried out, the detector can adapt to and adjust the corresponding early warning value according to the actual installation application environment.
4. The fire detection alarm system method based on the intelligent sampling algorithm of the environment of the internet of things according to claim 3, wherein the temperature sensor in the third step is an NTC sensor.
5. The fire detection alarm system method based on the intelligent sampling algorithm of the environment of the internet of things according to claim 3, wherein the BP neural network in the fourth step is a BP neural network based on a genetic algorithm.
6. The fire detection alarm system method based on the intelligent sampling algorithm of the environment of the Internet of things according to claim 3, wherein the construction of the BP neural network based on the genetic algorithm in the fourth step comprises the following steps of;
s1, initializing a population N;
s2, determining an adaptability function and control parameters;
s3, selecting, crossing and mutating operations are carried out, and a result is fed back to the BP neural network;
s4, if the optimal weight estimation is obtained, constructing a BP neural network based on a genetic algorithm, otherwise, returning to the step S3.
7. The fire detection alarm system method based on the intelligent sampling algorithm of the environment of the internet of things according to claim 6, wherein the function in S2 isWherein e is the maximum estimation error of the improved BP neural network, Y i For actual output, C i An output is desired.
8. The fire detection alarm system method based on the intelligent sampling algorithm of the environment of the internet of things according to claim 6, wherein the error function of the BP neural network comprises a penalty term.
9. The method of fire detection alarm system based on the intelligent sampling algorithm of the environment of the internet of things according to claim 3, wherein in the seventh step, when the output result is that the water vapor is too much, judgment is made to cause photoelectric early warning and temperature change to be that the installation adaptation environment is water vapor triggering alarm, and the fire is not caused, the judgment result is output, and the detector is reminded of needing manual dehumidification drying treatment.
10. The fire detection alarm system method based on the intelligent sampling algorithm of the environment of the internet of things according to claim 3, wherein the self-adapting and self-adjusting corresponding early warning values in the step eight comprises the following steps:
i, through data sampling of actual product installation environment temperature, humidity and air particles, combining the specific building and space use position names determined during the installation of a specific front-end app system;
II, after the detector triggers early warning due to smoke or dust, the detector is judged to be not a fire alarm but an installation and application environment after being judged by other various conditions such as temperature, humidity and the like;
III, after the detector triggers the early warning in the step II for a plurality of times within a certain period of time, the system and the equipment can comprehensively evaluate the sensitivity threshold value of the environmental smoke detection and carry out special individual adjustment on the automatic early warning threshold value;
IV, comprehensively adjusting the punishment early warning threshold value of the application installation environment according to early warning times and time frequency to avoid excessive early warning consuming the electric quantity of the composite fire detector and causing frequent early warning to cause habituation of multiple fire warning behaviors of a user, so that paralytic consciousness is formed and the importance of fire warning is ignored;
v, after non-fire early warning judgment, equipment local photoelectric early warning is automatically carried out, and then non-fire alarm automatic elimination early warning sound is judged, so that fire alarm accurate timing linkage multiparty pushing is realized, after early warning, non-fire alarm timely automatic background silencing cancellation early warning can not continuously disturb a user, and other staff can not be disturbed, so that false alarm is formed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311426807.7A CN117542174A (en) | 2023-10-30 | 2023-10-30 | Fire detection alarm based on Internet of things environment sampling intelligent algorithm and system thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311426807.7A CN117542174A (en) | 2023-10-30 | 2023-10-30 | Fire detection alarm based on Internet of things environment sampling intelligent algorithm and system thereof |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117542174A true CN117542174A (en) | 2024-02-09 |
Family
ID=89794950
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311426807.7A Pending CN117542174A (en) | 2023-10-30 | 2023-10-30 | Fire detection alarm based on Internet of things environment sampling intelligent algorithm and system thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117542174A (en) |
-
2023
- 2023-10-30 CN CN202311426807.7A patent/CN117542174A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP0877995B1 (en) | Method for dynamically adjusting fire detection criteria | |
| US20220044140A1 (en) | Event condition detection | |
| CN100390827C (en) | Fire and smoke detection and controlling system thereof | |
| JP6507155B2 (en) | Smart home hazard detector providing situation specific features and / or pre-alarm configuration | |
| US6166647A (en) | Fire detector | |
| EP2320397B1 (en) | Fire sensor and method for detecting fire | |
| US20150077737A1 (en) | System and methods for monitoring an environment | |
| CN209821986U (en) | Fire monitoring system | |
| CA3210157A1 (en) | Sensor fusion for fire detection & air quality monitoring | |
| JPS6325398B2 (en) | ||
| CN109166266A (en) | Intelligent fire alarm system and method | |
| WO1999022351A1 (en) | Electrical current saving combined smoke and fire detector | |
| CN207097187U (en) | A kind of intelligent free-standing smoke sensitive fire hazard detection alarm | |
| CN212256511U (en) | A mixed detector of temperature sensing smoke for detecting conflagration | |
| KR200470830Y1 (en) | Ceiling type sensor case having sensor intergration board | |
| CN117542174A (en) | Fire detection alarm based on Internet of things environment sampling intelligent algorithm and system thereof | |
| CN107437320A (en) | A kind of intelligent free-standing smoke sensitive fire hazard detection alarm and its alarm method | |
| CN210039021U (en) | Fire alarm system | |
| CN116246410A (en) | Fire monitoring method, system and storage medium based on smoke alarm | |
| CN209149522U (en) | Pyrotechnics identifies alarm system | |
| CN118038620B (en) | Intelligent fire safety monitoring system and method | |
| US11519622B2 (en) | HVAC monitoring system | |
| Qiyuan et al. | Experimental analysis on false alarms of fire detectors by cooking fumes | |
| KR200343946Y1 (en) | Fire alarm air cleaner assembly | |
| KR20240080008A (en) | Fire detection system comprising preventing from unwanted alarm and enhancing the fire adaptability for smoke detector |
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 |