CN213424156U - Fire prediction device - Google Patents
Fire prediction device Download PDFInfo
- Publication number
- CN213424156U CN213424156U CN202022360296.1U CN202022360296U CN213424156U CN 213424156 U CN213424156 U CN 213424156U CN 202022360296 U CN202022360296 U CN 202022360296U CN 213424156 U CN213424156 U CN 213424156U
- Authority
- CN
- China
- Prior art keywords
- sensor
- data
- coordinator
- environment
- fire
- 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.)
- Active
Links
Images
Landscapes
- Fire Alarms (AREA)
- Alarm Systems (AREA)
Abstract
The utility model provides a conflagration prediction device, include: the system comprises at least one environment sensor, at least one image sensor, at least one terminal node, a coordinator and an upper computer; the environment sensor and the image sensor are electrically connected with the terminal nodes in a one-to-one correspondence manner, the terminal nodes are in communication connection with the coordinator, the coordinator is in communication connection with the upper computer, and the output end of the upper computer is connected with the alarm; through setting up environmental sensor and image sensor, make the two of image data and the on-the-spot physical parameter of traditional environmental sensor measurement conflagration combine, realize the complementation of sensor data, restore the characteristic of scene from a plurality of dimensions to realize accurate high-efficient ground environmental fire's control prediction.
Description
Technical Field
The utility model relates to a conflagration prediction field, concretely relates to conflagration prediction device.
Background
A fire is a very common disaster. Along with the continuous development of society and the continuous increase of social wealth, the occurrence of fire often brings huge wealth loss, and moreover, the occurrence of fire also seriously threatens the life safety of people and influences the stable and prosperous development of society. The earlier and more accurate detection of a fire minimizes various losses to the home.
The existing fire prediction system is mostly based on the discrimination of a single sensor, no reliable judgment basis exists on data, the fire prediction accuracy rate cannot meet the requirements of people, and although a fire prediction device based on multiple sensors is also arranged at present, the multiple sensors are simply spliced and cannot accurately and efficiently monitor and predict the environmental fire.
SUMMERY OF THE UTILITY MODEL
In order to overcome the defect of the prior art, the utility model aims to solve the technical problem that a conflagration prediction device is proposed, through setting up environmental sensor and image sensor, do the combination with the two of the on-the-spot physical parameter of image data and traditional environmental sensor measurement conflagration, realize sensor data's complementation, restore the characteristic of scene from a plurality of dimensions to realize the control prediction of accurate high-efficient ground environment condition of a fire.
To achieve the purpose, the utility model adopts the following technical proposal:
the utility model provides a pair of conflagration prediction device, include: the system comprises at least one environment sensor, at least one image sensor, at least one terminal node, a coordinator and an upper computer; the environment sensor is used for collecting environment data, and the image sensor is used for collecting image data which are generated to the terminal node; the terminal node is used for receiving the environment data and sending the environment data to the coordinator; the coordinator generates packed data according to the environment data of the receiving terminal nodes and the environment data and sends the packed data to the upper computer; the upper computer is used for monitoring and receiving the packed data of the coordinator, generating a decision signal according to the packed data and sending the decision signal to the alarm; the environment sensor and the image sensor are electrically connected with the terminal nodes in a one-to-one correspondence mode, the terminal nodes are in communication connection with the coordinator, the coordinator is in communication connection with the upper computer, and the output end of the upper computer is connected with the alarm.
Preferably, the environment sensor further comprises a filter circuit, the environment sensor is electrically connected with an input end of the filter circuit, and an output end of the filter circuit is electrically connected with an input end of the terminal node.
Preferably, the upper computer comprises a convolutional neural network trainer and a monitor; the convolutional neural network trainer is used for performing characteristic extraction data splicing on the environmental data and the image data in the packed data to generate a fire probability and transmitting the fire probability to the monitor; the monitor is used for receiving the fire probability and displaying the fire probability; the convolutional neural network trainer is in communication connection with the coordinator, and is connected with the monitor.
Preferably, the environment sensor comprises a temperature sensor, a humidity sensor, a smoke sensor, a flame sensor and a CO concentration sensor, and the temperature sensor, the humidity sensor, the smoke sensor, the flame sensor and the CO concentration sensor are all electrically connected with the terminal node.
Preferably, the terminal nodes and the coordinator are CC2530 terminal nodes and CC2530 coordinator.
Preferably, the temperature sensor is a DS18B20 sensor.
Preferably, the humidity sensor is an SHT20 sensor.
Preferably, the smoke sensor is an MQ-2 smoke gas sensor.
Preferably, the flame sensor is an infrared light receiving sensor.
