Fire-fighting early warning management system and method

Abstract

The invention discloses a fire-fighting early warning management system and method, and the system comprises a monitoring module which is used for obtaining and analyzing real-time feature monitoring data of preset monitoring equipment, determining abnormal equipment, abnormal feature monitoring data and corresponding abnormal change features, and generating an early warning work order; the determination module is used for determining historical abnormal feature monitoring data, fire behavior change features, historical false abnormal feature monitoring data and hidden danger change features based on the historical fire behavior logs and the historical hidden danger logs; the analysis module is used for performing similarity analysis to obtain a first similarity and a second similarity; and the notification module is used for setting the credibility of the early warning work order according to the first similarity and the second similarity and generating a corresponding notification instruction, and by improving the judgment accuracy of the early warning work order, whether the early warning work order is a false alarm or a real fire behavior is accurately judged, and the accuracy and the early warning efficiency of fire fighting early warning are further improved.

Description

Technical Field

[0001] This application relates to the technical field of fire warning, and in particular to a fire warning management system and method. Background Art

[0002] When a device issues a fire alarm signal, the platform generates a corresponding warning work order. The inspectors in the area where the device is located and the enterprise administrators receive notifications. The area inspectors arrive at the scene to confirm whether it is a false alarm or a real fire, thereby judging the accuracy of the warning work order and issuing corresponding notifications. The above management method requires personnel to arrive at the scene for confirmation, which wastes time and cannot guarantee the safety of personnel, reducing the fire warning efficiency. Therefore, how to reduce the personnel cost and quickly judge the accuracy of the warning work order and improve the fire warning efficiency is a technical problem that needs to be solved at present. Summary of the Invention

[0003] To solve the above technical problems, this application provides a fire warning management system and method. By obtaining abnormal devices, abnormal feature monitoring data, and corresponding abnormal change features, and performing similarity analysis with the historical abnormal feature monitoring data and corresponding fire change features, historical pseudo-abnormal feature monitoring data and corresponding hidden danger change features of the same abnormal device, the first similarity and the second similarity are determined, and the credibility of the warning work order is obtained, thereby improving the judgment accuracy of the warning work order, reducing the personnel cost, and improving the fire warning efficiency and warning accuracy.

[0004] In some embodiments of this application, a fire warning management system is provided, including:

[0005] A monitoring module, configured to obtain and analyze the real-time feature monitoring data of preset monitoring devices, determine abnormal devices, abnormal feature monitoring data, and corresponding abnormal change features, and generate a warning work order;

[0006] A determination module, configured to obtain the historical fire log and historical hidden danger log of the abnormal device, determine the historical abnormal feature monitoring data and corresponding fire change features of the corresponding abnormal device based on the historical fire log, and determine the historical pseudo-abnormal feature monitoring data and corresponding hidden danger change features of the corresponding abnormal device based on the historical hidden danger log;

[0007] An analysis module, configured to perform similarity analysis on the abnormal feature monitoring data and corresponding abnormal change features of the abnormal device in the warning work order with the historical abnormal feature monitoring data and corresponding fire change features, historical pseudo-abnormal feature monitoring data and corresponding hidden danger change features of the corresponding abnormal device, to obtain the first similarity and the second similarity;

[0008] A notification module, configured to set the credibility of the warning work order according to the first similarity and the second similarity, and generate a corresponding notification instruction.

[0009] In some embodiments of the present application, determining abnormal devices, abnormal feature monitoring data, and corresponding abnormal change features includes:

[0010] Presetting a plurality of fire evaluation indicators in advance;

[0011] Obtaining real-time monitoring data of a preset monitoring device within a preset time period, and obtaining the historical correlation degree between each real-time monitoring data and each fire evaluation indicator. Setting the real-time monitoring data with a historical correlation degree greater than a preset correlation degree threshold as real-time feature monitoring data;

[0012] Taking the difference between the real-time feature monitoring data within the preset time period and the standard monitoring data of the corresponding fire evaluation indicator to obtain the data difference between the real-time feature monitoring data and the standard monitoring data;

[0013] Setting the real-time feature monitoring data with a data difference greater than a preset difference threshold as abnormal feature monitoring data, and setting the preset monitoring device with abnormal feature monitoring data as an abnormal device;

[0014] Determining the initial time node when the data difference of each abnormal feature monitoring data is greater than the preset difference threshold. Taking the initial time node as the starting point, setting data acquisition nodes at a preset time interval, obtaining the corresponding abnormal feature monitoring data according to the data acquisition nodes, and constructing a data change curve for each abnormal feature monitoring data;

[0015] Setting the change trend, change rate, and change magnitude value of the abnormal feature monitoring data in each data change curve as the abnormal change features of the corresponding abnormal feature monitoring data.

[0016] In some embodiments of the present application, determining the historical abnormal feature monitoring data and corresponding fire change features of the corresponding abnormal devices based on historical fire logs, and determining historical pseudo-abnormal feature monitoring data and corresponding hidden danger change features based on historical hidden danger logs includes:

[0017] Obtaining the historical fire logs of the abnormal devices, screening out the historical abnormal feature monitoring data in each historical fire log, and determining the first abnormal time period of the historical abnormal feature monitoring data;

[0018] Taking the first abnormal time period of each historical abnormal feature monitoring data as the time reference line, setting data acquisition nodes at a preset time interval, obtaining the corresponding historical abnormal feature monitoring data according to each data acquisition node, and mapping it to the corresponding time reference line to obtain the first historical data change curve of each historical abnormal feature detection data;

[0019] Set the change trend, change rate, and change magnitude of the historical anomaly feature monitoring data in each first historical data change curve as the fire change feature of the corresponding historical anomaly feature monitoring data;

[0020] Analyze all historical fire logs of the abnormal device, and sequentially obtain the fire change features of the historical anomaly feature monitoring data in each historical fire log of the corresponding abnormal device;

[0021] Obtain the historical hidden danger logs of the abnormal device, screen out the historical pseudo-anomaly feature monitoring data in each historical hidden danger log, and determine the second abnormal time period of the historical pseudo-anomaly feature monitoring data;

[0022] Taking the second abnormal time period of each historical pseudo-anomaly feature monitoring data as the time reference line, setting data acquisition nodes at a preset time interval, obtaining the corresponding historical pseudo-anomaly feature monitoring data according to each data acquisition node, and mapping it to the corresponding time reference line to obtain the second historical data change curve of each historical pseudo-anomaly feature detection data;

[0023] Set the change trend, change rate, and change magnitude of the historical pseudo-anomaly feature monitoring data in each second historical data change curve as the hidden danger change feature of the corresponding historical pseudo-anomaly feature monitoring data;

[0024] Analyze all historical hidden danger logs of the abnormal device, and sequentially obtain the hidden danger change features of the historical pseudo-anomaly feature monitoring data in each historical hidden danger log of the current abnormal device.

