CN111145517B - Artificial intelligence electric line fire risk real-time early warning method and system - Google Patents
Artificial intelligence electric line fire risk real-time early warning method and system Download PDFInfo
- Publication number
- CN111145517B CN111145517B CN202010005593.6A CN202010005593A CN111145517B CN 111145517 B CN111145517 B CN 111145517B CN 202010005593 A CN202010005593 A CN 202010005593A CN 111145517 B CN111145517 B CN 111145517B
- Authority
- CN
- China
- Prior art keywords
- abnormal
- risk
- early warning
- level
- 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
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B31/00—Predictive alarm systems characterised by extrapolation or other computation using updated historic data
-
- 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
- G08B17/06—Electric actuation of the alarm, e.g. using a thermally-operated switch
-
- 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
- G08B21/00—Alarms responsive to a single specified undesired or abnormal condition and not otherwise provided for
- G08B21/18—Status alarms
- G08B21/185—Electrical failure alarms
Landscapes
- Business, Economics & Management (AREA)
- Emergency Management (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Alarm Systems (AREA)
Abstract
The invention discloses a real-time early warning method and a real-time early warning system for fire risks of an artificial intelligent electric line, wherein the method comprises the following steps: collecting multidimensional data of each electric circuit in real time, wherein the multidimensional data comprise measuring point temperature rise, meter box temperature rise and load current, and the measuring point temperature rise and the meter box temperature rise are obtained through temperature field collection; carrying out abnormity judgment on the multi-dimensional data to obtain the abnormity grade of each electric circuit; according to the abnormal grade of each electric line and the duration of the abnormal grade, calculating to obtain a fire early warning risk value of each electric line. The invention has the technical characteristics of wide early warning area, high early warning accuracy, high early warning efficiency, real-time early warning and intelligent early warning.
Description
Technical Field
The invention belongs to the technical field of fire risk early warning, and particularly relates to an artificial intelligent electric line fire risk real-time early warning method and system.
Background
The fire disaster is a particularly serious disaster phenomenon, and is an important mark for the progress of social civilization by effectively controlling the fire disaster. In many fire accidents, because of the frequent occurrence of fires caused by electrical line problems, the number of the electrical line fires accounts for 30% of the total number of the fires in 2011 to 2016 according to the statistics of the public security fire service bureau, and the electrical line fires continuously threaten the safety of the public and the social development along with the continuous aging of the electrical line of the social building, and once the electrical line fires, the electrical line fires cause irreversible consequences on economy, scientific and technological innovation, historical relics and the like.
In the prior art, single-dimensional monitoring and threshold alarming are mostly adopted, such as smoke meters, temperature meters, circuit breakers, electric leakage detection and other products, when a fire disaster occurs, fire information can be monitored at the first time to alarm the fire disaster, but the technologies often have the phenomena of false alarm and false alarm failure or alarm processing is carried out when serious problems occur, if the processing is not timely carried out, a large amount of loss can be caused, and huge potential safety hazards are brought to the life safety of personnel and the normal operation of an electric power system.
Therefore, it is desirable to provide a solution that can provide early warning of electrical line fires.
Disclosure of Invention
The invention aims to provide an artificial intelligent electric line fire risk real-time early warning method and system, which have the technical characteristics of wide early warning area, high early warning accuracy, high early warning efficiency, real-time early warning and early warning intellectualization.
In order to solve the problems, the technical scheme of the invention is as follows:
an artificial intelligence electric line fire risk real-time early warning method comprises the following steps:
s1: collecting multidimensional data of each electric line in real time, wherein the multidimensional data comprise measuring point temperature rise, meter box temperature rise and load current, the measuring point temperature rise and the meter box temperature rise are collected and obtained through a temperature field, and the temperature field comprises measuring point temperature, meter box temperature and environment temperature;
s2: carrying out abnormity judgment on the multi-dimensional data to obtain the abnormity grade of each electric circuit;
s3: and calculating the fire early warning risk value of each electric line according to the abnormal grade of each electric line and the duration of the abnormal grade.
According to an embodiment of the present invention, step S2 specifically includes: judging abnormal points and measuring the distances between the abnormal points and a normal area of the multi-dimensional data through a fire risk abnormal level judgment model to obtain the abnormal level of each electric line; wherein the content of the first and second substances,
the establishment process of the fire risk abnormal grade judgment model comprises the following steps:
establishing a load temperature rise analysis model according to the corresponding relation between the temperature field and the load current;
and training to obtain a fire risk abnormal grade judgment model according to the load temperature rise analysis model and an artificial intelligent electrical fire clustering classification algorithm and an artificial intelligent electrical fire grading algorithm.
According to an embodiment of the invention, according to a load temperature rise analysis model, a fire risk abnormal grade judgment model is obtained through training by an artificial intelligence electrical fire clustering classification algorithm and an artificial intelligence electrical fire grading algorithm, and the specific process comprises the following steps:
a1: carrying out data grouping on big data based on multi-dimensional data according to the wire specification, the environmental temperature and the load current;
a2: classifying the grouped data according to the temperature rise of the measuring points and the temperature rise of the meter box, and clustering each level into one class;
a3: carrying out abnormal grade standardized classification on each group of classified data according to a preset standard classification value;
a4, counting the statistics of each level of each group of data after abnormal level standardized classification, and readjusting the abnormal levels of the rest levels of each group of data respectively on the basis of the abnormal level corresponding to the maximum statistics in each group of data to obtain an intelligent abnormal level judgment standard;
a5: and obtaining a fire risk abnormal grade judgment model according to the intelligent abnormal grade judgment standard.
According to an embodiment of the present invention, the abnormal level is divided from level 0 in the increasing direction of the severity, and step a4 specifically includes the following steps:
if the abnormal level corresponding to the maximum statistical value is equal to 0, calculating a relative safety range taking the level corresponding to the maximum statistical value as the center according to a preset percentage and the maximum statistical value, adjusting the abnormal level of the level which falls into the relative safety range and has the abnormal level less than or equal to 1 to be 0, and determining the abnormal level of the rest level according to the distance between the levels;
and if the abnormal level corresponding to the maximum statistical value is not equal to 0, reducing the abnormal levels corresponding to all levels before the level corresponding to the maximum statistical value by one level, and keeping the 0 level unchanged.
According to an embodiment of the present invention, step a5 is followed by step a 6:
and C, performing supervised learning on the multi-dimensional data acquired in real time through the fire risk abnormal level judgment model, updating the supervised learning result into an intelligent abnormal level judgment standard at preset time intervals, and executing the steps A4-A5 again to obtain a new fire risk abnormal level judgment model.
According to an embodiment of the present invention, in step S1, the multidimensional data further includes smoke information and humidity information.
According to an embodiment of the present invention, in step S3, the calculation process of the fire warning risk value risk is:
risk＝(level+counts/n)*10
wherein, level is the abnormal level, n is the real-time data acquisition frequency up to the current day, and counts is the cumulative frequency of the abnormal level up to the current day.
According to an embodiment of the present invention, the step S3 is followed by the step S4:
and performing risk ranking on each electric line according to the fire early warning risk value, and performing early warning according to the height of the fire early warning risk value.
According to an embodiment of the present invention, step S4 further includes:
and performing longitudinal and transverse analysis on a time dimension according to the fire early warning risk value, wherein the longitudinal analysis obtains the risk trend of each electric line and the probability of the same risk in the history, and the transverse analysis obtains the probability of the same risk in all the electric lines in the early warning area and the maximum risk condition of the whole early warning area.
