CN111047815A - Method and system for identifying false alarm of fire-fighting detector based on self-learning model - Google Patents
Method and system for identifying false alarm of fire-fighting detector based on self-learning model Download PDFInfo
- Publication number
- CN111047815A CN111047815A CN201911372561.3A CN201911372561A CN111047815A CN 111047815 A CN111047815 A CN 111047815A CN 201911372561 A CN201911372561 A CN 201911372561A CN 111047815 A CN111047815 A CN 111047815A
- Authority
- CN
- China
- Prior art keywords
- value
- fire
- detector
- alarm
- self
- 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
Images
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B17/00—Fire alarms; Alarms responsive to explosion
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B17/00—Fire alarms; Alarms responsive to explosion
- G08B17/10—Actuation by presence of smoke or gases, e.g. automatic alarm devices for analysing flowing fluid materials by the use of optical means
-
- 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
- G08B25/01—Alarm systems in which the location of the alarm condition is signalled to a central station, e.g. fire or police telegraphic systems characterised by the transmission medium
- G08B25/08—Alarm systems in which the location of the alarm condition is signalled to a central station, e.g. fire or police telegraphic systems characterised by the transmission medium using communication transmission lines
Landscapes
- Business, Economics & Management (AREA)
- Emergency Management (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Fire Alarms (AREA)
Abstract
The invention discloses a method and a system for identifying false alarm of a fire detector based on a self-learning model, belonging to a method for preventing false detection alarm, wherein the method comprises the following steps: a, a plurality of fire fighting detectors transmit the acquired current parameter values to a monitoring control center; b, adjusting a preset normal value by the monitoring control center through the area where the fire detector is located through self-learning, obtaining an acceleration sampling value and an alarm threshold value through the normal value, and transmitting the acceleration sampling value and the alarm threshold value to the fire detector; step C, the fire-fighting detector compares the current sampling parameter value with the acceleration sampling value; and D, when the parameter value sampled by the defense detector in the step C is continuously greater than the alarm threshold value for a plurality of times, the monitoring control center outputs an alarm signal. The normal value and the alarm value are adjusted in real time according to the installation area of the fire-fighting detector through the self-learning model, and the alarm signal is output after the multiple sampling and the alarm threshold value are compared, so that the situation of false alarm of the fire-fighting detector in the fire-fighting field can be effectively reduced and eliminated.
Description
Technical Field
The invention relates to a method for preventing false detection and alarm, in particular to a method for identifying false alarm of a fire detector based on a self-learning model.
Background
An automatic fire alarm system for fire control is an automatic fire-fighting equipment which is arranged in buildings and structures and is used for realizing early detection and alarm of fire, sending control signals to various fire-fighting equipment and realizing a preset fire-fighting function. The method comprises the following steps: 1) a fire detection alarm system; 2) a fire-fighting linkage control system; 3) a combustible gas detection alarm system; 4) an electrical fire control system. The principle is that when a fire disaster happens to a certain monitored site in a building or a structure, a fire detector detects signals such as smoke, high temperature, flame and gas specific to the fire disaster and converts the signals into electric signals to be transmitted to a fire alarm controller, the controller receives a fire alarm signal, the fire alarm signal is analyzed and compared with a normal state threshold value or parameter, and if the fire is confirmed, two loop signals are output: one path of the instruction alarm device gives out sound alarm, and the other path of the instruction starts the fire control equipment. The detection alarm system in the fire-fighting system most commonly comprises a temperature-sensitive detector and a temperature-sensitive detector. In general, after detectors such as smoke detectors, temperature detectors, manual alarms and the like give an alarm, an alarm signal is transmitted to a fire control room, and an operator on duty in the fire control room needs to perform alarm processing. It is often necessary to dispatch a person on duty or to notify a patrol security officer nearby to go to the fire scene to see if a confirmed alarm is true within minutes. For example, if the fire is a real fire, a fire emergency treatment process is started, if the fire is a real fire, the fire emergency treatment process is started. If the device is in a false alarm, the device needs to be silenced and the false alarm needs to be processed, so that the device is recovered to a normal working state. And making relevant records and reporting the processing conditions to the upper level. Therefore, the false alarm in the fire fighting field causes a plurality of reasons, and brings a great amount of waste of manpower, material resources and financial resources. And once the alarm occurs, the alarm cannot be processed, so that in case of a real fire alarm, the alarm cannot be processed in time, and property loss and even life danger of residents can be caused. Therefore, research and improvement on the identification method of the fire detector false alarm are needed.