[0025] In some embodiments of the present application, before performing the similarity analysis, it further includes:

[0026] Compare all the anomaly feature monitoring data of each abnormal device in the warning work order with the historical anomaly feature monitoring data of each historical fire log of the corresponding abnormal device to obtain the occurrence ratio of all the anomaly feature monitoring data in the historical anomaly feature monitoring data of each historical fire log;

[0027] If the occurrence ratio is less than the preset occurrence ratio threshold, then eliminate the corresponding historical fire log and the corresponding historical anomaly feature monitoring data;

[0028] Compare all the anomaly feature monitoring data of each abnormal device in the warning work order with the historical pseudo-anomaly feature monitoring data of each historical hidden danger log of the corresponding abnormal device to obtain the occurrence ratio of all the anomaly feature monitoring data in the historical pseudo-anomaly feature monitoring data of each historical hidden danger log;

[0029] If the occurrence ratio is less than the preset occurrence ratio threshold, then eliminate the corresponding historical hidden danger log and the corresponding historical pseudo-anomaly feature monitoring data.

[0030] In some embodiments of the present application, similarity analysis is performed, including:

[0031] Comparing the change trends of the historical abnormal feature monitoring data of the remaining historical fire logs of the same abnormal device with the corresponding abnormal feature monitoring data of the corresponding abnormal device in the early warning work order to obtain the first change trend difference degree;

[0032] Marking the abnormal feature monitoring data with the first change trend difference degree less than the preset difference degree, and calculating the difference degree difference;

[0033] Generating the first change trend similarity of the change trends between the marked abnormal feature monitoring data of the same abnormal device and the historical abnormal feature monitoring data of the corresponding same historical fire log according to the number of marked abnormal feature monitoring data and the difference degree difference;

[0034] The calculation formula of the first change trend similarity is:

[0035]

[0036] Wherein, B1 is the first change trend similarity, n1 is the number of marks of the abnormal feature monitoring data with the first change trend difference degree less than the preset difference degree, m is the total number of abnormal feature monitoring data of the abnormal device, △Ci is the difference degree difference between the i-th marked abnormal feature monitoring data and the corresponding historical abnormal feature monitoring data, di is the weight coefficient of the i-th marked abnormal feature monitoring data, and c0 is the first similarity conversion coefficient;

[0037] Re-removing the historical fire logs with the first change trend similarity less than the preset change trend similarity threshold, and obtaining the similar time periods of the change trends between the historical abnormal feature monitoring data of each remaining historical fire log and the marked abnormal feature monitoring data in the early warning work order, wherein the similar time periods include multiple data acquisition nodes;

[0038] Obtaining the change amount values and change rates at each data acquisition node of the marked abnormal feature monitoring data and the corresponding historical abnormal feature monitoring data in the corresponding similar time period, and taking the difference to obtain the change amount value difference and change rate difference at each data acquisition node;

[0039] Generating the first change amount value sub-similarity and the first change rate sub-similarity at the corresponding data acquisition nodes between the marked abnormal feature monitoring data and the corresponding historical abnormal feature monitoring data according to the change amount value difference and change rate difference at each data acquisition node;

[0040] Monitor the multiple first change amount sub-similarities and first change rate sub-similarities of the marked abnormal feature monitoring data and the corresponding historical abnormal feature monitoring data in the corresponding similar time periods, and generate a first similarity by combining the first change trend similarity;

[0041] The calculation formula of the first similarity is:

[0042]

[0043] Wherein, X1 is the first similarity, h3 is the weight coefficient of the change trend similarity, d 1 is the weight coefficient of the abnormal feature monitoring data of the first marked one, d n1 is the weight coefficient of the abnormal feature monitoring data of the n1th marked one, u1 1 is the total number of data acquisition nodes of the abnormal feature monitoring data of the first marked one and the corresponding historical abnormal feature detection data in the similar time period, L1 s,1 is the first change amount sub-similarity of the abnormal feature monitoring data of the first marked one and the corresponding historical abnormal feature detection data at the s-th data acquisition node in the similar time period, h1 is the weight coefficient of the change amount sub-similarity, h2 is the weight coefficient of the change rate sub-similarity, K1 s,1 is the first change rate sub-similarity of the abnormal feature monitoring data of the first marked one and the corresponding historical abnormal feature detection data at the s-th data acquisition node in the similar time period, g1 n1 is the total number of data acquisition nodes of the abnormal feature monitoring data of the n1th marked one and the corresponding historical abnormal feature detection data in the similar time period, L1 v,n1 is the first change amount sub-similarity of the abnormal feature monitoring data of the n1th marked one and the corresponding historical abnormal feature detection data at the v-th data acquisition node in the similar time period, K1 v,n1 is the first change rate sub-similarity of the abnormal feature monitoring data of the n1th marked one and the corresponding historical abnormal feature detection data at the v-th data acquisition node in the similar time period.

[0044] In some embodiments of the present application, for similarity analysis, it further includes:

[0045] Compare the change trends of the historical pseudo-abnormal feature monitoring data of the remaining historical hidden danger logs of the same abnormal device with the corresponding abnormal feature monitoring data of the corresponding abnormal device in the early warning work order to obtain a second change trend difference degree;

[0046] Mark the abnormal feature monitoring data with the second change trend difference degree less than the preset difference degree, and calculate the difference degree difference;

[0047] Generate a second change trend similarity of the change trend between the abnormal feature monitoring data marked for the same abnormal device and the historical pseudo-abnormal feature monitoring data of the corresponding same historical hidden danger log according to the number of abnormal feature monitoring data marked and the difference degree difference value.

[0048] The calculation formula for the second change trend similarity is as follows:

[0049]

[0050] Where B2 is the second change trend similarity, n2 is the number of marks of abnormal feature monitoring data with the second change trend difference less than the preset difference degree, and △Wi is the difference degree difference value between the abnormal feature monitoring data of the i-th mark and the corresponding historical pseudo-abnormal feature monitoring data.

[0051] Eliminate the historical hidden danger logs with the second change trend similarity less than the preset change trend similarity threshold, and obtain the similar time periods of the change trends between the historical pseudo-abnormal feature monitoring data of each remaining historical hidden danger log and the abnormal feature monitoring data marked in the warning work order, where the similar time periods include multiple data acquisition nodes.

[0052] Obtain the change amount value and change rate at each data acquisition node of the marked abnormal feature monitoring data and the corresponding historical pseudo-abnormal feature monitoring data in the corresponding similar time period, and perform a difference operation to obtain the change amount value difference and change rate difference at each data acquisition node.

[0053] Generate a second change amount value sub-similarity and a second change rate sub-similarity at the corresponding data acquisition node of the marked abnormal feature monitoring data and the corresponding historical pseudo-abnormal feature monitoring data according to the change amount value difference and change rate difference at each data acquisition node.

[0054] Generate a second similarity according to multiple second change amount value sub-similarities and second change rate sub-similarities of the marked abnormal feature monitoring data and the corresponding historical pseudo-abnormal feature monitoring data in the corresponding similar time period, and combine with the second change trend similarity.