The invention also provides an artificial intelligence electric line fire risk real-time early warning system, which comprises:
the system comprises one or more Internet of things acquisition sensing terminals, a power supply and control module and a power supply and control module, wherein each Internet of things acquisition sensing terminal comprises a load current detection module, a measuring point temperature detection module, a meter box temperature detection module and an environment temperature detection module and is used for acquiring multi-dimensional data of an electric circuit in real time, the multi-dimensional data comprises measuring point temperature rise, meter box temperature rise and load current, the measuring point temperature rise and the meter box temperature rise are acquired through a temperature field, and the temperature field comprises measuring point temperature, meter box temperature and environment temperature;
the intelligent gateway is in signal connection with the Internet of things acquisition sensing terminal and is used for transmitting multi-dimensional data acquired in real time;
the data processing server is in data communication with the intelligent gateway and comprises an abnormal grade judging module and a risk processing module, wherein the abnormal grade judging module is used for performing abnormal judgment on the multidimensional data to obtain the abnormal grade of each electric line; and the risk processing module is used for calculating and obtaining the fire early warning risk value of each electric line according to the abnormal grade of each electric line and the duration of the abnormal grade.
According to an embodiment of the invention, the internet of things acquisition sensing terminal further comprises a smoke detection module and a humidity detection module.
According to an embodiment of the invention, the system further comprises an early warning monitoring platform, which is used for carrying out risk ranking on each electric line according to the fire early warning risk value and carrying out early warning according to the height of the fire early warning risk value.
According to an embodiment of the invention, the early warning monitoring platform is further configured to perform longitudinal and transverse analysis in a time dimension according to the fire early warning risk value, the longitudinal analysis obtains risk trends of each electrical line and probabilities of the same risk appearing in the history, and the transverse analysis obtains probabilities of the same risk appearing in all electrical lines in the early warning area and the maximum risk condition of the whole early warning area.
Compared with the prior art, the invention has the following advantages and positive effects:
1) the invention introduces a temperature field to monitor an electric circuit, respectively monitors the temperature of a measuring point, the temperature of a meter box and the ambient temperature, simultaneously changes three temperature dimensions into two dimension temperature rises through the temperature field, namely the temperature rise of the measuring point and the temperature rise of the meter box, reduces the dimension of original data, effectively reduces the data processing amount for the later big data early warning analysis, improves the efficiency of the overall fire early warning, adopts the temperature rise of the measuring point and the temperature rise of the meter box to carry out fire early warning, eliminates the interference of the ambient temperature, simultaneously calculates the fire early warning risk value based on the abnormal grade and the abnormal grade duration, eliminates the accidental abnormality and improves the accuracy of the fire early warning;
2) the invention carries out abnormity judgment based on multi-dimensional data to obtain the abnormity grade of each electric line, concretely carries out judgment of an abnormal point and measurement of the distance between the abnormal point and a normal area through a fire risk abnormity grade judgment model to determine the abnormity grade, wherein, the fire risk is converted into the discrimination of abnormal points and the measurement of the distance between the abnormal points and the normal area through a load temperature rise analysis model, thereby facilitating the determination of abnormal grade, training based on the big data of multi-dimensional data to obtain a fire risk abnormal grade judgment model through a clustering classification algorithm and a grading algorithm, realizing the accurate determination of the abnormal grade, the fire hazard abnormal grade judgment model can realize real-time, quick and accurate fire hazard early warning of the electric line in a large area, and achieves the technical effects of wide early warning area, high early warning accuracy, high early warning efficiency and real-time early warning;
3) the invention combines the characteristics of the power industry and the characteristics of acquired data, adopts a clustering classification algorithm and a classification algorithm which are fused to obtain a fire risk abnormal grade judgment model, wherein abnormal points can be judged based on the clustering algorithm, the quantitative measurement of the abnormal grades is realized based on the measurement of the distances between the abnormal points and normal regions of the classification algorithm, the division of the abnormal grades is flexibly determined based on the classification algorithm, different standards can be automatically formed for the abnormal grades according to different conditions by the classification algorithm, in other words, the standards formed for the abnormal grades are different according to different wire specifications, different environmental temperatures and different currents, the grades can be newly added according to the actual conditions, the temperature rise value is not limited, so that the intelligent abnormal grade judgment standard can be obtained, and the fire risk abnormal grade judgment model based on the classification algorithm can realize self-adaptive fire early warning, the fire early warning universality and the intelligent degree are improved, and the technical effects of wide application range and early warning intelligence are achieved;
4) according to the fire disaster early warning method, auxiliary abnormal grade judgment is carried out through other multi-dimensional data such as smoke information and humidity information, so that the reliability and accuracy of the fire disaster early warning risk value are improved;
5) according to the invention, not only can early warning be carried out according to the fire early warning risk value, but also a plurality of aspects of risk analysis can be carried out, wherein, the risk trend of each electric line, the probability of the same risk appearing in the history, the probability of the same risk appearing in all the electric lines in the early warning area and the maximum risk condition of the whole early warning area can be obtained by carrying out longitudinal and transverse analysis on the multidimensional data of each electric line, so that the individual fire early warning analysis of the unit electric line and the regional fire early warning analysis of a plurality of electric lines can be carried out, thus, the reliability, the comprehensiveness and the accuracy of the fire early warning are improved.
Drawings
FIG. 1 is a schematic main flow chart of a real-time fire risk early warning method for an artificial intelligence electrical line according to the present invention;
FIG. 2 is a schematic diagram of a temperature field of a real-time fire risk early warning method for an artificial intelligence electrical line according to the present invention;
FIG. 3 is a schematic diagram of temperature field acquisition of a real-time fire risk early warning method for an artificial intelligence electrical line of the present invention;
FIG. 4 is a schematic diagram of a model building process of the artificial intelligence electric line fire risk real-time early warning method of the present invention;
fig. 5 is a schematic structural diagram of an artificial intelligence electric line fire risk real-time early warning system according to the present invention.
Description of reference numerals:
1-an internet of things acquisition sensing terminal; 2-an intelligent gateway; 3-a data processing server; 4-early warning monitoring platform.
Detailed Description
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following description will be made with reference to the accompanying drawings. It is obvious that the drawings in the following description are only some examples of the invention, and that for a person skilled in the art, other drawings and embodiments can be derived from them without inventive effort.
For the sake of simplicity, the drawings only schematically show the parts relevant to the present invention, and they do not represent the actual structure as a product. In addition, in order to make the drawings concise and understandable, components having the same structure or function in some of the drawings are only schematically illustrated or only labeled. In this document, "one" means not only "only one" but also a case of "more than one".
The method and system for real-time early warning of fire risk in an artificial intelligence electrical line provided by the invention are further described in detail with reference to the accompanying drawings and specific embodiments.
According to the artificial intelligent electric line fire early warning method and system, the important information in the collected multidimensional data is analyzed and monitored in real time by using modern technology such as data collection, real-time transmission, big data storage, big data analysis and the like of the Internet of things, and the abnormal condition of the electric line is found and early warned in time, so that the real-time condition of the monitored line can be effectively, timely and accurately displayed, management personnel can be assisted to handle the situation, and the risk can be controlled within a safety range.