Disclosure of Invention
One of the objectives of the present invention is to provide a method for identifying a false alarm of a fire detector based on a self-learning model, so as to solve the technical problems in the prior art that the false alarm of the fire detector is easy to occur, a large amount of manpower, material resources and financial resources are wasted, and a rescue opportunity of a real fire situation is missed.
In order to solve the technical problems, the invention adopts the following technical scheme:
the invention provides a method for identifying false alarm of a fire detector based on a self-learning model, which comprises the following steps:
a, a plurality of fire fighting detectors transmit the acquired current parameter values to a monitoring control center;
b, adjusting a preset normal value by the monitoring control center through the area where the fire detector is located through self-learning, obtaining an acceleration sampling value and an alarm threshold value through the normal value, and transmitting the acceleration sampling value and the alarm threshold value to the fire detector; the acceleration sampling value and the alarm threshold value are both larger than normal values;
the self-learning mode is to collect data collected by the plurality of fire fighting detectors in the step A in unit time; then preprocessing the collected data set to remove abnormal data; finally, calculating the average value of the data set without abnormal data, and taking the average value as the current normal value;
step C, the fire-fighting detector compares the currently sampled parameter value with the acceleration sampling value, judges whether the currently sampled parameter value is larger than the acceleration sampling value, if so, the parameter sampling period is changed into one half of the normal sampling period, and continues sampling; otherwise, keeping the current parameter value sampling period;
and D, when the parameter value sampled by the defense detector in the step C is continuously greater than the alarm threshold value for a plurality of times, the monitoring control center outputs an alarm signal.
Preferably, the further counting technical scheme is as follows: and E, when the parameter values sampled by the detector for multiple times in the step C are all larger than the acceleration sampling value and continuously rise, but do not exceed the alarm threshold value, the monitoring control center outputs an early warning signal.
The further counting technical scheme is as follows: and B, removing abnormal data in the data set in the step B by calculating the standard deviation of the data set, comparing each value in the data set to determine whether the value is more than three times of the standard deviation, and deleting the sample if the value is more than three times of the standard deviation.
The further counting technical scheme is as follows: the data set is a data set of at least 20 fire fighting probes, and each fire fighting probe is used for more than 50 times.
The further counting technical scheme is as follows: the alarm threshold is 4% above the normal value.
The further counting technical scheme is as follows: the fire detector is provided with a smoke sensor and a temperature sensor.
The invention provides a system for identifying false alarm of a fire detector based on a self-learning model, which comprises a plurality of fire detectors and a monitoring control center, wherein the fire detectors are connected to the monitoring control center, and the system comprises: the fire fighting detectors are used for transmitting the acquired current parameter values to the monitoring control center; the monitoring control center is used for adjusting a preset normal value through the area where the fire-fighting detector is located through self-learning, obtaining an acceleration sampling value and an alarm threshold value through the normal value, and transmitting the acceleration sampling value and the alarm threshold value to the fire-fighting detector; the acceleration sampling value and the alarm threshold value are both larger than normal values; the self-learning mode is to collect data collected by a plurality of fire fighting detectors in unit time; then preprocessing the collected data set to remove abnormal data; finally, calculating the average value of the data set without abnormal data, and taking the average value as the current normal value; the fire-fighting detector is also used for comparing the currently sampled parameter value with the acceleration sampling value, judging whether the currently sampled parameter value is larger than the acceleration sampling value, if so, changing the parameter sampling period to be one half of the normal sampling period, and continuing sampling; otherwise, keeping the current parameter value sampling period; and when the sampled parameter value is continuously greater than the alarm threshold value for a plurality of times, the monitoring control center outputs an alarm signal.