[0055] The calculation formula for the second similarity is as follows:

[0056]

[0057] Where X2 is the second similarity, u2 1 is the total number of data acquisition nodes of the abnormal feature monitoring data of the first mark and the corresponding historical pseudo-abnormal feature detection data in the similar time period, L2 z,1 is the second change amount value sub-similarity of the abnormal feature monitoring data of the first mark and the corresponding historical pseudo-abnormal feature detection data at the z-th data acquisition node in the similar time period, K2z,1 The second change rate sub-similarity of the abnormal feature monitoring data of the first label and the corresponding historical pseudo-abnormal feature detection data at the z-th data acquisition node in a similar time period, g2 n2 The total number of data acquisition nodes of the abnormal feature monitoring data of the n2-th label and the corresponding historical pseudo-abnormal feature detection data in a similar time period, L2 r,n2 The second change amount value sub-similarity of the abnormal feature monitoring data of the n2-th label and the corresponding historical pseudo-abnormal feature detection data at the r-th data acquisition node in a similar time period, K2 r,n2 The second change rate sub-similarity of the abnormal feature monitoring data of the n2-th label and the corresponding historical pseudo-abnormal feature detection data at the r-th data acquisition node in a similar time period.

[0058] In some embodiments of the present application, the credibility of the warning work order is set according to the first similarity and the second similarity, and the corresponding notification instruction is generated, including:

[0059] A similarity threshold and a credibility threshold are preset in advance;

[0060] If the first similarity is greater than the similarity threshold and the second similarity is less than the similarity threshold, the credibility of the warning work order is set to be greater than the credibility threshold, and a fire alarm notification instruction is generated according to the warning work order. The fire alarm notification instruction includes triggering a fire protection linkage and an emergency plan, and sending fire information to the fire department;

[0061] If the second similarity is greater than the similarity threshold and the first similarity is less than the similarity threshold, the credibility of the warning work order is set to be less than the credibility threshold, and a hidden danger notification instruction is generated according to the warning work order. The hidden danger notification instruction includes reminding the inspection personnel to perform operation and maintenance;

[0062] In some embodiments of the present application, it further includes a fire warning management method:

[0063] Obtain the real-time feature monitoring data of the preset monitoring equipment and analyze it to determine the abnormal equipment, abnormal feature monitoring data and the corresponding abnormal change features, and generate a warning work order;

[0064] Obtain the historical fire log and historical hidden danger log of the abnormal equipment, determine the historical abnormal feature monitoring data and the corresponding fire change features of the corresponding abnormal equipment based on the historical fire log, and determine the historical pseudo-abnormal feature monitoring data and the corresponding hidden danger change features of the corresponding abnormal equipment based on the historical hidden danger log;

[0065] Analyze the similarity between the abnormal feature monitoring data of abnormal devices in the early warning work order and the corresponding abnormal change features, respectively, and the historical abnormal feature monitoring data of the corresponding abnormal devices and the corresponding fire change features, as well as the historical pseudo-abnormal feature monitoring data and the corresponding hidden danger change features, to obtain the first similarity and the second similarity;

[0066] Set the credibility of the early warning work order according to the first similarity and the second similarity, and generate the corresponding notification instruction.

[0067] Compared with the prior art, a fire early warning management system and method according to an embodiment of the present application have the beneficial effects that:

[0068] By obtaining abnormal devices, abnormal feature monitoring data, and the corresponding abnormal change features, and analyzing the similarity with the historical abnormal feature monitoring data of the same abnormal device and the corresponding fire change features, as well as the historical pseudo-abnormal feature monitoring data and the corresponding hidden danger change features, the first similarity and the second similarity are determined, and the credibility of the early warning work order is obtained, thereby improving the judgment accuracy of the early warning work order, reducing the personnel cost, and improving the fire early warning efficiency and warning accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 is a schematic diagram of a fire early warning management system in a preferred embodiment of an embodiment of the present application;

[0070] Figure 2 is a schematic flowchart of a fire early warning management system in a preferred embodiment of an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0071] The following further describes in detail the specific embodiments of the present application in conjunction with the drawings and embodiments. The following embodiments are used to illustrate the present application, but are not used to limit the scope of the present application.

[0072] As Figure 1 shown, a fire early warning management system according to an embodiment of the present application includes:

[0073] A monitoring module, configured to obtain real-time feature monitoring data of preset monitoring devices, analyze the data, determine abnormal devices, abnormal feature monitoring data, and corresponding abnormal change features, and generate an early warning work order;

[0074] A determination module, configured to obtain the historical fire log and historical hidden danger log of the abnormal device, determine the historical abnormal feature monitoring data and the corresponding fire change features of the corresponding abnormal device based on the historical fire log, and determine the historical pseudo-abnormal feature monitoring data and the corresponding hidden danger change features of the corresponding abnormal device based on the historical hidden danger log;

[0075] An analysis module is configured to perform similarity analysis on the abnormal feature monitoring data of abnormal devices in the early warning work order and the corresponding abnormal change features respectively with the historical abnormal feature monitoring data of the corresponding abnormal devices and the corresponding fire change features, and the historical pseudo-abnormal feature monitoring data of the corresponding abnormal devices and the corresponding hidden danger change features, to obtain a first similarity and a second similarity;

[0076] A notification module is configured to set the credibility of the early warning work order according to the first similarity and the second similarity, and generate a corresponding notification instruction.

[0077] In this embodiment, the preset monitoring devices include temperature sensors, smoke concentration sensors, etc. The preset monitoring devices are devices that monitor in real time whether a fire has occurred, and the abnormal devices are devices that send out fire alarm signals.

[0078] In this embodiment, the abnormal feature monitoring data refers to the real-time monitoring data associated with the fire evaluation index and the data difference between the real-time monitoring data and the standard monitoring data is greater than the preset data difference. The abnormal change feature refers to the change trend, change rate, and change magnitude of the abnormal feature monitoring data after the initial time node when the data difference is greater than the preset data difference.

[0079] In this embodiment, the historical fire log refers to the historical monitoring log of the corresponding preset monitoring device when a real fire occurs, including but not limited to the historical abnormal monitoring data of the corresponding abnormal device and the corresponding change features when a real fire occurs, or the simulated abnormal monitoring data of the corresponding abnormal device and the simulated change features when simulating various real fire situations. The historical hidden danger log refers to the historical monitoring log when the device sends out a fire alarm signal but is confirmed as a false alarm, including but not limited to the historical pseudo-abnormal monitoring data of the corresponding abnormal device and the corresponding change features when a false alarm occurs, and the simulated pseudo-abnormal monitoring data of the corresponding abnormal device and the simulated change features when simulating different false alarm situations.

[0080] In this embodiment, the historical fire log and the historical hidden danger log include the monitoring data and the corresponding change features of the corresponding preset monitoring device in the face of multiple real fire or false alarm situations. By comparing the abnormal feature monitoring data and the abnormal change features of the preset monitoring device with the historical abnormal feature monitoring data and the corresponding fire change features, and the historical pseudo-abnormal feature monitoring data and the corresponding hidden danger change features of each historical fire log or historical hidden danger log for similarity analysis, the credibility of the current early warning work order is determined and a notification instruction is generated, thereby improving the fire warning efficiency and accuracy.