Example 1
Referring to fig. 1, the application provides a real-time early warning method for fire risk of an artificial intelligence electrical line, comprising the following steps:
s1: collecting multidimensional data of each electric line in real time, wherein the multidimensional data comprise measuring point temperature rise, meter box temperature rise and load current, the measuring point temperature rise and the meter box temperature rise are collected and obtained through a temperature field, and the temperature field comprises measuring point temperature, meter box temperature and environment temperature;
s2: carrying out abnormity judgment on the multi-dimensional data to obtain the abnormity grade of each electric circuit;
s3: and calculating the fire early warning risk value of each electric line according to the abnormal grade of each electric line and the duration of the abnormal grade.
The present embodiment will now be described in detail:
according to the characteristics of the electric circuit, the electric circuit fire can occur in 6 stages: the real-time fire risk early warning can be carried out from the fault heating stage, the full-stage fire early warning can be carried out, corresponding treatment is carried out according to different stages, and the risk is controlled within a reasonable range. Thereby preventing a fire from occurring.
1) In the embodiment, in order to cover the whole process of a fire, multi-dimensional data collection is adopted in step S1, the multi-dimensional data include measuring point temperature, ambient temperature, electric meter box temperature, load current, smoke, humidity and time, wherein the embodiment introduces a concept of a temperature field, referring to fig. 2 and fig. 3, a temperature field is formed together based on the measuring point temperature, the ambient temperature and the temperature of the electric meter box, the three respectively monitor the temperature of an electric circuit, the ambient temperature outside the electric meter box and the temperature inside the electric meter box, and one surface is formed by the three temperatures, so that the risk can be judged more comprehensively. Through the temperature field, three temperature dimensions are changed into two dimension temperature rises, a measuring point temperature rise and a meter box temperature rise, the measuring point temperature rise indicates the difference value of the measuring point temperature minus the ambient temperature, the representing temperature rise indicates that the meter box temperature minus the ambient temperature, and in the subsequent big data analysis, the dimension reduction is carried out on the original data, so that the data processing amount is effectively reduced, the efficiency of the whole fire early warning is improved, and the influence of the ambient temperature on risk judgment is removed.
In the embodiment, the temperature field is introduced to monitor the electric circuit, the measuring point temperature, the meter box temperature and the environment temperature are respectively monitored, three temperature dimensions are changed into two dimension temperature rises through the temperature field, namely the measuring point temperature rise and the meter box temperature rise, the dimension of original data is reduced, the processing amount of data is effectively reduced for the later big data early warning analysis, the efficiency of overall fire early warning is improved, and the measuring point temperature rise and the meter box temperature rise are adopted to carry out fire early warning, so that the interference of the environment temperature is eliminated.
2) Step S2 of this embodiment specifically includes: judging abnormal points and measuring the distance between the abnormal points and the stable range of the multi-dimensional data through a fire risk abnormal level judgment model to obtain the abnormal level of each electric line; wherein the content of the first and second substances,
the establishment process of the fire risk abnormal grade judgment model is as follows: establishing a load temperature rise analysis model according to the corresponding relation between the temperature field and the load current; and training to obtain a fire risk abnormal grade judgment model according to the load temperature rise analysis model through an artificial intelligent electrical fire clustering classification algorithm and an artificial intelligent electrical fire grading algorithm.
In the embodiment, a load temperature rise analysis model is established according to the corresponding relation between the load current and the temperature field, the risk degree is converted into the discrimination of an abnormal point and the measurement of the distance between the abnormal point and a stable range through a load temperature rise analysis model, wherein, based on the big data statistics of the multidimensional data, in the corresponding relation of the load current and the temperature field, most data points of the electric circuit are in a relatively stable range, which can be regarded as a relatively normal temperature value, but abnormal points also appear, these outliers, which are not as far from the relatively stable range, refract different relative risk situations, under the same load current, the temperature rise is different and represents different risk degrees, the farther the abnormal point is from the stable range, the greater the risk degree is, therefore, the model can convert the risk degree into the discrimination of the abnormal point and the measurement of the distance between the abnormal point and the stable range.
Referring to fig. 4, in the embodiment, a fire risk abnormal level judgment model is obtained through training of an artificial intelligence electrical fire clustering classification algorithm and an artificial intelligence electrical fire classification algorithm, and the specific process includes the following steps:
a1: carrying out data grouping on big data based on multi-dimensional data according to the wire specification, the environmental temperature and the load current;
a2: classifying the grouped data according to the temperature rise of the measuring points and the temperature rise of the meter box, and clustering each level into one class;
a3: carrying out abnormal grade standardized classification on each group of classified data according to a preset standard classification value;
a4, counting the statistics of each level of each group of data after abnormal level standardized classification, and readjusting the abnormal levels of the rest levels of each group of data respectively on the basis of the abnormal level corresponding to the maximum statistics in each group of data to obtain an intelligent abnormal level judgment standard;
a5: and obtaining a fire risk abnormal grade judgment model according to the intelligent abnormal grade judgment standard.
Taking an example:
the big data of the multi-dimensional data is firstly classified into whole data through the artificial intelligence electric fire cluster classification algorithm: 1) classifying the data according to the same wire specification, the same environmental temperature and the same current, and setting a fixed size interval for each dimension classification according to actual needs; 2) the temperature rise is set up as a hierarchy according to the fixed temperature rise, and each hierarchy is grouped into one class, so that the distance measurement problem can be solved, and thus, the data distribution as shown in table 1 can be obtained:
TABLE 1 post-clustering-algorithm data distribution
| t_at | counts | proportion |
| (0.0，2.0] | 1143412 | 55.6925％ |
| (2.0，4.0] | 725189 | 35.3220％ |
| (4.0，6.0] | 133317 | 6.4935％ |
| (6.0，8.0] | 28148 | 1.3710％ |
| (8.0，10.0] | 9380 | 0.4569％ |
| (10.0，12.0] | 5629 | 0.2742％ |
| (12.0，14.0] | 3283 | 0.1599％ |
| (14.0，16.0] | 1857 | 0.0904％ |
| (16.0，18.0] | 1140 | 0.0555％ |
| (18.0，20.0] | 877 | 0.0427％ |
| (20.0，22.0] | 410 | 0.0200％ |
| (22.0，24.0] | 207 | 0.0101％ |
| (24.0，26.0] | 165 | 0.0080％ |
| (26.0，28.0] | 60 | 0.0029％ |
| (28.0，30.0] | 7 | 0.0003％ |
As shown in table 1, t _ at is the temperature rise range, counts is the number of data points corresponding to the tat range, and contribution is the data distribution probability corresponding to the tat range, where the number of statistics of temperature rise at each level represents the probability of the temperature rise, in other words, the different degrees of risk represented. According to the industry characteristics, the higher the temperature rise, the greater the risk degree, and the distribution conforms to the industry characteristics. The number in the reasonable range is far greater than the number of abnormal points, and the measuring point temperature rise (-4.0-6.0) accounts for 97.8807%, so that the algorithm can reveal the characteristics of risk distribution and can also meet the requirements of abnormal point judgment and abnormal point and stable range distance measurement. Based on the above data distribution, the classification of the abnormality level can be performed easily, but in actual data, as in table 2, the case often occurs:
TABLE 2 other data distribution scenarios
| tat | counts | tat | counts | tat | counts |
| (-0.001，2.0] | 1023 | (-0.001，2.0] | 4 | (-0.001，2.0] | 0 |
| (2.0，4.0] | 264 | (2.0，4.0] | 11 | (2.0，4.0] | 0 |
| (4.0，6.0] | 276 | (4.0，6.0] | 55 | (4.0，6.0] | 0 |
| (6.0，8.0] | 142 | (6.0，8.0] | 20 | (6.0，8.0] | 0 |
| (8.0，10.0] | 282 | (8.0，10.0] | 16 | (8.0，10.0] | 0 |
| (10.0，12.0] | 538 | (10.0，12.0] | 92 | (10.0，12.0] | 0 |
| (12.0，14.0] | 293 | (12.0，14.0] | 88 | (12.0.14.0] | 0 |
| (14.0，16.0] | 176 | (14.0，16.0] | 3 | (14.0，16.0] | 0 |
| (16.0，18.0] | 135 | (16.0，18.0] | 10 | (16.0，18.0] | 0 |
| (18.0，20.0] | 88 | (18.0，20.0] | 60 | (18.0，20.0] | 4 |
| (20.0，22.0] | 14 | (20.0，22.0] | 30 | (20.0，22.0] | 5 |
| (22.0，24.0] | 0 | (22.0，24.0] | 0 | (22.0，24.0] | 14 |
| (24.0，26.0] | 0 | (24.0，26.0] | 0 | (24.0，26.0] | 0 |
| (26.0，28.0] | 0 | (26.0，28.0] | 0 | (26.0.28.0] | 0 |
| (28.0，30.0] | 0 | (28.0，30.0] | 0 | (28.0，30.0] | 0 |
As shown in table 2, three columns of data represent statistical grouping data of temperature rises of the measurement points under different conditions, respectively, and due to temperature rise distribution and idealization as in table 1, the conditions for determination of the abnormal point, determination of the abnormal point and the stable range distance may not be satisfied if based on one of the abnormal level determination criteria alone.