Preferably, the further technical scheme is as follows: the fire-fighting detector is provided with a smoke sensor and a temperature sensor; the data set is a data set of at least 20 parameter sensors and more than 50 times of each parameter sensor.
The further technical scheme is as follows: the fire-fighting detector is also used for outputting an early warning signal by the monitoring control center if the parameter values of multiple sampling of the fire-fighting detector are all larger than the acceleration sampling value and are continuously increased but do not exceed the alarm threshold value; the alarm threshold is 4% above the normal value.
The further technical scheme is as follows: the mode of removing the abnormal data in the data set by the monitoring control center is to calculate the standard deviation of the data set, then compare each value in the data set to determine whether the value is more than three times of the standard deviation, and finally delete the sample if the value is more than three times of the standard deviation.
Compared with the prior art, the invention has the following beneficial effects: the method for identifying the fire fighting detector by the aid of the self-learning model is realized on the basis of the self-learning model, can be realized by means of existing sensors and control modules, and is suitable for being applied to various fire fighting places.
Drawings
FIG. 1 is a flow chart of a method for illustrating one embodiment of the present invention;
FIG. 2 is a flow chart for illustrating self-learning in one embodiment of the present invention;
FIG. 3 is a system block diagram illustrating one embodiment of the invention;
FIG. 4 is a distribution plot illustrating data preprocessing in one embodiment of the invention.
Detailed Description
The invention is further elucidated with reference to the drawing.
Referring to fig. 1, an embodiment of the present invention is a method for identifying false alarms of a fire detector based on a self-learning model, in which a normal value is increased according to a specific environment of an installation place of the fire detector, for example, a region due to air quality, or a period of time due to construction, decoration, etc., which is very likely to cause false alarms if a fixed alarm value is used; the method comprises the following steps and is preferably performed in sequence:
s1, transmitting the acquired current parameter values to a monitoring control center by a plurality of fire-fighting detectors; the fire detector is provided with a smoke sensor and a temperature sensor, and can simultaneously acquire smoke concentration and temperature;
s2, the monitoring control center adjusts a preset normal value through self-learning in the area where the fire detector is located, obtains an acceleration sampling value and an alarm threshold value through the normal value, and transmits the acceleration sampling value and the alarm threshold value to the fire detector; the acceleration sampling value and the alarm threshold value are both larger than normal values; and the acceleration sampling value is arranged between the normal value and the alarm threshold value; and generally speaking, the alarm threshold needs to be 4% higher than the normal value.
Referring to fig. 2, the self-learning manner is to collect data collected by a plurality of fire detectors in step a in a unit time; then preprocessing the collected data set to remove abnormal data; finally, calculating the average value of the data set without abnormal data, and taking the average value as the current normal value; taking smoke detection as an example, the minimum alarm concentration is 4 points (i.e. 4%, that is, the normal concentration is 1% OBS/M, and the alarm concentration is 5% OBS/M) greater than the normal concentration, and the normal smoke concentration is increased or exceeds the minimum value in the above-mentioned area due to air quality or in some time due to construction, decoration, etc. Under the condition, the concentration value detected by the smoke detector of the area is learned to the normal concentration value of the area, and then a certain numerical value is added on the basis to be used as an alarm concentration value;
through the learning of the mode, the preset data of the fire detector is adjusted, so that the fire detector judges the smoke concentration and the temperature by using a new threshold value in the detection process, the possibility of generating false alarm by the system is reduced, and the next step is continuously executed on the basis again;
s3, the fire-fighting detector compares the current sampling parameter value with the acceleration sampling value, judges whether the current sampling parameter value is larger than the acceleration sampling value, if yes, the parameter sampling period is changed to one half of the normal sampling period, and continues sampling; otherwise, keeping the current parameter value sampling period; in this step, also taking smoke as an example, if the sampled concentration is greater than the threshold (W1 > Y1), the sampling period is half of the normal sampling period (denoted as T2, T2=1/2 × T1), and if the sampled concentration is less than the threshold (W1 < Y1), the sampling period is returned to T1 (T2 = T1);
step S4, when the parameter value sampled by the defense detector in the step S3 is continuously greater than the alarm threshold value for a plurality of times, the monitoring control center outputs an alarm signal;
in the above steps, when the parameter values sampled by the defense detector for multiple times in step S3 are all greater than the acceleration sample value and continuously increase, but do not exceed the alarm threshold, the monitoring control center outputs the early warning signal.