[0081] In some embodiments of the present application, determining the abnormal device, the abnormal feature monitoring data, and the corresponding abnormal change features includes:

[0082] Pre-setting a plurality of fire evaluation indexes;

[0083] Obtain the real-time monitoring data of the preset monitoring device within the preset time period, and obtain the historical correlation degree between each real-time monitoring data and each fire evaluation index. Set the real-time monitoring data with a historical correlation degree greater than the preset correlation degree threshold as the real-time characteristic monitoring data;

[0084] Subtract the standard monitoring data of the real-time characteristic monitoring data within the preset time period from the corresponding fire evaluation index to obtain the data difference between the real-time characteristic monitoring data and the standard monitoring data;

[0085] Set the real-time characteristic monitoring data with a data difference greater than the preset difference threshold as the abnormal characteristic monitoring data, and set the preset monitoring device with abnormal characteristic monitoring data as the abnormal device;

[0086] Determine the initial time node when the data difference of each abnormal characteristic monitoring data is greater than the preset difference threshold. Starting from the initial time node, set data acquisition nodes at preset time intervals, obtain the corresponding abnormal characteristic monitoring data according to the data acquisition nodes, and construct the data change curve of each abnormal characteristic monitoring data;

[0087] Set the change trend, change rate, and change magnitude value of the abnormal characteristic monitoring data in each data change curve as the abnormal change characteristics of the corresponding abnormal characteristic monitoring data.

[0088] In this embodiment, the fire evaluation index refers to the evaluation index when a fire occurs, which is set in advance according to expert experience or data verification. The historical correlation degree refers to the influence degree of the real-time monitoring data on each fire evaluation index. When the influence degree is large, the corresponding historical correlation degree is large, that is, the real-time monitoring data is the characteristic monitoring data of the corresponding fire evaluation index.

[0089] In this embodiment, by determining the abnormal characteristic monitoring data, abnormal devices, and obtaining the abnormal change characteristics of the abnormal characteristic monitoring data, it lays a foundation for setting the credibility of the subsequent warning work order, improves the accuracy of the warning work order, thereby reducing the false alarm probability and timely warning and handling of real fires.

[0090] In some embodiments of the present application, based on the historical fire log, determine the historical abnormal characteristic monitoring data and the corresponding fire change characteristics of the corresponding abnormal device, and based on the historical hidden danger log, determine the historical pseudo-abnormal characteristic monitoring data and the corresponding hidden danger change characteristics, including:

[0091] Obtain the historical fire log of the abnormal device, screen out the historical abnormal characteristic monitoring data in each historical fire log, and determine the first abnormal time period of the historical abnormal characteristic monitoring data;

[0092] Taking the first abnormal period of each historical abnormal feature monitoring data as the time reference line, setting data acquisition nodes at a preset time interval, obtaining the corresponding historical abnormal feature monitoring data according to each data acquisition node, and mapping it to the corresponding time reference line to obtain the first historical data change curve of each historical abnormal feature detection data;

[0093] Setting the change trend, change rate, and change amount value of the historical abnormal feature monitoring data in each first historical data change curve as the fire change characteristics of the corresponding historical abnormal feature monitoring data;

[0094] Analyzing all historical fire logs of the abnormal device to obtain the fire change characteristics of the historical abnormal feature monitoring data in each historical fire log of the corresponding abnormal device in turn;

[0095] Obtaining the historical hidden danger logs of the abnormal device, screening out the historical pseudo-abnormal feature monitoring data in each historical hidden danger log, and determining the second abnormal period of the historical pseudo-abnormal feature monitoring data;

[0096] Taking the second abnormal period of each historical pseudo-abnormal feature monitoring data as the time reference line, setting data acquisition nodes at a preset time interval, obtaining the corresponding historical pseudo-abnormal feature monitoring data according to each data acquisition node, and mapping it to the corresponding time reference line to obtain the second historical data change curve of each historical pseudo-abnormal feature detection data;

[0097] Setting the change trend, change rate, and change amount value of the historical pseudo-abnormal feature monitoring data in each second historical data change curve as the hidden danger change characteristics of the corresponding historical pseudo-abnormal feature monitoring data;

[0098] Analyzing all historical hidden danger logs of the abnormal device to obtain the hidden danger change characteristics of the historical pseudo-abnormal feature monitoring data in each historical hidden danger log of the current abnormal device in turn.

[0099] In this embodiment, the first abnormal period refers to the historical initial node where the data difference between the historical abnormal feature monitoring data and the corresponding standard monitoring data in the corresponding historical fire log is greater than the preset difference threshold to the end node where the data difference is less than the preset difference threshold, and the second abnormal period refers to the historical initial node where the data difference between the historical pseudo-abnormal feature monitoring data and the corresponding standard monitoring data in the corresponding historical hidden danger log is greater than the preset difference threshold to the end node where the data difference is less than the preset difference threshold.

[0100] In this embodiment, by determining the historical abnormal feature monitoring data in each historical fire log and the fire change characteristics of the historical abnormal feature monitoring data, as well as the hidden danger change characteristics of the historical pseudo-abnormal feature monitoring data in each historical hidden danger log, a foundation is laid for subsequently judging the abnormal change characteristics of the abnormal feature monitoring data of the abnormal device, the accuracy of judging the abnormal change characteristics is improved, it is accurately judged whether it is a real fire or a false alarm, and the accuracy of fire warning is improved.

[0101] In some embodiments of the present application, before performing the similarity analysis, it further includes:

[0102] Compare all the abnormal feature monitoring data of each abnormal device in the warning work order with the historical abnormal feature monitoring data of each historical fire log of the corresponding abnormal device to obtain the occurrence ratio of all the abnormal feature monitoring data in the historical abnormal feature monitoring data of each historical fire log;

[0103] If the occurrence ratio is less than the preset occurrence ratio threshold, then eliminate the corresponding historical fire log and the corresponding historical abnormal feature monitoring data;

[0104] Compare all the abnormal feature monitoring data of each abnormal device in the warning work order with the historical pseudo-abnormal feature monitoring data of each historical hidden danger log of the corresponding abnormal device to obtain the occurrence ratio of all the abnormal feature monitoring data in the historical pseudo-abnormal feature monitoring data of each historical hidden danger log;

[0105] If the occurrence ratio is less than the preset occurrence ratio threshold, then eliminate the corresponding historical hidden danger log and the corresponding historical pseudo-abnormal feature monitoring data.

[0106] In this embodiment, the occurrence ratio refers to the ratio of the quantity of all the abnormal feature monitoring data to the occurrence quantity of all the abnormal feature monitoring data in the historical abnormal feature monitoring data of each historical fire log and the historical pseudo-abnormal feature monitoring data of each historical hidden danger log. The more the quantity of all the abnormal feature monitoring data appears in the historical abnormal feature monitoring data of each historical fire log and the historical pseudo-abnormal feature monitoring data of each historical hidden danger log, the more similar it may be to the corresponding historical fire log and historical hidden danger log, that is, retain the corresponding historical fire log and historical hidden danger log and perform similarity analysis.