Therefore, the present embodiment further combines with an artificial intelligence electrical fire classification algorithm to obtain an intelligent abnormal level judgment standard of the abnormal level judgment standard, and specifically, if data after the above clustering classification algorithm is distributed as in table 3, where tat is a temperature rise value, which may be temperature rise of a measurement point or temperature rise of a meter box, counts is the number of data points corresponding to tat, ambient temperature, 30 degrees celsius in the table, a _ port is a load current, 21 amperes in the table, the temperature rise values are classified according to actual needs, and each level is clustered into one class according to 2 degrees celsius, and clustered into 30 classes.
TABLE 3 temperature data distribution
Currently, specific classification is performed through an artificial intelligence electrical fire classification algorithm:
and carrying out abnormal grade standardized classification on each group of classified data according to a preset standard classification value, wherein the standard classification value is as follows:
[0，0，0，1，1，1，2，2，2，3，3，3，4，4，4，5，5，5，6，6，6，7，7，7，8，8，8，9，9，9]
thus, the judgment standard after the abnormal grade is classified in a standardized way can be obtained, the judgment standard is only a preliminary standard, the statistic value of each grade is counted, and the rest grades are determined based on the abnormal grade with the maximum statistic value in each group of data and based on the measurement of the distance:
if the abnormal level corresponding to the maximum statistical value is equal to 0, calculating a relative safety range taking the level corresponding to the maximum statistical value as the center according to a preset percentage and the maximum statistical value, adjusting the abnormal level of the level which falls into the relative safety range and has the abnormal level less than or equal to 1 to be 0, and determining the abnormal level of the rest level according to the distance between the levels;
and if the abnormal level corresponding to the maximum statistical value is not equal to 0, reducing the abnormal levels corresponding to all levels before the level corresponding to the maximum statistical value by one level, and keeping the 0 level unchanged.
Based on the above, the corresponding intelligent abnormality level determination criteria after determination in table 3 are shown in table 4:
TABLE 4 rating case
The hierarchy corresponding to the count of 234 in the original table 3 is originally the abnormal level 1, and is adjusted to be the abnormal level 0, and so on, the subsequent abnormal levels are all corrected, and thus, the intelligent abnormal level judgment standard is adaptively adjusted according to the current conditions. Based on the intelligent abnormal grade judgment standard, the fire risk abnormal grade judgment model of the embodiment can be obtained, and the abnormal grade judgment of collecting the multidimensional data of each electric line in real time is executed.
Based on the above example, the abnormal level of the intelligent abnormal level judgment standard obtained based on the classification algorithm in this embodiment may automatically form different standards according to different conditions, in other words, the formed standards are different for different wire specifications, different environmental temperatures, and different currents.
Preferably, referring to fig. 4, step B5 is further followed by step B6: and carrying out supervision learning on the real-time acquired multidimensional data according to an intelligent abnormal grade judgment standard of the fire risk abnormal grade judgment model, classifying to obtain an information quantity corresponding table of the monitored electric line, writing back the information quantity corresponding table to the intelligent abnormal grade judgment standard every day through the information quantity table, and reclassifying according to the classification algorithm to obtain a new artificial intelligent measuring point temperature rise abnormal grade standard table.
By this processing, the four dimensional risks can be compared. The comprehensive risk value calculation formula is the maximum value of the measuring point temperature rise risk value, the meter box temperature rise risk value, the smoke risk value and the environment temperature risk value.
In the embodiment, the characteristics of the power industry and the characteristics of acquired data are combined, a clustering classification algorithm and a classification algorithm which are fused are adopted to obtain a fire risk abnormal grade judgment model, wherein abnormal points can be judged based on the clustering algorithm, the quantitative measurement of the abnormal grades is realized based on the measurement of the distances between the abnormal points and normal regions of the classification algorithm, the division of the abnormal grades is flexibly determined based on the classification algorithm, different standards can be automatically formed for the abnormal grades according to different conditions in the classification algorithm, in other words, the standards formed for different wire specifications, different environmental temperatures and different currents are different, the grades can be newly added according to actual conditions, the temperature rise value is not limited, so that the intelligent abnormal grade judgment standard can be obtained, and the fire risk abnormal grade judgment model based on the method can realize self-adaptive fire early warning, the fire early warning universality and the intelligent degree are improved, and the technical effects of wide application range and early warning intelligence are achieved.
3) In step S3 of this embodiment, for convenience of understanding and expression, the risk value of this embodiment is designed to have a specific risk value range of 0 to 100, and considering the influence of time, such as 5-level, some lasting for 10 minutes, some lasting for 30 minutes, and the risk values corresponding to these two cases obviously have differences, so a fire early warning risk value is specifically obtained according to the duration of the abnormal level and the abnormal level, and the calculation process of the fire early warning risk value risk is as follows:
risk＝(level+counts/n)*10
wherein, level is the abnormal level, n is the real-time data acquisition frequency up to the current day, and counts is the cumulative frequency of the abnormal level up to the current day.
In order to more comprehensively show specific risk conditions, the fire early warning risk value of the embodiment is calculated according to other multidimensional data, specifically, the environmental temperature and the smoke risk are processed in a segmented manner, within a certain threshold, the abnormal level of the temperature rise of the measuring point and the temperature rise of the meter box is consistent, and when the certain threshold is exceeded, a higher abnormal level is set, for example, smoke is treated with the risk that the abnormal level of the temperature rise of the measuring point and the temperature rise of the meter box is equal to or lower than 4, and the risk is equal to the abnormal level of the temperature rise of the measuring point and the temperature rise of the meter box. Above level 4, levels 5-7 are directly equal to level 8 of the exception level, and levels 8-9 are directly equal to level 9 of the exception level. Ambient temperature may be treated similarly. Through the processing, the four dimensionality risks can be compared, and the final comprehensive risk value is the maximum value of the measuring point temperature rise risk value, the meter box temperature rise risk value, the smoke risk value and the environment temperature risk value.