Based on the steps, under the condition that the sampled parameter value exceeds the accelerated sampling value, the sampling frequency is accelerated, the data adopted each time is compared with the alarm threshold value, if the sampled parameter value exceeds the alarm minimum value for a plurality of times, a fire alarm is normally reported (for example, the data collected for three times continuously exceed the alarm threshold value), and under the condition of accelerated collection, the data collected each time generally rises, but the data does not reach the alarm threshold value, a fire alarm early warning is reported to the upper management center (the concentration continuously rises, but the alarm threshold value is not reached);
and after normal fire alarm, continuously adopting the original normal value in a new sampling period, and reporting alarm cancellation information to the monitoring control center if the parameter value is determined to be less than the alarm threshold value after the parameter value is determined to be continuously maintained for a period of time.
In the embodiment, the normal value and the alarm value are adjusted in real time according to the installation area of the fire detector through the self-learning model, and the alarm signal is output after the multiple sampling and alarm threshold value comparison, so that the condition of false alarm of the fire detector in the fire field can be effectively reduced and eliminated, the waste of manpower, material resources and financial resources is avoided, and the delay of the rescue opportunity of real fire is avoided.
According to another embodiment of the present invention, in order to further improve the accuracy of the self-learning of the method, in the learning manner, the manner of removing the abnormal data in the data set may be to calculate the standard deviation of the data set, then compare each value in the data set to determine whether the value is greater than three times the standard deviation, and finally delete the sample if the value is greater than three times the standard deviation; in addition, in the process of sampling and constructing the data set, at least 20 fire detectors and more than 50 times of data sets of each fire detector are required to be collected so as to ensure the adaptability of the learning result and the installation environment of the fire detectors; usually, the reported parameter values collected by the devices in an area (usually, a building in the same region of the networking unit) are collected. Since the self-learning function is one of the keys for solving the technical problems, the self-learning function described in this embodiment is still summarized as follows by taking the smoke sensation as an example:
1) in the initial smoke sensation state, for example, a smoke sensation value is sampled every 5s (denoted as T1), the smoke sensation value and the date are recorded (denoted as W1 and D1), and the data are reported to a superior management center;
2) through learning in a section of the area where smoke is sensed, for example, concentration values (W1, D1) collected by a plurality of detectors in a building, and according to the concentration values collected by the area, a normal concentration value is adjusted, and the learning method is as follows:
a. data acquisition: collecting concentration values reported by devices in an area (generally, buildings in the same region of the networking unit), wherein each device is collected at least 50 times, the number of the devices is at least more than 20 (generally, 6 stories in a building, the number of the 1000 square meters of each floor of the building is about 300, and the installation distance required by relevant specifications is less than 15 m);
b. data preprocessing: assuming that the data is normally distributed, the anomaly-removed data (normally distributed data is almost entirely concentrated in the (μ -3 σ, μ +3 σ) interval, and the probability of exceeding this range is only less than 0.3%, as shown in fig. 4); adopting an algorithm to realize error data elimination, firstly calculating the standard deviation of a data column to be detected; then comparing whether each value of the data column is larger than 3 times of the standard deviation; finally deleting the sample which is larger than 3 times of the standard deviation;
in fig. 4, the dark regions are a range of values within less than one standard deviation from the mean, the range representing a percentage of 68% of the total values in the positive distribution, and the percentages within two standard deviations combining to 95% according to the positive distribution; ratios within three standard deviations combine to be 99%;
c. and (3) data analysis: after data preprocessing, rejecting abnormal data, and calculating an arithmetic mean value of the equipment data to be used as a reference value of a normal concentration value of smoke sensation in the area, namely a normal value called in the embodiment;