Preferably, the CO concentration sensor is a MS2200 carbon monoxide sensor.
Preferably, the image sensor is an AI-SU500C color CCD camera.
The utility model has the advantages that:
the utility model provides a fire prediction device, gather data through setting up two kinds of sensors of environmental sensor and image sensor, environmental sensor mainly used gathers environmental data such as environment temperature, humidity, smog, flame, CO concentration in the environment, and image sensor mainly used gathers the image data of environment, and pass environmental data and image data transmission to the host computer in, the host computer carries out the concatenation of feature extraction data through the convolutional neural network training ware and produces the conflagration probability, wherein set up a plurality of terminal nodes, each terminal node all is connected with temperature sensor, humidity transducer, the smog sensor, the flame sensor, CO concentration sensor and image sensor, combine the advantage of a plurality of terminal nodes, can acquire environmental data and image data from a plurality of angles, produce the conflagration probability after the concatenation of feature extraction data of rethread host computer, the accuracy of fire identification is greatly improved, and the real-time and reliable alarm effect of early fire is realized.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without inventive exercise.
Fig. 1 is a schematic structural diagram provided in the embodiment of the present invention.
The attached drawings are as follows: 1 an environmental sensor; 2 an image sensor; 3, a terminal node; 4, a coordinator; 5, an upper computer; 6, an alarm; 7 a filter circuit.
Detailed Description
The technical solutions in the embodiments of the present invention will be described clearly and completely with reference to the accompanying drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments in the present invention, all other embodiments obtained by a person skilled in the art without creative efforts belong to the protection scope of the present invention.
The technical solution of the present invention is further explained by the following embodiments with reference to the accompanying drawings.
As shown in fig. 1, the fire prediction apparatus provided in the present embodiment includes: the system comprises at least one environment sensor 1, at least one image sensor 2, at least one terminal node 3, a coordinator 4 and an upper computer 5; the environment sensor 1 is used for collecting environment data, the image sensor 2 is used for collecting image data, and the environment data and the image data are all generated to the terminal node 3; the terminal node 3 is used for receiving the environment data and sending the environment data to the coordinator 4; the coordinator 4 generates packed data according to the environment data and the environment data of the receiving terminal node 3 and sends the packed data to the upper computer 5; the upper computer 5 is used for monitoring and receiving the packed data of the coordinator 4, generating a decision signal according to the packed data and sending the decision signal to the alarm 6; the environment sensor 1 and the image sensor 2 are electrically connected with the terminal nodes 3 in a one-to-one correspondence mode, the terminal nodes 3 are connected with the coordinator 4 in a communication mode, the coordinator 4 is connected with the upper computer 5 in a communication mode, and the output end of the upper computer 5 is connected with the alarm 6. Specifically, the environment sensor 1 and the image sensor 2 are electrically connected with a terminal node 3, the terminal node 3 and the coordinator 4 can be electrically connected and can also be connected by a wireless network to enable the terminal node and the coordinator to communicate, the coordinator 4 and the upper computer 5 can be electrically connected and can also be connected by a wireless network to enable the coordinator to communicate, and the upper computer 5 is electrically connected with the alarm 6; in this embodiment, be provided with two terminal nodes 3, every terminal node 3 can place in a plurality of different positions, and every terminal node 3 all is connected with environmental sensor 1 and image sensor 2, gathers environmental data and image data through terminal node 3, and unified transmission is carried out data interaction by coordinator 4 in host computer 5 to coordinator 4, and host computer 5 carries out analysis processes after receiving data and generates the decision-making model, and control alarm 6 reports to the police.
Preferably, the environment sensor device further comprises a filter circuit 7, wherein the environment sensor 1 is electrically connected with an input end of the filter circuit 7, and an output end of the filter circuit 7 is electrically connected with an input end of the terminal node 3.
Preferably, the environment sensor 1 includes a temperature sensor, a humidity sensor, a smoke sensor, a flame sensor, and a CO concentration sensor, and the temperature sensor, the humidity sensor, the smoke sensor, the flame sensor, and the CO concentration sensor are all electrically connected to the terminal node 3. Specifically, every terminal node 3 is connected with temperature sensor, humidity transducer, smoke transducer, flame sensor, CO concentration sensor and image sensor 2, after the terminal node 3 board carried the data that the ADC module was used for gathering corresponding sensor, through IIC, the USART acquires relevant sensor data, carry out hardware filtering to analog voltage data through filter circuit 7, and then the data of gathering terminal node 3 are after the median is filtered, obtain real-time environmental data, terminal node 3 passes through wireless transmission to coordinator 4, coordinator 4 changes the USB agreement through TTL, reach host computer 5 with data, host computer 5 integrates data.