Referring to fig. 1, step S3 of the present embodiment is followed by step S4: performing risk ranking on each electric line according to the fire early warning risk value, and performing early warning according to the height of the fire early warning risk value; and performing longitudinal and transverse analysis on a time dimension according to the fire early warning risk value, wherein the longitudinal analysis obtains the risk trend of each electric line and the probability of the same risk in the history, and the transverse analysis obtains the probability of the same risk in all the electric lines in the early warning area and the maximum risk condition of the whole early warning area.
Specifically, by ranking risk values today, the risk status of a certain area of the day can be obtained. For the risky electric lines, firstly, corresponding early warning is given according to different risk grades; secondly, analyzing from the time dimension, transversely and longitudinally respectively, and judging the risk condition more comprehensively through history:
by ranking the risk values of the day, the overall risk condition of the current area can be obtained, and different early warning treatments can be timely made according to the risk grades;
by longitudinally comparing the time dimensions of the same line, the probability (historical occupation ratio) of the same risk appearing in the history and the current risk trend (risk trend) can be obtained, and the current line risks can be seen, and the situations that the current line risks are at the high level or the low level of the historical risks, and are frequently or suddenly appeared can be seen. For example: the risk is large today, the history proportion is small, the trend is an ascending trend, the risk degree of the line is high, and attention is needed;
by transversely comparing different line time dimensions, the probability (daily risk ratio) of the same risk occurring in all monitored lines in the area and the maximum risk condition of the reference area can be obtained. From this, it can be obtained whether the line risk condition is a ubiquitous or abnormal condition. For example: if the risk value and the regional maximum value do not differ much, the risk share on the day is smaller. It can be determined that the line has a high risk degree in the area and needs attention.
Through the steps, the current and historical risk conditions of each line can be clearly judged, and timely and comprehensive evaluation can be made. Finally, according to this evaluation, according to the security level gradient descent method: continuously paying attention to the lines in the risk controllable range according to the risk conditions; and (4) processing the line beyond the risk controllable range, and reducing the risk level of the monitored line to be within the controllable range, so that the risk can be always controlled to be within the safe range.
This embodiment not only can carry out the early warning according to conflagration early warning risk value, can also carry out many-sided risk analysis, wherein, carry out vertically, horizontal analysis through the multidimension degree data of each electric line, can obtain the risk trend of each electric line, the probability that equal risk appears in the history, the probability that equal risk appears in all electric lines in the early warning region, the regional biggest risk condition of whole early warning, thereby can carry out the individuality conflagration early warning analysis of unit electric line, and the regional conflagration early warning analysis of numerous electric lines, therefore, the reliability, the comprehensiveness, the accuracy of conflagration risk early warning has been improved.
Example 2
Referring to fig. 5, the present application provides an artificial intelligence electrical line fire risk real-time early warning system based on embodiment 1, including:
the system comprises one or more Internet of things acquisition sensing terminals 1, wherein each Internet of things acquisition sensing terminal comprises a load current detection module, a measuring point temperature detection module, a meter box temperature detection module and an environment temperature detection module and is used for acquiring multi-dimensional data of an electric circuit in real time, the multi-dimensional data comprises measuring point temperature rise, meter box temperature rise and load current, the measuring point temperature rise and the meter box temperature rise are acquired through a temperature field, and the temperature field comprises measuring point temperature, meter box temperature and environment temperature;
the intelligent gateway 2 is in signal connection with the Internet of things acquisition sensing terminal and is used for transmitting multi-dimensional data acquired in real time;
the data processing server 3 is in data communication with the intelligent gateway and comprises an abnormal grade judging module and a risk processing module, and the abnormal grade judging module is used for performing abnormal judgment on the multidimensional data to obtain the abnormal grade of each electric line; and the risk processing module is used for calculating and obtaining the fire early warning risk value of each electric line according to the abnormal grade of each electric line and the duration of the abnormal grade.
The present embodiment will now be described in detail:
according to the characteristics of the electric circuit, the electric circuit fire can occur in 6 stages: the artificial intelligent real-time early warning system for the fire risk of the electrical line can start early warning from a fault heating stage, can perform full-stage fire early warning, performs corresponding treatment according to different stages, and controls the risk within a reasonable range. Thereby preventing a fire from occurring.
In the embodiment, for forming the whole process of covering a fire, the internet of things acquisition sensing terminal adopts multi-dimensional data acquisition and is provided with a load current detection module, a measuring point temperature detection module, a meter box temperature detection module, an environment temperature detection module, a smoke detection module and a humidity detection module, wherein the multi-dimensional data comprises measuring point temperature, environment temperature, electricity meter box temperature, load current, smoke, humidity and time, the concept of a temperature field is introduced in the embodiment, referring to fig. 2 and fig. 3, a temperature field is formed jointly based on the measuring point temperature, the environment temperature and the meter box temperature, the three respectively monitor the temperature of an electric circuit, the environment temperature outside the electricity meter box and the temperature inside the electricity meter box, and one surface is formed by three temperatures, so that the risk can be judged more comprehensively. Through the temperature field, three temperature dimensions are changed into two dimension temperature rises, a measuring point temperature rise and a meter box temperature rise, the measuring point temperature rise indicates the difference value of the measuring point temperature minus the ambient temperature, the representing temperature rise indicates that the meter box temperature minus the ambient temperature, and in the subsequent big data analysis, the dimension reduction is carried out on the original data, so that the data processing amount is effectively reduced, the efficiency of the whole fire early warning is improved, and the influence of the ambient temperature on risk judgment is removed.
In the embodiment, the temperature field is introduced to monitor the electric circuit, the measuring point temperature, the meter box temperature and the environment temperature are respectively monitored, three temperature dimensions are changed into two dimension temperature rises through the temperature field, namely the measuring point temperature rise and the meter box temperature rise, the dimension of original data is reduced, the processing amount of data is effectively reduced for the later big data early warning analysis, the efficiency of overall fire early warning is improved, and the measuring point temperature rise and the meter box temperature rise are adopted to carry out fire early warning, so that the interference of the environment temperature is eliminated.
The intelligent gateway of this embodiment accessible loRa, LoRaWAN, NB-lot etc. technique and thing networking collection perception terminal signal connection to receive the data that thing networking collection perception terminal gathered in real time, and the accessible is wired or send the data of gathering for data processing server with wireless ground mode, and intelligent gateway has the characteristics of multiple thing networking communication mode, simple and easy nimble parameter configuration, multichannel data cache, visual operation suggestion, satisfies in big data acquisition's needs.
The data processing server specifically comprises a data storage server and an artificial intelligent machine learning server, and is matched with the data storage server to realize real-time fire risk early warning of the multi-dimensional data, wherein the abnormal grade judgment module judges abnormal points and measures the distance between the abnormal points and a normal area of the multi-dimensional data through a fire risk abnormal grade judgment model to obtain the abnormal grade of each electric line; the establishment process of the fire risk abnormal grade judgment model comprises the following steps: establishing a load temperature rise analysis model according to the corresponding relation between the temperature field and the load current; and training to obtain a fire risk abnormal grade judgment model according to the load temperature rise analysis model and an artificial intelligent electrical fire clustering classification algorithm and an artificial intelligent electrical fire grading algorithm.