referring to fig. 3, another embodiment of the present invention is a system for identifying false alarm of fire detector based on self-learning model based on the above method, the system includes a plurality of fire detectors and a monitoring control center, the plurality of fire detectors are connected to the monitoring control center, wherein:
the fire fighting detectors are used for transmitting the acquired current parameter values to the monitoring control center;
the monitoring control center is used for adjusting a preset normal value through the area where the fire-fighting detector is located through self-learning, obtaining an acceleration sampling value and an alarm threshold value through the normal value, and transmitting the acceleration sampling value and the alarm threshold value to the fire-fighting detector; the acceleration sampling value and the alarm threshold value are both larger than normal values;
the self-learning mode is to collect data collected by a plurality of fire fighting detectors in unit time; then preprocessing the collected data set to remove abnormal data; finally, calculating the average value of the data set without abnormal data, and taking the average value as the current normal value;
the fire-fighting detector is also used for comparing the currently sampled parameter value with the acceleration sampling value, judging whether the currently sampled parameter value is larger than the acceleration sampling value, if so, changing the parameter sampling period to be one half of the normal sampling period, and continuing sampling; otherwise, keeping the current parameter value sampling period; and when the sampled parameter value is continuously greater than the alarm threshold value for a plurality of times, the monitoring control center outputs an alarm signal.
As with the method and embodiment described above, the fire detector in the system described above has a smoke sensor and a temperature sensor; and the data set used by the monitoring control center for self-learning is a data set with at least 20 parameter sensors and more than 50 times of each parameter sensor. In the embodiment, the fire detector is also used for outputting an early warning signal by the monitoring control center if the parameter values of multiple sampling of the fire detector are all larger than the acceleration sampling value and are continuously increased, but the parameter values do not exceed the alarm threshold value; the alarm threshold is 4% above the normal value. The mode of removing the abnormal data in the data set by the monitoring control center is to calculate the standard deviation of the data set, then compare each value in the data set to determine whether the value is more than three times of the standard deviation, and finally delete the sample if the value is more than three times of the standard deviation.
The system described in the above embodiment includes a monitoring control center and a smoke detection terminal, and the functions implemented by the two parts in the system are counted again as follows:
1) the monitoring control center:
a. collecting data reported by a terminal fire detector, carrying out normal distribution to remove abnormality, calculating an average value, and learning a normal value, an acceleration sampling value and an alarm threshold value of an area;
b. these results are sent to the smoke sensing equipment;
2) smoke detection terminal:
a. according to the collection period, the collected data is reported to the monitoring center
b. Receiving the learning result sent by the monitoring center, including the reference value of normal concentration value, the accelerated sampling concentration value and the alarm threshold value
c. Comparing the data acquired each time with a sampling acceleration value and an alarm threshold value, and if the data is greater than the sampling acceleration value, halving the sampling period and accelerating the sampling;
d. if the sampling value is larger than the acceleration sampling value but does not reach the alarm threshold value, reporting fire early warning information if the sampling value continuously rises three times;
e. if the sampling value is greater than the lower limit of the alarm threshold value, reporting fire alarm information for three consecutive times;
f. the sampling value is less than the alarm threshold value for three times continuously, if the sampling value is already alarmed, the alarm canceling information is sent
The self-learning model adopted by the invention can effectively eliminate false alarm, for example, smoke alarm caused by large ash layer on site, smoke alarm caused by dirty filter screen of smoke detector, alarm caused by large air quantity on site, and alarm scene caused by large oil smoke on site, can eliminate false alarm by base line learning adjustment and continuous data acquisition, and can bring extra benefit, for example, the manufacturer can be informed to maintain equipment by identifying the dirty condition of filter screen.