[0107] In this embodiment, by eliminating and retaining the corresponding historical fire log and the corresponding historical abnormal feature monitoring data, historical hidden danger log and the corresponding historical pseudo-abnormal feature monitoring data, and performing subsequent similarity analysis, the quantity analysis amount and processing amount are reduced, and the similarity analysis efficiency and the judgment efficiency of the credibility of the warning work order are improved.

[0108] In some embodiments of the present application, performing the similarity analysis includes:

[0109] Compare the change trends of the historical abnormal feature monitoring data of the remaining historical fire logs of the same abnormal device with the corresponding abnormal feature monitoring data of the corresponding abnormal device in the early warning work order to obtain the first change trend difference degree;

[0110] Mark the abnormal feature monitoring data with the first change trend difference degree less than the preset difference degree, and calculate the difference degree difference;

[0111] Generate the first change trend similarity of the change trends of the marked abnormal feature monitoring data of the same abnormal device and the historical abnormal feature monitoring data of the corresponding same historical fire log according to the number of marked abnormal feature monitoring data and the difference degree difference;

[0112] The calculation formula of the first change trend similarity is:

[0113]

[0114] Wherein, B1 is the first change trend similarity, n1 is the number of marks of the abnormal feature monitoring data with the first change trend difference degree less than the preset difference degree, m is the total number of abnormal feature monitoring data of the abnormal device, △Ci is the difference degree difference between the i-th marked abnormal feature monitoring data and the corresponding historical abnormal feature monitoring data, di is the weight coefficient of the i-th marked abnormal feature monitoring data, and c0 is the first similarity conversion coefficient;

[0115] Re-remove the historical fire logs with the first change trend similarity less than the preset change trend similarity threshold, and obtain the similar time periods of the change trends of the historical abnormal feature monitoring data of each remaining historical fire log and the marked abnormal feature monitoring data in the early warning work order, where the similar time periods include multiple data acquisition nodes;

[0116] Obtain the change amount value and change rate at each data acquisition node of the marked abnormal feature monitoring data and the corresponding historical abnormal feature monitoring data in the corresponding similar time period, and perform subtraction to obtain the change amount value difference and change rate difference at each data acquisition node;

[0117] Generate the first change amount value sub-similarity and the first change rate sub-similarity at the corresponding data acquisition node of the marked abnormal feature monitoring data and the corresponding historical abnormal feature monitoring data according to the change amount value difference and change rate difference at each data acquisition node;

[0118] Generate the first similarity according to the multiple first change amount value sub-similarities and first change rate sub-similarities of the marked abnormal feature monitoring data and the corresponding historical abnormal feature monitoring data in the corresponding similar time period, and combine the first change trend similarity;

[0119] The calculation formula for the first similarity is as follows:

[0120]

[0121] where X1 is the first similarity, h3 is the weight coefficient of the change trend similarity, d 1 is the weight coefficient of the abnormal feature monitoring data of the 1st tag, d n1 is the weight coefficient of the abnormal feature monitoring data of the n1th tag, u1 1 is the total number of data acquisition nodes of the abnormal feature monitoring data of the 1st tag and the corresponding historical abnormal feature detection data at similar time periods, L1 s,1 is the first change amount value sub-similarity of the abnormal feature monitoring data of the 1st tag and the corresponding historical abnormal feature detection data at the s-th data acquisition node at similar time periods, h1 is the weight coefficient of the change amount value sub-similarity, h2 is the weight coefficient of the change rate sub-similarity, K1 s,1 is the first change rate sub-similarity of the abnormal feature monitoring data of the 1st tag and the corresponding historical abnormal feature detection data at the s-th data acquisition node at similar time periods, g1 n1 is the total number of data acquisition nodes of the abnormal feature monitoring data of the n1th tag and the corresponding historical abnormal feature detection data at similar time periods, L1 v,n1 is the first change amount value sub-similarity of the abnormal feature monitoring data of the n1th tag and the corresponding historical abnormal feature detection data at the v-th data acquisition node at similar time periods, K1 v,n1 is the first change rate sub-similarity of the abnormal feature monitoring data of the n1th tag and the corresponding historical abnormal feature detection data at the v-th data acquisition node at similar time periods.

[0122] In this embodiment, s = 1, 2,..., u1, v = 1, 2,..., g1.

[0123] In this embodiment, L1 s,1 = △l1 s,1 × c1, △l1 s,1 is the change amount value difference of the abnormal feature monitoring data of the 1st tag and the corresponding historical abnormal feature detection data at the s-th data acquisition node at similar time periods, c1 is the second similarity conversion coefficient. The second similarity conversion coefficient refers to the coefficient that converts the change amount value difference into the same dimension as the first change amount value sub-similarity. When the change amount value difference is smaller, the first change amount value sub-similarity is larger, and vice versa, K1 s,1 = △k1 s,1 × c2, △k1 s,1is the difference in the change rate of the abnormal feature monitoring data of the first - marked node and the corresponding historical abnormal feature detection data at the s - th data acquisition node in a similar period. c2 is the third similarity conversion coefficient, which is a coefficient used to convert the difference in the change rate into the same dimension as the first change - rate sub - similarity. When the difference in the change rate is smaller, the first change - rate sub - similarity is larger, and vice versa.

[0124] In this embodiment, L1 v,n1 =△l1 v,n1 ×c1, △l1 v,n1 is the difference in the change amount of the abnormal feature monitoring data of the n1 - th marked node and the corresponding historical abnormal feature detection data at the v - th data acquisition node in a similar period. K1 v,n1 =△k1 v,n1 ×c2, △k1 v,n1 is the difference in the change rate of the abnormal feature monitoring data of the n1 - th marked node and the corresponding historical abnormal feature detection data at the v - th data acquisition node in a similar period.

[0125] In this embodiment, the difference - degree difference = preset difference degree - first change - trend difference degree. The first similarity conversion coefficient is a coefficient used to convert the difference - degree difference into the same dimension as the change - trend similarity. When the difference - degree difference is smaller, the converted first change - trend similarity is larger, and vice versa.

[0126] In this embodiment, the first similarity refers to the similarity degree between the abnormal change characteristics of all abnormal feature monitoring data of each abnormal device in the early - warning work order and the fire - situation change characteristics of the corresponding historical abnormal feature monitoring data in each historical fire - situation log of the corresponding abnormal device. Whether it is a real fire is judged according to the first similarity, which improves the accuracy and efficiency of fire - fighting early - warning.