In the embodiment, a load temperature rise analysis model is established according to the corresponding relation between the load current and the temperature field, the risk degree is converted into the discrimination of an abnormal point and the measurement of the distance between the abnormal point and a stable range through a load temperature rise analysis model, wherein, based on the big data statistics of the multidimensional data, in the corresponding relation of the load current and the temperature field, most data points of the electric circuit are in a relatively stable range, which can be regarded as a relatively normal temperature value, but abnormal points also appear, these outliers, which are not as far from the relatively stable range, refract different relative risk situations, under the same load current, the temperature rise is different and represents different risk degrees, the farther the abnormal point is from the stable range, the greater the risk degree is, therefore, the model can convert the risk degree into the discrimination of the abnormal point and the measurement of the distance between the abnormal point and the stable range.
According to the load temperature rise analysis model, a fire risk abnormal grade judgment model is obtained through training by an artificial intelligent electrical fire clustering classification algorithm and an artificial intelligent electrical fire grading algorithm, and the specific process is as follows: a1: carrying out data grouping on big data based on multi-dimensional data according to the wire specification, the environmental temperature and the load current; a2: classifying the grouped data according to the temperature rise of the measuring points and the temperature rise of the meter box, and clustering each level into one class; a3: carrying out abnormal grade standardized classification on each group of classified data according to a preset standard classification value; a4, counting the statistics of each level of each group of data after abnormal level standardized classification, and readjusting the abnormal levels of the rest levels of each group of data respectively on the basis of the abnormal level corresponding to the maximum statistics in each group of data to obtain an intelligent abnormal level judgment standard; a5: and obtaining a fire risk abnormal grade judgment model according to the intelligent abnormal grade judgment standard.
Specifically, the abnormality levels are divided from level 0 in the increasing direction of the severity, and step a4 specifically includes the following steps: if the abnormal level corresponding to the maximum statistical value is equal to 0, calculating a relative safety range taking the level corresponding to the maximum statistical value as the center according to a preset percentage and the maximum statistical value, adjusting the abnormal level of the level which falls into the relative safety range and has the abnormal level less than or equal to 1 to be 0, and determining the abnormal level of the rest level according to the distance between the levels; and if the abnormal level corresponding to the maximum statistical value is not equal to 0, reducing the abnormal levels corresponding to all levels before the level corresponding to the maximum statistical value by one level, and keeping the 0 level unchanged.
The abnormal grade judging module of this embodiment can automatically form different standards according to different conditions based on the abnormal grade of the intelligent abnormal grade judging standard obtained by the grading algorithm, in other words, the formed standards are different under different wire specifications, different environmental temperatures and different currents.
In the embodiment, the characteristics of the power industry and the characteristics of acquired data are combined, a clustering classification algorithm and a classification algorithm which are fused are adopted to obtain a fire risk abnormal grade judgment model, wherein abnormal points can be judged based on the clustering algorithm, the quantitative measurement of the abnormal grades is realized based on the measurement of the distances between the abnormal points and normal regions of the classification algorithm, the division of the abnormal grades is flexibly determined based on the classification algorithm, different standards can be automatically formed for the abnormal grades according to different conditions in the classification algorithm, in other words, the standards formed for different wire specifications, different environmental temperatures and different currents are different, the grades can be newly added according to actual conditions, the temperature rise value is not limited, so that the intelligent abnormal grade judgment standard can be obtained, and the fire risk abnormal grade judgment model based on the method can realize self-adaptive fire early warning, the fire early warning universality and the intelligent degree are improved, and the technical effects of wide application range and early warning intelligence are achieved.
Specifically, for convenience of understanding and expression, the range of the risk value of the design of the risk value of the embodiment is 0-100, and considering the influence of time, such as 5-level, some lasting 10 minutes, and some lasting 30 minutes, the risk values corresponding to the two cases obviously have differences, so the fire early warning risk value is obtained specifically according to the duration of the abnormal level and the abnormal level, and the calculation process of the fire early warning risk value risk of the risk processing module is as follows:
risk＝(level+counts/n)*10
wherein, level is the abnormal level, n is the real-time data acquisition frequency up to the current day, and counts is the cumulative frequency of the abnormal level up to the current day.
In order to more comprehensively show specific risk conditions, the fire early warning risk value of the embodiment is calculated according to other multidimensional data, specifically, the environmental temperature and the smoke risk are processed in a segmented manner, within a certain threshold, the abnormal level of the temperature rise of the measuring point and the temperature rise of the meter box is consistent, and when the certain threshold is exceeded, a higher abnormal level is set, for example, smoke is treated with the risk that the abnormal level of the temperature rise of the measuring point and the temperature rise of the meter box is equal to or lower than 4, and the risk is equal to the abnormal level of the temperature rise of the measuring point and the temperature rise of the meter box. Above level 4, levels 5-7 are directly equal to level 8 of the exception level, and levels 8-9 are directly equal to level 9 of the exception level. Ambient temperature may be treated similarly. Through the processing, the four dimensionality risks can be compared, and the final comprehensive risk value is the maximum value of the measuring point temperature rise risk value, the meter box temperature rise risk value, the smoke risk value and the environment temperature risk value.
Preferably, referring to fig. 5, the embodiment further includes an early warning monitoring platform 4, where the early warning monitoring platform may be an individual mobile terminal, a regional utility service platform, an executed fire-fighting execution platform, and the like, and the data processing server serves the early warning monitoring platform through the Web server, where the early warning monitoring platform may rank risks of each electrical line according to fire early warning risk values, and perform early warning according to the level of the fire early warning risk values. The early warning monitoring platform can also perform longitudinal and transverse analysis in a time dimension according to the fire early warning risk value, the longitudinal analysis obtains the risk trend of each electric line and the probability of the same risk in the history, and the transverse analysis obtains the probability of the same risk in all the electric lines in the early warning area and the maximum risk condition of the whole early warning area.
Specifically, by ranking risk values today, the risk status of a certain area of the day can be obtained. For the risky electric lines, firstly, corresponding early warning is given according to different risk grades; secondly, analyzing from the time dimension, transversely and longitudinally respectively, and judging the risk condition more comprehensively through history:
by ranking the risk values of the day, the overall risk condition of the current area can be obtained, and different early warning treatments can be timely made according to the risk grades;
by longitudinally comparing the time dimensions of the same line, the probability (historical occupation ratio) of the same risk appearing in the history and the current risk trend (risk trend) can be obtained, and the current line risks can be seen, and the situations that the current line risks are at the high level or the low level of the historical risks, and are frequently or suddenly appeared can be seen. For example: the risk is large today, the history proportion is small, the trend is an ascending trend, the risk degree of the line is high, and attention is needed;
by transversely comparing different line time dimensions, the probability (daily risk ratio) of the same risk occurring in all monitored lines in the area and the maximum risk condition of the reference area can be obtained. From this, it can be obtained whether the line risk condition is a ubiquitous or abnormal condition. For example: if the risk value and the regional maximum value do not differ much, the risk share on the day is smaller. It can be determined that the line has a high risk degree in the area and needs attention.
Through the steps, the current and historical risk conditions of each line can be clearly judged, and timely and comprehensive evaluation can be made. Finally, according to this evaluation, according to the security level gradient descent method: continuously paying attention to the lines in the risk controllable range according to the risk conditions; and (4) processing the line beyond the risk controllable range, and reducing the risk level of the monitored line to be within the controllable range, so that the risk can be always controlled to be within the safe range.