In addition to the foregoing, it should be noted that reference throughout this specification to "one embodiment," "another embodiment," "an embodiment," or the like, means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment described generally throughout this application. The appearances of the same phrase in various places in the specification are not necessarily all referring to the same embodiment. Further, when a particular feature, structure, or characteristic is described in connection with any embodiment, it is submitted that it is within the scope of the invention to effect such feature, structure, or characteristic in connection with other embodiments.
Although the invention has been described herein with reference to a number of illustrative embodiments thereof, it should be understood that numerous other modifications and embodiments can be devised by those skilled in the art that will fall within the spirit and scope of the principles of this disclosure. More specifically, various variations and modifications are possible in the component parts and/or arrangements of the subject combination arrangement within the scope of the disclosure, the drawings and the appended claims. In addition to variations and modifications in the component parts and/or arrangements, other uses will also be apparent to those skilled in the art.
Claims (10)
1. A method for identifying false alarm of a fire detector based on a self-learning model is characterized by comprising the following steps:
a, a plurality of fire fighting detectors transmit the acquired current parameter values to a monitoring control center;
b, adjusting a preset normal value by the monitoring control center through the area where the fire detector is located through self-learning, obtaining an acceleration sampling value and an alarm threshold value through the normal value, and transmitting the acceleration sampling value and the alarm threshold value to the fire detector; the acceleration sampling value and the alarm threshold value are both larger than normal values;
the self-learning mode is to collect data collected by the plurality of fire fighting detectors in the step A in unit time; then preprocessing the collected data set to remove abnormal data; finally, calculating the average value of the data set without abnormal data, and taking the average value as the current normal value;
step C, the fire-fighting detector compares the currently sampled parameter value with the acceleration sampling value, judges whether the currently sampled parameter value is larger than the acceleration sampling value, if so, the parameter sampling period is changed into one half of the normal sampling period, and continues sampling; otherwise, keeping the current parameter value sampling period;
and D, when the parameter value sampled by the defense detector in the step C is continuously greater than the alarm threshold value for a plurality of times, the monitoring control center outputs an alarm signal.
2. The self-learning model-based method for identifying fire detector false alarms according to claim 1, characterized in that: and E, when the parameter values sampled by the detector for multiple times in the step C are all larger than the acceleration sampling value and continuously rise, but do not exceed the alarm threshold value, the monitoring control center outputs an early warning signal.
3. The self-learning model-based method for identifying fire detector false alarms according to claim 1, characterized in that: and B, removing abnormal data in the data set in the step B by calculating the standard deviation of the data set, comparing each value in the data set to determine whether the value is more than three times of the standard deviation, and deleting the sample if the value is more than three times of the standard deviation.
4. The method for identifying false alarms of fire detectors based on a self-learning model according to claim 1 or 3, characterized in that: the data set is a data set of at least 20 fire fighting probes, and each fire fighting probe is used for more than 50 times.
5. The self-learning model-based method for identifying fire detector false alarms according to claim 1, characterized in that: the alarm threshold is 4% above the normal value.
6. The self-learning model-based method for identifying fire detector false alarms according to claim 1, characterized in that: the fire detector is provided with a smoke sensor and a temperature sensor.
7. The utility model provides a system for discerning fire control detector false positive based on self-learning model which characterized in that: the system comprises a plurality of fire-fighting detectors and a monitoring control center, wherein the fire-fighting detectors are connected to the monitoring control center, and the system comprises:
the fire fighting detectors are used for transmitting the acquired current parameter values to the monitoring control center;
the monitoring control center is used for adjusting a preset normal value through the area where the fire-fighting detector is located through self-learning, obtaining an acceleration sampling value and an alarm threshold value through the normal value, and transmitting the acceleration sampling value and the alarm threshold value to the fire-fighting detector; the acceleration sampling value and the alarm threshold value are both larger than normal values;
the self-learning mode is to collect data collected by a plurality of fire fighting detectors in unit time; then preprocessing the collected data set to remove abnormal data; finally, calculating the average value of the data set without abnormal data, and taking the average value as the current normal value;
the fire-fighting detector is also used for comparing the currently sampled parameter value with the acceleration sampling value, judging whether the currently sampled parameter value is larger than the acceleration sampling value, if so, changing the parameter sampling period to be one half of the normal sampling period, and continuing sampling; otherwise, keeping the current parameter value sampling period; and when the sampled parameter value is continuously greater than the alarm threshold value for a plurality of times, the monitoring control center outputs an alarm signal.
8. The self-learning model-based method for identifying fire detector false alarms of claim 7, wherein: the fire-fighting detector is provided with a smoke sensor and a temperature sensor; the data set is a data set of at least 20 parameter sensors and more than 50 times of each parameter sensor.
9. The self-learning model-based fire detector false alarm identification system of claim 7, wherein: the fire-fighting detector is also used for outputting an early warning signal by the monitoring control center if the parameter values of multiple sampling of the fire-fighting detector are all larger than the acceleration sampling value and are continuously increased but do not exceed the alarm threshold value; the alarm threshold is 4% above the normal value.
10. The self-learning model-based fire detector false alarm identification system of claim 7, wherein: the mode of removing the abnormal data in the data set by the monitoring control center is to calculate the standard deviation of the data set, then compare each value in the data set to determine whether the value is more than three times of the standard deviation, and finally delete the sample if the value is more than three times of the standard deviation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911372561.3A CN111047815A (en) | 2019-12-27 | 2019-12-27 | Method and system for identifying false alarm of fire-fighting detector based on self-learning model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911372561.3A CN111047815A (en) | 2019-12-27 | 2019-12-27 | Method and system for identifying false alarm of fire-fighting detector based on self-learning model |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111047815A true CN111047815A (en) | 2020-04-21 |
Family
ID=70239268
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911372561.3A Pending CN111047815A (en) | 2019-12-27 | 2019-12-27 | Method and system for identifying false alarm of fire-fighting detector based on self-learning model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111047815A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111613037A (en) * | 2020-04-30 | 2020-09-01 | 杭州拓深科技有限公司 | Method for reducing composite smoke sense false alarm based on intelligent algorithm |
| CN112017386A (en) * | 2020-08-14 | 2020-12-01 | 中国科学院上海微系统与信息技术研究所 | Forest and grassland fire monitoring system |
| CN112071015A (en) * | 2020-09-22 | 2020-12-11 | 中煤科工集团重庆研究院有限公司 | Self-adaptive smoke identification method and device of gas-sensitive smoke sensor for mine |
| CN113570809A (en) * | 2021-07-13 | 2021-10-29 | 中车南京浦镇车辆有限公司 | Fire alarm system for continuously sampling and modeling environment and calculating alarm threshold value |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101219231B1 (en) * | 2011-10-10 | 2013-01-09 | 강성철 | An environment recognition device having requesting help emergency call and transporting sensing image |