[0127] In some embodiments of the present application, when performing similarity analysis, it further includes:

[0128] Comparing the change trend of the historical pseudo - abnormal feature monitoring data of the remaining historical hidden - danger logs of the same abnormal device with the corresponding abnormal feature monitoring data of the corresponding abnormal device in the early - warning work order to obtain a second change - trend difference degree;

[0129] Marking the abnormal feature monitoring data with the second change - trend difference degree less than the preset difference degree, and calculating the difference - degree difference;

[0130] Generating a second change - trend similarity of the change trend between the marked abnormal feature monitoring data of the same abnormal device and the historical pseudo - abnormal feature monitoring data of the corresponding same historical hidden - danger log according to the number of the marked abnormal feature monitoring data and the difference - degree difference;

[0131] The calculation formula for the similarity of the second change trend is as follows:

[0132]

[0133] Among them, B2 is the similarity of the second change trend, n2 is the number of marks of the abnormal feature monitoring data with the difference degree of the second change trend less than the preset difference degree, and △Wi is the difference value of the difference degree between the abnormal feature monitoring data of the i-th mark and the corresponding historical pseudo-abnormal feature monitoring data;

[0134] Eliminate the historical hidden danger logs with the similarity of the second change trend less than the preset change trend similarity threshold, and obtain the similar time periods of the change trends of the historical pseudo-abnormal feature monitoring data of each remaining historical hidden danger log and the abnormal feature monitoring data marked in the early warning work order. Among them, the similar time period includes multiple data acquisition nodes;

[0135] Obtain the change amount value and change rate of the marked abnormal feature monitoring data and the corresponding historical pseudo-abnormal feature monitoring data at each data acquisition node in the corresponding similar time period, and perform a difference operation to obtain the change amount value difference and change rate difference at each data acquisition node;

[0136] Generate the second change amount value sub-similarity and the second change rate sub-similarity of the marked abnormal feature monitoring data and the corresponding historical pseudo-abnormal feature monitoring data at the corresponding data acquisition node according to the change amount value difference and change rate difference of each data acquisition node;

[0137] Generate the second similarity according to multiple second change amount value sub-similarities and second change rate sub-similarities of the marked abnormal feature monitoring data and the corresponding historical pseudo-abnormal feature monitoring data in the corresponding similar time period, and combine the similarity of the second change trend;

[0138] The calculation formula for the second similarity is as follows:

[0139]

[0140] Among them, X2 is the second similarity, u2 1 is the total number of data acquisition nodes of the abnormal feature monitoring data of the first mark and the corresponding historical pseudo-abnormal feature detection data in the similar time period, L2 z,1 is the second change amount value sub-similarity of the abnormal feature monitoring data of the first mark and the corresponding historical pseudo-abnormal feature detection data at the z-th data acquisition node in the similar time period, K2 z,1 is the second change rate sub-similarity of the abnormal feature monitoring data of the first mark and the corresponding historical pseudo-abnormal feature detection data at the z-th data acquisition node in the similar time period, g2 n2The total number of data acquisition nodes of the abnormal feature monitoring data of the n2th mark and the corresponding historical pseudo-abnormal feature detection data in a similar period is L2 r,n2 The second change amount sub-similarity of the abnormal feature monitoring data of the n2th mark and the corresponding historical pseudo-abnormal feature detection data at the rth data acquisition node in a similar period is K2 r,n2 The second change rate sub-similarity of the abnormal feature monitoring data of the n2th mark and the corresponding historical pseudo-abnormal feature detection data at the rth data acquisition node in a similar period

[0141] In this embodiment, z = 1, 2,..., u2, r = 1, 2,..., g2

[0142] In this embodiment, L2 z,1 = △l2 z,1 × c1, △l2 z,1 The change amount difference of the abnormal feature monitoring data of the 1st mark and the corresponding historical pseudo-abnormal feature detection data at the zth data acquisition node in a similar period is K2 z,1 = △k2 z,1 × c2, △k2 z,1 The change rate difference of the abnormal feature monitoring data of the 1st mark and the corresponding historical abnormal feature detection data at the zth data acquisition node in a similar period

[0143] In this embodiment, L2 r,n2 = △l2 r,n2 × c1, △l2 r,n2 The change amount difference of the abnormal feature monitoring data of the n2th mark and the corresponding historical pseudo-abnormal feature detection data at the rth data acquisition node in a similar period is K2 r,n2 = △k2 r,n2 × c2, △k2 r,n2 The change rate difference of the abnormal feature monitoring data of the n2th mark and the corresponding historical abnormal feature detection data at the rth data acquisition node in a similar period

[0144] In this embodiment, the second similarity refers to the similarity degree between the abnormal change characteristics of all abnormal feature monitoring data of each abnormal device in the early warning work order and the hidden danger change characteristics of the corresponding historical pseudo-abnormal feature monitoring data in each historical hidden danger log of the corresponding abnormal device. Whether it is a false alarm situation is judged according to the second similarity, so as to improve the accuracy and efficiency of fire warning

[0145] In some embodiments of the present application, the credibility of the early warning work order is set according to the first similarity and the second similarity, and the corresponding notification instruction is generated, including:

[0146] Pre-set a similarity threshold and a credibility threshold

[0147] When the first similarity is greater than the similarity threshold and the second similarity is less than the similarity threshold, set the credibility of the warning work order to be greater than the credibility threshold, and generate a fire notification instruction according to the warning work order. The fire notification instruction includes triggering fire protection linkage and emergency plans, and sending fire information to the fire department;

[0148] When the second similarity is greater than the similarity threshold and the first similarity is less than the similarity threshold, set the credibility of the warning work order to be less than the credibility threshold, and generate a hidden danger notification instruction according to the warning work order. The hidden danger notification instruction includes reminding the inspection personnel to perform operation and maintenance;

[0149] In this embodiment, the credibility refers to the accuracy of the fire in the warning work order, and the similarity threshold is the maximum value of the preset similarity degree. The notification instruction of the warning work order is determined by the similarity threshold and the credibility threshold, which improves the judgment accuracy of the warning work order, and improves the fire warning accuracy and warning efficiency.

[0150] In this embodiment, when both the first similarity and the second similarity are less than the similarity threshold, obtain the abnormal feature monitoring data and abnormal change features of the abnormal equipment in the next preset time period and re-analyze the similarity, which improves the judgment result of the credibility of the warning work order and ensures the fire warning accuracy.

[0151] In some embodiments of the present application, as Figure 2 shown, there is also a fire warning management method:

[0152] Step S201: Obtain the real-time feature monitoring data of the preset monitoring equipment and analyze it to determine the abnormal equipment, abnormal feature monitoring data and corresponding abnormal change features, and generate a warning work order;

[0153] Step S202: Obtain the historical fire log and historical hidden danger log of the abnormal equipment, determine the historical abnormal feature monitoring data and corresponding fire change features of the corresponding abnormal equipment based on the historical fire log, and determine the historical pseudo-abnormal feature monitoring data and corresponding hidden danger change features of the corresponding abnormal equipment based on the historical hidden danger log;

[0154] Step S203: Perform similarity analysis on the abnormal feature monitoring data and corresponding abnormal change features of the abnormal equipment in the warning work order respectively with the historical abnormal feature monitoring data and corresponding fire change features, historical pseudo-abnormal feature monitoring data and corresponding hidden danger change features of the corresponding abnormal equipment to obtain the first similarity and the second similarity;

[0155] Step S204: Set the credibility of the warning work order according to the first similarity and the second similarity, and generate the corresponding notification instruction.

[0156] The above are only the preferred embodiments of the present application. It should be noted that for those of ordinary skill in the art, without departing from the technical principle of the present application, several improvements and substitutions can be made, and these improvements and substitutions should also be regarded as the protection scope of the present application.

Claims

1. A fire warning management system, characterized in that: include: The monitoring module is used to obtain and analyze the real-time characteristic monitoring data of the preset monitoring equipment, determine the abnormal equipment, abnormal characteristic monitoring data and the corresponding abnormal change characteristics, and generate an early warning work order; A determination module is used to obtain the historical fire log and the historical hidden danger log of the abnormal equipment, determine the historical abnormal characteristic monitoring data of the corresponding abnormal equipment and the corresponding fire condition change characteristics based on the historical fire log, and determine the historical pseudo-abnormal characteristic monitoring data of the corresponding abnormal equipment and the corresponding hidden danger change characteristics based on the historical hidden danger log; An analysis module is used to perform similarity analysis on the abnormal feature monitoring data of the abnormal equipment in the early warning work order and the corresponding abnormal change features with the historical abnormal feature monitoring data of the corresponding abnormal equipment and the corresponding fire change features, the historical pseudo-abnormal feature monitoring data and the corresponding hidden danger change features, to obtain a first similarity and a second similarity; The notification module is used to set the credibility of the early warning work order according to the first similarity and the second similarity, and generate a corresponding notification instruction.

2. The fire early warning management system according to claim 1, characterized in that: Identify abnormal equipment, abnormal feature monitoring data, and corresponding abnormal change features, including: Pre-set multiple fire assessment indicators; Acquire the real-time monitoring data of the preset monitoring equipment within the preset time period, and obtain the historical correlation degree between each real-time monitoring data and each fire evaluation index, and set the real-time monitoring data with a historical correlation degree greater than a preset correlation degree threshold as the real-time characteristic monitoring data; Subtract the real-time characteristic monitoring data within a preset period from the standard monitoring data of the corresponding fire evaluation index to obtain the data difference between the real-time characteristic monitoring data and the standard monitoring data; The real-time characteristic monitoring data with a data difference greater than a preset difference threshold is set as abnormal characteristic monitoring data, and the preset monitoring device with abnormal characteristic monitoring data is set as an abnormal device; Determine the initial time node at which the data difference of each abnormal feature monitoring data is greater than a preset difference threshold, take the initial time node as the starting point, set the data collection node at a preset time interval, obtain the corresponding abnormal feature monitoring data according to the data collection node, and construct a data change curve for each abnormal feature monitoring data; The change trend, change rate and change value of the abnormal feature monitoring data in each data change curve are set as the abnormal change characteristics of the corresponding abnormal feature monitoring data.

3. The fire early warning management system according to claim 2, characterized in that: Based on the historical fire log, the historical abnormal characteristic monitoring data of the corresponding abnormal equipment and the corresponding fire change characteristics are determined; based on the historical hidden danger log, the historical pseudo-abnormal characteristic monitoring data and the corresponding hidden danger change characteristics are determined, including: Obtain historical fire logs of abnormal equipment, filter out historical abnormal feature monitoring data in each historical fire log, and determine the first abnormal period of the historical abnormal feature monitoring data; Taking the first abnormal period of each historical abnormal feature monitoring data as the time reference line, setting the data collection node at the preset time interval, obtaining the corresponding historical abnormal feature monitoring data according to each data collection node, and mapping it to the corresponding time reference line, to obtain the first historical data change curve of each historical abnormal feature detection data; The change trend, change rate and change value of the historical abnormal characteristic monitoring data in each first historical data change curve are set as the fire condition change characteristics of the corresponding historical abnormal characteristic monitoring data; Analyze all historical fire logs of abnormal equipment, and obtain the fire change characteristics of historical abnormal characteristic monitoring data in each historical fire log corresponding to the abnormal equipment in turn; Obtain historical hidden danger logs of abnormal equipment, filter out historical pseudo-abnormal feature monitoring data in each historical hidden danger log, and determine the second abnormal period of the historical pseudo-abnormal feature monitoring data; Taking the second abnormal period of each historical pseudo-abnormal feature monitoring data as the time reference line, setting the data collection node at a preset time interval, obtaining the corresponding historical pseudo-abnormal feature monitoring data according to each data collection node, and mapping it to the corresponding time reference line, to obtain the second historical data change curve of each historical pseudo-abnormal feature detection data; The change trend, change rate and change value of the historical pseudo-abnormal feature monitoring data in each second historical data change curve are set as the hidden danger change characteristics of the corresponding historical pseudo-abnormal feature monitoring data; All historical hidden danger logs of abnormal equipment are analyzed, and hidden danger change characteristics of historical pseudo-abnormal feature monitoring data in each historical hidden danger log of the current abnormal equipment are obtained in turn.

4. The fire early warning management system according to claim 3, characterized in that: Before similarity analysis, it also includes: Compare all abnormal feature monitoring data of each abnormal device in the early warning work order with the historical abnormal feature monitoring data of each historical fire log of the corresponding abnormal device to obtain the appearance ratio of all abnormal feature monitoring data in the historical abnormal feature monitoring data of each historical fire log; If the occurrence ratio is less than the preset occurrence ratio threshold, the corresponding historical fire log and the corresponding historical abnormal feature monitoring data are eliminated; Compare all abnormal feature monitoring data of each abnormal device in the early warning work order with the historical pseudo abnormal feature monitoring data of each historical hidden danger log of the corresponding abnormal device, and obtain the appearance ratio of all abnormal feature monitoring data in the historical pseudo abnormal feature monitoring data of each historical hidden danger log; If the occurrence ratio is less than the preset occurrence ratio threshold, the corresponding historical hidden danger log and the corresponding historical pseudo-abnormal feature monitoring data will be eliminated.

5. The fire early warning management system according to claim 4, characterized in that: Perform similarity analysis, including: Compare the change trends of the historical abnormal feature monitoring data of the remaining historical fire logs of the same abnormal device with the corresponding abnormal feature monitoring data of the corresponding abnormal device in the early warning work order to obtain a first change trend difference; Marking the abnormal characteristic monitoring data whose first change trend difference is less than the preset difference, and calculating the difference value of the difference; Generate a first change trend similarity between the change trend of the abnormal feature monitoring data marked by the same abnormal device and the historical abnormal feature monitoring data of the same corresponding historical fire log according to the number of marked abnormal feature monitoring data and the difference in difference; The calculation formula of the first change trend similarity is: Wherein, B1 is the first change trend similarity, n1 is the number of marks of abnormal feature monitoring data whose difference of the first change trend is less than the preset difference, m is the total number of abnormal feature monitoring data of abnormal equipment, △Ci is the difference between the abnormal feature monitoring data of the ith mark and the corresponding historical abnormal feature monitoring data, di is the weight coefficient of the abnormal feature monitoring data of the ith mark, and c0 is the first similarity conversion coefficient; The historical fire logs whose first change trend similarity is less than a preset change trend similarity threshold are eliminated again, and the similar period of the change trend of the historical abnormal feature monitoring data of each remaining historical fire log and the abnormal feature monitoring data marked in the early warning work order is obtained, wherein the similar period includes multiple data collection nodes; Obtain the change value and change rate of the marked abnormal feature monitoring data and the corresponding historical abnormal feature monitoring data at each data collection node in the corresponding similar period, and make a difference to obtain the change value difference and change rate difference at each data collection node; Generate a first change value sub-similarity and a first change rate sub-similarity between the marked abnormal feature monitoring data and the corresponding historical abnormal feature monitoring data at the corresponding data collection node according to the change value difference and the change rate difference of each data collection node; Generate a first similarity based on a plurality of first change value sub-similarity and a first change rate sub-similarity between the marked abnormal feature monitoring data and the corresponding historical abnormal feature monitoring data in a corresponding similar time period, and in combination with the first change trend similarity; The calculation formula of the first similarity is: Among them, X1 is the first similarity, h3 is the weight coefficient of the change trend similarity, d1 is the weight coefficient of the abnormal feature monitoring data of the first mark, and d n1 is the weight coefficient of the abnormal feature monitoring data of the n1th mark, u11 is the total number of data collection nodes of the abnormal feature monitoring data of the first mark and the corresponding historical abnormal feature detection data in the similar period, L1 s,1 is the first change value sub-similarity between the first marked abnormal feature monitoring data and the corresponding historical abnormal feature detection data at the sth data collection node in a similar period, h1 is the weight coefficient of the change value sub-similarity, h2 is the weight coefficient of the change rate sub-similarity, K1 s,1 is the first change rate sub-similarity between the first marked abnormal feature monitoring data and the corresponding historical abnormal feature detection data at the sth data collection node in a similar period, g1 n1 L1 is the total number of data collection nodes in the similar period between the abnormal feature monitoring data of the n1th mark and the corresponding historical abnormal feature detection data, v,n1 K1 is the first change value sub-similarity between the abnormal feature monitoring data of the n1th mark and the corresponding historical abnormal feature detection data at the vth data collection node in a similar period, v,n1 It is the first change rate sub-similarity between the abnormal feature monitoring data of the n1th mark and the corresponding historical abnormal feature detection data at the vth data collection node in a similar period.

6. The fire early warning management system according to claim 5, characterized in that: Similarity analysis also includes: Compare the change trends of the historical pseudo-abnormal feature monitoring data in the remaining historical hidden danger logs of the same abnormal device with the corresponding abnormal feature monitoring data of the corresponding abnormal device in the early warning work order to obtain a second change trend difference; Mark the abnormal characteristic monitoring data whose second change trend difference is less than the preset difference, and calculate the difference value; Generate a second change trend similarity between the change trend of the abnormal feature monitoring data marked by the same abnormal device and the historical pseudo-abnormal feature monitoring data of the same historical hidden danger log according to the number of marked abnormal feature monitoring data and the difference in difference; The calculation formula of the second change trend similarity is: Wherein, B2 is the similarity of the second change trend, n2 is the number of marks of abnormal feature monitoring data whose difference of the second change trend is less than the preset difference, and △Wi is the difference between the abnormal feature monitoring data of the ith mark and the corresponding historical pseudo-abnormal feature monitoring data; Eliminate the historical hidden danger logs whose second change trend similarity is less than the preset change trend similarity threshold, and obtain the similar period of the change trend of the historical pseudo-abnormal feature monitoring data of each remaining historical hidden danger log and the abnormal feature monitoring data marked in the early warning work order, wherein the similar period includes multiple data collection nodes; Obtain the change value and change rate of each data collection node in the corresponding similar time period of the marked abnormal feature monitoring data and the corresponding historical pseudo-abnormal feature monitoring data, and make a difference to obtain the change value difference and change rate difference at each data collection node; Generate a second change value sub-similarity and a second change rate sub-similarity between the marked abnormal feature monitoring data and the corresponding historical pseudo abnormal feature monitoring data at the corresponding data collection node according to the change value difference and the change rate difference of each data collection node; Generate a second similarity based on a plurality of second change value sub-similarity and a second change rate sub-similarity between the marked abnormal feature monitoring data and the corresponding historical pseudo abnormal feature monitoring data in a corresponding similar time period, and in combination with a second change trend similarity; The calculation formula of the second similarity is: Among them, X2 is the second similarity, u21 is the total number of data collection nodes in the similar period between the first marked abnormal feature monitoring data and the corresponding historical pseudo-abnormal feature detection data, and L2 z,1 K2 is the second change value sub-similarity between the first marked abnormal feature monitoring data and the corresponding historical pseudo-abnormal feature detection data at the zth data collection node in a similar period, z,1 is the second change rate sub-similarity between the first marked abnormal feature monitoring data and the corresponding historical pseudo-abnormal feature detection data at the zth data collection node in a similar period, g2 n2 is the total number of data collection nodes in the similar period between the abnormal feature monitoring data of the n2th mark and the corresponding historical pseudo-abnormal feature detection data, L2 r,n2 K2 is the second change value sub-similarity between the abnormal feature monitoring data of the n2th mark and the corresponding historical pseudo-abnormal feature detection data at the rth data collection node in a similar period, r,n2 It is the second change rate sub-similarity between the abnormal feature monitoring data of the n2th mark and the corresponding historical pseudo-abnormal feature detection data at the rth data collection node in a similar period.

7. The fire early warning management system according to claim 6, characterized in that: The credibility of the warning work order is set according to the first similarity and the second similarity, and a corresponding notification instruction is generated, including: Pre-set similarity threshold and credibility threshold; If the first similarity is greater than the similarity threshold and the second similarity is less than the similarity threshold, the credibility of the early warning work order is set to be greater than the credibility threshold, and a fire notification instruction is generated according to the early warning work order, wherein the fire notification instruction includes triggering fire linkage and emergency plan, and sending fire information to the fire department; If the second similarity is greater than the similarity threshold and the first similarity is less than the similarity threshold, the credibility of the warning work order is set to be less than the credibility threshold, and a hidden danger notification instruction is generated according to the warning work order, wherein the hidden danger notification instruction includes reminding the inspection personnel to perform operation and maintenance.

8. A fire warning management method, characterized in that: include: Obtain and analyze the real-time feature monitoring data of the preset monitoring equipment, determine the abnormal equipment, abnormal feature monitoring data and corresponding abnormal change characteristics, and generate an early warning work order; Obtain the historical fire log and historical hidden danger log of the abnormal equipment, determine the historical abnormal characteristic monitoring data of the corresponding abnormal equipment and the corresponding fire condition change characteristics based on the historical fire log, and determine the historical pseudo-abnormal characteristic monitoring data of the corresponding abnormal equipment and the corresponding hidden danger change characteristics based on the historical hidden danger log; Perform similarity analysis on the abnormal feature monitoring data of the abnormal equipment in the early warning work order and the corresponding abnormal change features with the historical abnormal feature monitoring data of the corresponding abnormal equipment and the corresponding fire change features, the historical pseudo-abnormal feature monitoring data and the corresponding hidden danger change features, to obtain a first similarity and a second similarity; The credibility of the early warning work order is set according to the first similarity and the second similarity, and a corresponding notification instruction is generated.

Images