This embodiment not only can carry out the early warning according to conflagration early warning risk value, can also carry out many-sided risk analysis, wherein, carry out vertically, horizontal analysis through the multidimension degree data of each electric line, can obtain the risk trend of each electric line, the probability that equal risk appears in the history, the probability that equal risk appears in all electric lines in the early warning region, the regional biggest risk condition of whole early warning, thereby can carry out the individuality conflagration early warning analysis of unit electric line, and the regional conflagration early warning analysis of numerous electric lines, therefore, the reliability, the comprehensiveness, the accuracy of conflagration risk early warning has been improved.
The embodiments of the present invention have been described in detail with reference to the accompanying drawings, but the present invention is not limited to the above embodiments. Even if various changes are made to the present invention, it is still within the scope of the present invention if they fall within the scope of the claims of the present invention and their equivalents.
Claims (10)
1. An artificial intelligence electric line fire risk real-time early warning method is characterized by comprising the following steps:
s1: the method comprises the steps that multi-dimensional data of each electric line are collected in real time, wherein the multi-dimensional data comprise measuring point temperature rise, meter box temperature rise and load current, the measuring point temperature rise and the meter box temperature rise are collected through a temperature field, and the temperature field comprises measuring point temperature, meter box temperature and environment temperature;
s2: carrying out abnormity judgment on the multi-dimensional data to obtain the abnormity grade of each electric circuit;
the step S2 specifically includes: judging abnormal points and measuring the distances between the abnormal points and a normal area of the multi-dimensional data through a fire risk abnormal level judgment model to obtain the abnormal level of each electric line;
wherein the content of the first and second substances,
the establishment process of the fire risk abnormal grade judgment model is as follows:
establishing a load temperature rise analysis model according to the corresponding relation between the temperature field and the load current;
training to obtain a fire risk abnormal grade judgment model through an artificial intelligent electrical fire clustering classification algorithm and an artificial intelligent electrical fire grading algorithm according to the load temperature rise analysis model;
according to the load temperature rise analysis model, a fire risk abnormal grade judgment model is obtained through training by an artificial intelligence electrical fire clustering classification algorithm and an artificial intelligence electrical fire grading algorithm, and the method specifically comprises the following steps:
a1: carrying out data grouping on big data based on the multi-dimensional data according to the wire specification, the ambient temperature and the load current;
a2: classifying the grouped data according to the temperature rise of the measuring points and the temperature rise of the meter box, and clustering each level into one class;
a3: carrying out abnormal grade standardized classification on each group of classified data according to a preset standard classification value;
a4, counting the statistics of each level of each group of data after abnormal level standardized classification, and readjusting the abnormal levels of the rest levels of each group of data respectively on the basis of the abnormal level corresponding to the maximum statistics in each group of data to obtain an intelligent abnormal level judgment standard;
the abnormality levels are divided from level 0 in the increasing direction of the severity, and the step a4 specifically includes the following steps:
if the abnormal level corresponding to the maximum statistical value is equal to 0, calculating a relative safety range taking the level corresponding to the maximum statistical value as the center according to a preset percentage and the maximum statistical value, adjusting the abnormal level of the level which falls into the relative safety range and has the abnormal level less than or equal to 1 to be 0, and determining the abnormal level of the rest level according to the distance between the levels;
if the abnormal level corresponding to the maximum statistical value is not equal to 0, reducing the abnormal levels corresponding to all levels before the level corresponding to the maximum statistical value by one level, and keeping the 0 level unchanged;
a5: obtaining the fire risk abnormal grade judgment model according to the intelligent abnormal grade judgment standard;
s3: and calculating the fire early warning risk value of each electric line according to the abnormal grade of each electric line and the duration of the abnormal grade.
2. The artificial intelligence electrical wiring fire risk real-time warning method as claimed in claim 1, further comprising step a6 after the step a 5:
and performing supervised learning on the multi-dimensional data acquired in real time through the fire risk abnormal level judgment model, updating the result of the supervised learning into the intelligent abnormal level judgment standard at preset time intervals, and re-executing the step A4-A5 to obtain a new fire risk abnormal level judgment model.
3. The real-time early warning method for fire risk of artificial intelligence electrical line of claim 1, wherein in step S1, the multi-dimensional data further comprises smoke information and humidity information.
4. The method according to claim 1, wherein in step S3, the fire warning risk value risk is calculated by:
risk＝(level+counts/n)*10
wherein, level is the abnormal level, n is the real-time data acquisition frequency up to the current day, and counts is the accumulated frequency of the abnormal level up to the current day.
5. The real-time early warning method for fire risk of artificial intelligence electrical wiring of claim 1, further comprising step S4 after the step S3:
and performing risk ranking on each electric line according to the fire early warning risk value, and performing early warning according to the height of the fire early warning risk value.
6. The real-time early warning method for fire risk of artificial intelligence electrical wiring according to claim 5, wherein the step S4 further comprises:
and performing longitudinal and transverse analysis on a time dimension according to the fire early warning risk value, obtaining the risk trend of each electric line and the probability of the same risk in the history through the longitudinal analysis, and obtaining the probability of the same risk in all the electric lines in the early warning area and the maximum risk condition of the whole early warning area through the transverse analysis.
7. The utility model provides a real-time early warning system of artificial intelligence electric line conflagration risk which characterized in that includes:
the system comprises one or more Internet of things acquisition sensing terminals, wherein each Internet of things acquisition sensing terminal comprises a load current detection module, a measuring point temperature detection module, a meter box temperature detection module and an environment temperature detection module and is used for acquiring multi-dimensional data of an electric circuit in real time, the multi-dimensional data comprises measuring point temperature rise, meter box temperature rise and load current, the measuring point temperature rise and the meter box temperature rise are acquired through temperature field acquisition, and the temperature field comprises measuring point temperature, meter box temperature and environment temperature;
the intelligent gateway is in signal connection with the Internet of things acquisition sensing terminal and is used for transmitting the multi-dimensional data acquired in real time;
the data processing server is in data communication with the intelligent gateway and comprises an abnormal grade judging module and a risk processing module;
the abnormal grade judging module is used for judging the abnormality of the multidimensional data to obtain the abnormal grade of each electric circuit;
wherein the obtaining of the abnormal grade of each electric line is to judge the abnormal point and measure the distance between the abnormal point and the normal area of the multi-dimensional data through a fire risk abnormal grade judging model to obtain the abnormal grade of each electric line,
the establishment process of the fire risk abnormal grade judgment model is as follows:
establishing a load temperature rise analysis model according to the corresponding relation between the temperature field and the load current;
training to obtain a fire risk abnormal grade judgment model through an artificial intelligent electrical fire clustering classification algorithm and an artificial intelligent electrical fire grading algorithm according to the load temperature rise analysis model;
a1: carrying out data grouping on big data based on the multi-dimensional data according to the wire specification, the ambient temperature and the load current;
a2: classifying the grouped data according to the temperature rise of the measuring points and the temperature rise of the meter box, and clustering each level into one class;
a3: carrying out abnormal grade standardized classification on each group of classified data according to a preset standard classification value;
a4, counting the statistics of each level of each group of data after abnormal level standardized classification, and readjusting the abnormal levels of the rest levels of each group of data respectively on the basis of the abnormal level corresponding to the maximum statistics in each group of data to obtain an intelligent abnormal level judgment standard;
the abnormality levels are divided from level 0 in the increasing direction of the severity, and the step a4 specifically includes the following steps:
if the abnormal level corresponding to the maximum statistical value is equal to 0, calculating a relative safety range taking the level corresponding to the maximum statistical value as the center according to a preset percentage and the maximum statistical value, adjusting the abnormal level of the level which falls into the relative safety range and has the abnormal level less than or equal to 1 to be 0, and determining the abnormal level of the rest level according to the distance between the levels;
if the abnormal level corresponding to the maximum statistical value is not equal to 0, reducing the abnormal levels corresponding to all levels before the level corresponding to the maximum statistical value by one level, and keeping the 0 level unchanged;
a5: obtaining the fire risk abnormal grade judgment model according to the intelligent abnormal grade judgment standard;
and the risk processing module is used for calculating and obtaining the fire early warning risk value of each electric line according to the abnormal grade of each electric line and the duration of the abnormal grade.