| CN103191533A (en) * | 2013-03-25 | 2013-07-10 | 武汉大学 | Intelligent fire monitoring and evacuation indicating system and method |
| CN106073764A (en) * | 2016-05-31 | 2016-11-09 | 深圳市理邦精密仪器股份有限公司 | Reduce the method and device of dynamic electrocardiogram (ECG) data recording equipment power consumption |
| CN110475223A (en) * | 2019-07-01 | 2019-11-19 | 深圳市中电数通智慧安全科技股份有限公司 | A kind of method for safety monitoring, device and terminal device |
| CN110517438A (en) * | 2019-10-11 | 2019-11-29 | 四川长虹电器股份有限公司 | A kind of method and warning device of adaptive adjustment smoke alarm threshold value |
-
2019
- 2019-12-27 CN CN201911372561.3A patent/CN111047815A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101219231B1 (en) * | 2011-10-10 | 2013-01-09 | 강성철 | An environment recognition device having requesting help emergency call and transporting sensing image |
| CN103191533A (en) * | 2013-03-25 | 2013-07-10 | 武汉大学 | Intelligent fire monitoring and evacuation indicating system and method |
| CN106073764A (en) * | 2016-05-31 | 2016-11-09 | 深圳市理邦精密仪器股份有限公司 | Reduce the method and device of dynamic electrocardiogram (ECG) data recording equipment power consumption |
| CN110475223A (en) * | 2019-07-01 | 2019-11-19 | 深圳市中电数通智慧安全科技股份有限公司 | A kind of method for safety monitoring, device and terminal device |
| CN110517438A (en) * | 2019-10-11 | 2019-11-29 | 四川长虹电器股份有限公司 | A kind of method and warning device of adaptive adjustment smoke alarm threshold value |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111613037A (en) * | 2020-04-30 | 2020-09-01 | 杭州拓深科技有限公司 | Method for reducing composite smoke sense false alarm based on intelligent algorithm |
| CN111613037B (en) * | 2020-04-30 | 2022-09-13 | 杭州拓深科技有限公司 | Method for reducing composite smoke sense false alarm based on intelligent algorithm |
| CN112017386A (en) * | 2020-08-14 | 2020-12-01 | 中国科学院上海微系统与信息技术研究所 | Forest and grassland fire monitoring system |
| CN112071015A (en) * | 2020-09-22 | 2020-12-11 | 中煤科工集团重庆研究院有限公司 | Self-adaptive smoke identification method and device of gas-sensitive smoke sensor for mine |
| CN113570809A (en) * | 2021-07-13 | 2021-10-29 | 中车南京浦镇车辆有限公司 | Fire alarm system for continuously sampling and modeling environment and calculating alarm threshold value |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111047815A (en) | Method and system for identifying false alarm of fire-fighting detector based on self-learning model | |
| CN106485871B (en) | System and method for providing early prediction and prediction of false alarms by applying statistical inference models | |
| CN102693600B (en) | Unit disaster control method and system | |
| US20170337485A1 (en) | Event condition detection | |
| CN209821986U (en) | Fire monitoring system | |
| CN116246407A (en) | Agriculture and forestry area fire early warning supervisory systems based on artificial intelligence | |
| CN113053053B (en) | Interval demonstration multi-sensor fusion disaster cellular alarm linkage system based on particle swarm optimization | |
| CN104850047A (en) | Safety monitoring system for electric power equipment | |
| CN110562818B (en) | Monitoring system and monitoring method for evaluating elevator running quality by time dimension | |
| CN104036623A (en) | Method for correcting false data information alarm | |
| CN107895453A (en) | Building safety warning system and method | |
| CN207097187U (en) | A kind of intelligent free-standing smoke sensitive fire hazard detection alarm | |
| CN113741258B (en) | Rail transit station fire monitoring system based on Internet of things and optimization method thereof | |
| JP2003281658A (en) | Safety reporting system | |
| CN107808504A (en) | Alarm method and warning system | |
| CN206331546U (en) | Automatic fire alarm system based on RF wireless sensor networks | |
| CN115359641B (en) | Regular false alarm judging method for photoelectric sensing equipment | |
| CN110796397A (en) | Alarm system and method | |
| KR102298848B1 (en) | Apparatus and method for controlling fire receiver interlocking with digital substation system | |
| CN117729314B (en) | AI vision-based multi-region monitoring system and method | |
| CN211885100U (en) | Fire-fighting monitoring system | |
| US11276284B1 (en) | System and method for detecting events in a system | |
| JP4658462B2 (en) | Environmental monitoring system | |
| CN114038146B (en) | Camera identification alarm device supporting internet connection | |
| CN114593773B (en) | Data acquisition device with cloud data tracking function |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200421 |
|
| RJ01 | Rejection of invention patent application after publication |