Preferably, the upper computer 5 comprises a convolutional neural network trainer and a monitor; the convolutional neural network trainer is used for performing characteristic extraction data splicing on the environmental data and the image data in the packed data to generate a fire probability and transmitting the fire probability to the monitor; the monitor is used for receiving the fire probability and displaying the fire probability; the convolutional neural network trainer is in communication connection with the coordinator 4 and is connected with the monitor. Specifically, the process of the upper computer 5 receiving the data of the coordinator 4 for integration includes firstly, the convolutional neural network trainer of the upper computer 5 splices the data collected by the environment sensor 1 into a one-dimensional feature vector S1, performs noise reduction filtering on the color image collected by the image sensor 2 to obtain final data of the image, performs deep learning feature extraction on the image data to obtain an image convolution feature vector, passes through a full connection layer, and maps the multidimensional image convolution feature vector to the one-dimensional feature vector S2. Further, the data of the environment sensor 1 and the one-dimensional feature vector S1 extracted by the convolutional neural network are spliced to obtain a one-dimensional feature vector S3, and particularly, in order to ensure that the decisive action of each sensing data on the network is close, normalization processing can be preferentially considered on the features S2 and S1 respectively before data splicing. Further, a full connection relation is established between the spliced characteristic vector S3 and the output result, in order to obtain the probability value of the fire, the output result is processed by adopting an activation function, generally, the activation function adopts a Sigmoid function, finally, the probability value of the fire is obtained and sent to a monitor, the monitor monitors and displays the probability value of the fire, and when the probability value of the fire exceeds a preset value, a decision signal is sent to the alarm 6 for alarm prompt.
The above process is a prediction process of a single scene, also called a forward propagation process of a network. For multiple scenes, the process is similar and is not described in detail. The fire prediction device disclosed by the application further comprises a training link, namely a network back propagation process. Sufficient sample data needs to be collected prior to training. The sample collection procedure was as follows: firstly, various sensor data of a single scene are collected, and the fire condition is manually calibrated, and the fire condition is divided into two conditions of fire occurrence and fire non-occurrence. Changing scene environment, such as increasing pedestrian walking links, or changing physical parameters of ambient light, temperature, humidity and the like, or intentionally igniting a certain object under the condition of ensuring safety to cause the condition similar to fire and the like, repeatedly collecting sample data of various conditions to obtain a sample set, extracting various data, and according to the following steps of 7: 3 into training and test sets. And for the loss function in the training process, adopting a cross entropy loss function, solving the cross entropy of the prediction result and the label value, and reversely updating the weight values of each layer by adopting an Adam optimization algorithm according to the set learning rate.
Preferably, the terminal node 3 and coordinator 4 are CC2530 terminal node 3 and CC2530 coordinator 4.
Preferably, the temperature sensor is a DS18B20 sensor.
Preferably, the humidity sensor is an SHT20 sensor.
Preferably, the smoke sensor is an MQ-2 smoke gas sensor.
Preferably, the flame sensor is an infrared light receiving sensor.
Preferably, the CO concentration sensor is a MS2200 carbon monoxide sensor.
Preferably, the image sensor 2 is an AI-SU500C color CCD camera.
The sensors can also be the same type of sensors which can achieve similar acquisition effects.
The above description is only a preferred embodiment of the present invention, and should not be taken as limiting the invention, and any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. A fire prediction apparatus, comprising: the system comprises at least one environment sensor (1), at least one image sensor (2), at least one terminal node (3), a coordinator (4) and an upper computer (5);
the environment sensor (1) is used for collecting environment data, the image sensor (2) is used for collecting image data, and the environment data and the image data are all generated to the terminal node (3);
the terminal node (3) is used for receiving environment data and sending the environment data to the coordinator (4);
the coordinator (4) generates packed data according to the environment data of the receiving terminal node (3) and the environment data, and sends the packed data to the upper computer (5);
the upper computer (5) is used for monitoring and receiving the packed data of the coordinator (4), generating a decision signal according to the packed data and sending the decision signal to the alarm (6);
environmental sensor (1) with image sensor (2) all with terminal node (3) one-to-one electricity is connected, terminal node (3) all with coordinator (4) communication is connected, coordinator (4) with host computer (5) communication is connected, the output of host computer (5) with alarm (6) are connected.
2. A fire prediction device as defined in claim 1, wherein: the environment sensor is characterized by further comprising a filter circuit (7), the environment sensor (1) is electrically connected with the input end of the filter circuit (7), and the output end of the filter circuit (7) is electrically connected with the input end of the terminal node (3).
3. A fire prediction device as defined in claim 1, wherein: the upper computer (5) comprises a convolutional neural network trainer and a monitor;
the convolutional neural network trainer is used for generating fire probability after feature extraction data splicing is carried out on the environmental data and the image data in the packed data, and transmitting the fire probability to the monitor;
the monitor is used for receiving the fire probability and displaying the fire probability;
the convolutional neural network trainer is in communication connection with the coordinator (4), and is connected with the monitor.
4. A fire prediction device as defined in claim 1, wherein: the environment sensor (1) comprises a temperature sensor, a humidity sensor, a smoke sensor, a flame sensor and a CO concentration sensor, and the temperature sensor, the humidity sensor, the smoke sensor, the flame sensor and the CO concentration sensor are all electrically connected with the terminal node (3).
5. A fire prediction device as defined in claim 4, wherein: the temperature sensor is a DS18B20 sensor.
6. A fire prediction device as defined in claim 4, wherein: the terminal node (3) and the coordinator (4) are a CC2530 terminal node (3) and a CC2530 coordinator (4).
7. A fire prediction device as defined in claim 4, wherein: the smoke sensor is an MQ-2 smoke gas-sensitive sensor.
8. A fire prediction device as defined in claim 4, wherein: the flame sensor is an infrared light receiving sensor.
9. A fire prediction device as defined in claim 4, wherein: the CO concentration sensor is an MS2200 carbon monoxide sensor.
10. A fire prediction device as defined in claim 4, wherein: the image sensor (2) is an AI-SU500C color CCD camera.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202022360296.1U CN213424156U (en) | 2020-10-21 | 2020-10-21 | Fire prediction device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202022360296.1U CN213424156U (en) | 2020-10-21 | 2020-10-21 | Fire prediction device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN213424156U true CN213424156U (en) | 2021-06-11 |
Family
ID=76268378
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202022360296.1U Active CN213424156U (en) | 2020-10-21 | 2020-10-21 | Fire prediction device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN213424156U (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115273385A (en) * | 2022-07-11 | 2022-11-01 | 杭州海康威视数字技术股份有限公司 | Camera for flame detection |
-
2020
- 2020-10-21 CN CN202022360296.1U patent/CN213424156U/en active Active
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115273385A (en) * | 2022-07-11 | 2022-11-01 | 杭州海康威视数字技术股份有限公司 | Camera for flame detection |
| CN115273385B (en) * | 2022-07-11 | 2024-03-26 | 杭州海康威视数字技术股份有限公司 | A camera for flame detection |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104966375A (en) | Security monitoring system and monitoring method | |
| CN106652285A (en) | Distributed multi-defence-area vibration optical fiber perimeter alarm system and perimeter monitoring method | |
| AU2021100365A4 (en) | A multi-sensor-based intelligent monitoring and early warning system and method for dam safety | |
| CN102208018A (en) | Method for recognizing fire disaster of power transmission line based on video variance analysis | |
| CN213424156U (en) | Fire prediction device | |
| CN117319451B (en) | Urban fire-fighting Internet of things supervision system based on multi-mode big data and method thereof | |
| CN113192283B (en) | Wireless fire early warning system with multi-sensor information fusion | |
| CN113160513A (en) | Flame detection device based on multisensor | |
| CN104240418B (en) | A kind of signal processing method and warning device | |
| CN201091014Y (en) | Fire detecting device | |
| CN108010254A (en) | One kind is based on four wave band infrared flame detectors and its flame identification algorithm | |
| CN114460080A (en) | Rice disease and pest intelligent monitoring system | |
| CN112785803A (en) | Monitoring system based on Internet of things | |
| CN110930632B (en) | Early warning system based on artificial intelligence | |
| CN106548592A (en) | A kind of Household security system based on Internet of Things | |
| CN111986436A (en) | Comprehensive flame detection method based on ultraviolet and deep neural networks | |
| CN210271155U (en) | Composite smoke-sensing detection labyrinth | |
| CN114255562A (en) | Wisdom fire control early warning system based on thing networking | |
| CN112947147A (en) | Fire-fighting robot based on multi-sensor and cloud platform algorithm | |
| CN116863629A (en) | Alarm device, alarm method and electronic equipment based on multiple sensors | |
| CN115931141A (en) | Temperature identification method of infrared temperature measurement map based on improved ANN algorithm | |
| CN212569977U (en) | Fire identification alarm device based on grey and RBF double-layer neural networks | |
| CN113553996A (en) | Dangerous person discovering system based on deep learning neural network emotion recognition | |
| CN205264012U (en) | High accuracy natural gas leakage detector based on thing networking | |
| CN210222823U (en) | Community population flow analysis system based on probe |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| GR01 | Patent grant | ||
| GR01 | Patent grant |