8. The artificial intelligence electrical line fire risk real-time early warning system of claim 7, wherein the internet of things collection and sensing terminal further comprises a smoke detection module and a humidity detection module.
9. The real-time fire risk early warning system for the artificial intelligence electrical lines as claimed in claim 7, further comprising an early warning monitoring platform for performing risk ranking on each electrical line according to the fire early warning risk value and performing early warning according to the level of the fire early warning risk value.
10. The system according to claim 9, wherein the early warning monitoring platform is further configured to perform longitudinal and transverse analysis in a time dimension according to the fire early warning risk value, the longitudinal analysis obtains risk trends of the electrical lines and probabilities of the same risks appearing in histories, and the transverse analysis obtains probabilities of the same risks appearing in all the electrical lines in the early warning area and the maximum risk condition of the whole early warning area.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010005593.6A CN111145517B (en) | 2020-01-03 | 2020-01-03 | Artificial intelligence electric line fire risk real-time early warning method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010005593.6A CN111145517B (en) | 2020-01-03 | 2020-01-03 | Artificial intelligence electric line fire risk real-time early warning method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111145517A CN111145517A (en) | 2020-05-12 |
| CN111145517B true CN111145517B (en) | 2022-03-25 |
Family
ID=70523493
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010005593.6A Active CN111145517B (en) | 2020-01-03 | 2020-01-03 | Artificial intelligence electric line fire risk real-time early warning method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111145517B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112833955A (en) * | 2021-01-08 | 2021-05-25 | 三一汽车起重机械有限公司 | Method for establishing hydraulic oil quality monitoring model, monitoring method, device and system |
| CN112946348A (en) * | 2021-01-29 | 2021-06-11 | 威胜信息技术股份有限公司 | Electrical fire monitoring method and system |
| CN113792477A (en) * | 2021-08-18 | 2021-12-14 | 珠海派诺科技股份有限公司 | Power utilization abnormity identification method, system and device and fire early warning system |
| CN114120598A (en) * | 2021-12-01 | 2022-03-01 | 北京磁浮有限公司 | Rail transit power supply early warning method and device, computer equipment and storage medium |
| CN114186876B (en) * | 2021-12-14 | 2022-11-22 | 广州淏华实业有限公司 | Intelligent electricity safety supervision and electric energy management method and system |
| CN117521978B (en) * | 2024-01-04 | 2024-04-26 | 深圳市一广建设工程有限公司 | Intelligent high-rise building fire control monitored control system |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107564231A (en) * | 2017-09-15 | 2018-01-09 | 山东建筑大学 | Building fire early warning and fire disaster situation assessment system and method based on Internet of Things |
| CN108171450A (en) * | 2018-03-26 | 2018-06-15 | 国网江苏省电力有限公司苏州供电分公司 | Risk evaluating method |
| CN108376310A (en) * | 2018-02-06 | 2018-08-07 | 深圳前海大观信息技术有限公司 | Building fire risk class appraisal procedure |
| CN109166262A (en) * | 2018-09-20 | 2019-01-08 | 上海枫昱能源科技有限公司 | Electric wiring fire alarm Internet of Things perception terminal and system based on Current Mutual Inductance |
| US10509061B2 (en) * | 2017-08-04 | 2019-12-17 | Sensus Spectrum, Llc | Method and system for high temperature detection in electric meters |
| CN111915842A (en) * | 2020-07-02 | 2020-11-10 | 广东技术师范大学 | Abnormity monitoring method and device, computer equipment and storage medium |
-
2020
- 2020-01-03 CN CN202010005593.6A patent/CN111145517B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10509061B2 (en) * | 2017-08-04 | 2019-12-17 | Sensus Spectrum, Llc | Method and system for high temperature detection in electric meters |
| CN107564231A (en) * | 2017-09-15 | 2018-01-09 | 山东建筑大学 | Building fire early warning and fire disaster situation assessment system and method based on Internet of Things |
| CN108376310A (en) * | 2018-02-06 | 2018-08-07 | 深圳前海大观信息技术有限公司 | Building fire risk class appraisal procedure |
| CN108171450A (en) * | 2018-03-26 | 2018-06-15 | 国网江苏省电力有限公司苏州供电分公司 | Risk evaluating method |
| CN109166262A (en) * | 2018-09-20 | 2019-01-08 | 上海枫昱能源科技有限公司 | Electric wiring fire alarm Internet of Things perception terminal and system based on Current Mutual Inductance |
| CN111915842A (en) * | 2020-07-02 | 2020-11-10 | 广东技术师范大学 | Abnormity monitoring method and device, computer equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111145517A (en) | 2020-05-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111145517B (en) | Artificial intelligence electric line fire risk real-time early warning method and system | |
| CN110763929A (en) | Intelligent monitoring and early warning system and method for convertor station equipment | |
| CN112505549B (en) | New energy automobile battery abnormity detection method based on isolated forest algorithm | |
| CN106372735B (en) | Relay protection state evaluation method | |
| CN116304766B (en) | Multi-sensor-based quick assessment method for state of switch cabinet | |
| CN104020754A (en) | Method for enabling state monitoring information of transformer station primary main equipment to access to regulation and control system | |
| CN111157045B (en) | Artificial intelligence electric line abnormal risk degree monitoring method and system | |
| CN117040138B (en) | Power distribution cabinet operation dynamic safety evaluation system | |
| CN110866616A (en) | Fault early warning method and device for secondary equipment of transformer substation | |
| CN112328961A (en) | On-line monitoring device quality evaluation system based on fault tree and Bayesian network | |
| CN109491339B (en) | Big data-based substation equipment running state early warning system | |
| CN113111596A (en) | Carbon peak-reaching real-time visual measuring and calculating method based on electric brain center | |
| CN115995886A (en) | Power monitoring system of power equipment | |
| CN112417763A (en) | Defect diagnosis method, device and equipment for power transmission line and storage medium | |
| CN114928168A (en) | Offshore platform unmanned data edge computing device | |
| CN117349624A (en) | Electric power energy monitoring method, system, terminal equipment and storage medium | |
| CN112488361A (en) | Transformer area low voltage prediction method and device based on big data | |
| CN111158338A (en) | Chemical risk monitoring method based on principal component analysis | |
| CN115358355B (en) | Method and device for judging main transformer oil temperature gauge and top layer oil temperature abnormity | |
| CN2570789Y (en) | Device for monitoring moisture in transformer oil | |
| CN115829168A (en) | Remote coordination operation and maintenance management system for new energy equipment based on data analysis | |
| CN115600879A (en) | Circuit breaker abnormity early warning method, system and related device | |
| CN115511156A (en) | Main transformer overload early warning method and system based on load prediction | |
| CN115358336A (en) | Power utilization abnormity detection method and device and electronic equipment | |
| CN114967628A (en) | Health food production management and control system that disinfects based on 5G intelligence mill |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |