CN-121982859-A - Personnel perception early warning system based on data fusion analysis

CN121982859ACN 121982859 ACN121982859 ACN 121982859ACN-121982859-A

Abstract

The invention relates to the technical field of perception early warning and discloses a personnel perception early warning system based on data fusion analysis, which comprises a data acquisition and preprocessing module, a data analysis module and a correlation early warning module, wherein the data acquisition and preprocessing module is used for acquiring actual residence data and district flow data of key personnel, preprocessing the acquired actual residence data and district flow data, removing invalid and abnormal data, adding a time stamp and geographic coordinates, and the model construction module is used for constructing a man-machine ground data model and a man-machine vehicle data model according to the preprocessed actual residence data and district flow data, and the data analysis module is used for respectively acquiring abnormal risk values and correlation early warning values of the key personnel according to the real residence data and district flow data updated in real time. And carrying out normalization processing and weighted fusion calculation on the acquired various data by constructing a man-machine ground and man-machine model, acquiring an abnormal risk value and an associated early warning value of key personnel, and carrying out hierarchical early warning notification according to the acquired abnormal risk value and the associated early warning value.

Inventors

WANG TIANSHENG
ZHANG SHUO
WANG WEI
CHEN YI
CHEN GUANGZHU

Assignees

池州市公安局

Dates

Publication Date: 20260505
Application Date: 20260122

Claims (8)

1. A personnel perception early warning system based on data fusion analysis is characterized by comprising the following modules: The data acquisition and preprocessing module is used for acquiring actual residence data and cell flow data of key personnel, preprocessing the acquired actual residence data and cell flow data, removing invalid and abnormal data, and adding a time stamp and geographic coordinates; the model construction module is used for constructing a man-machine ground data model and a man-machine vehicle data model according to the preprocessed actual living data and the cell flow data; the data analysis module is used for respectively acquiring the abnormal risk value and the associated early warning value of the key personnel according to the real living data and the district flow data updated in real time; And the early warning notification module is used for comparing the acquired abnormal risk value with the associated early warning value and the set safety threshold value and carrying out early warning notification according to the comparison result.
2. The personnel perception early warning system based on data fusion analysis of claim 1, wherein the method for working by the data analysis module is as follows: Extracting key data in the man-machine ground data model and the man-machine vehicle data model; Carrying out normalization processing on key data in the man-machine ground data model and the man-machine vehicle data model; And respectively calculating and acquiring an abnormal risk value and an associated early warning value according to the normalized man-machine ground data model and key data in the man-machine vehicle data model.
3. The personnel perception early warning system based on data fusion analysis according to claim 2, wherein the method for acquiring the abnormal risk value is as follows: acquiring the water and gas use matching degree, the frequency deviation degree of entering and exiting and the technology detection data association degree according to the normalized key data; according to the acquired water and gas use matching degree, calculating an acquired water and gas abnormal amplification value, according to the frequency deviation degree, calculating an acquired frequency attenuation coefficient, and according to the acquired technical detection data association degree, calculating an acquired technical detection association coefficient; And calculating and obtaining the living abnormal risk value of the key personnel according to the preset distribution weight coefficient by the obtained water-air abnormal amplification value, the frequency attenuation coefficient and the technical detection correlation coefficient.
4. The personnel perception early warning system based on data fusion analysis according to claim 2, wherein the method for acquiring the association early warning value is as follows: Acquiring the coincidence degree of personnel, mobile phones and vehicles in the same time and space according to the key data in the normalized man-machine vehicle data model and the key personnel behavior track data, and acquiring the time and space synchronization degree; Counting the recording times of people, mobile phones and vehicles meeting effective synchronous conditions in a preset adjacent time window, and obtaining associated frequencies; Judging the deviation level of the current behavior of the key personnel according to the historical track base line of the key personnel, and obtaining the behavior anomaly degree; Acquiring a space-time synchronization value according to the space-time synchronization degree calculation, acquiring an associated frequency attenuation coefficient according to the associated frequency, and acquiring a behavior anomaly amplification coefficient according to the behavior anomaly degree; And calculating and obtaining a human-computer vehicle association reliability factor through the space-time synchronization value and the association frequency attenuation coefficient, carrying out weighted summation according to the human-computer vehicle association reliability factor and the abnormal behavior amplification coefficient and a preset distribution weight coefficient, and calculating and obtaining the association early warning value of the key personnel.
5. The personnel perception early warning system based on data fusion analysis according to claim 1, wherein the early warning notification module works by the following method: The acquired abnormal risk value and the associated early warning value are respectively compared with an abnormal risk threshold value and an associated early warning threshold value: when the abnormal risk value is greater than or equal to the abnormal risk threshold value, triggering abnormal living and floating population risk early warning, and pushing actual living data and community floating data to corresponding personnel; Triggering a first-level early warning when the associated early warning value is greater than or equal to a first associated early warning threshold value and the abnormal risk value is greater than or equal to an abnormal risk threshold value; And triggering the second-level early warning when the associated early warning value is smaller than the first associated early warning threshold and larger than the second associated early warning threshold and the abnormal risk value is larger than or equal to the abnormal risk threshold.
6. The personnel perception early warning system based on data fusion analysis according to claim 1, wherein the method for working by the data acquisition and preprocessing module is as follows: Acquiring actual living data and district flow data, wherein the actual living data comprises newly-added data such as face snapshot files, vehicle snapshot files, shared bicycle and the like, marital registration, real estate registration, resident water, resident gas, industrial and commercial registration, a dispatching system and grid member data; the cell flow data counts the frequent entering and exiting of the cell personnel according to the frequency of the entering and exiting of the cell personnel and the frequency of the vehicle driver to form a cell flow number; unifying field formats and coding rules of the data, and adding a collected time stamp and a geographic coordinate label of the position of the collecting equipment for each data; And identifying and eliminating invalid data and repeated data in the actual residence data and the cell flow data.
7. The personnel perception early warning system based on data fusion analysis of claim 6, wherein the process of rejecting invalid data is as follows: Marking the obtained fuzzy face, incomplete license plate and invalid mobile phone number as candidate invalid data, obtaining associated data of the same acquisition equipment in the same time window, judging as absolute invalid data and eliminating if the candidate invalid data has no corresponding vehicle snapshot record and mobile phone probe data and no clear data obtained by the same equipment, and reserving data and marking as to-be-manually rechecked if the invalid data has associated data; and (3) checking the integrity of key fields of the property, the water vapor and the registration information, marking the key fields as candidate invalid data, reserving the candidate invalid data after the candidate invalid data is complemented by other data association, and rejecting the candidate invalid data if the candidate invalid data cannot be complemented in association and no history valid record exists.
8. The personnel perception early warning system based on data fusion analysis of claim 6, wherein the process of eliminating repeated data is as follows: Generating a space-time reference fingerprint, a core attribute fingerprint and a device association fingerprint for each piece of data, if the space-time reference fingerprints of the two pieces of data are completely matched with the core attribute fingerprint, judging the two pieces of data as repeated data, and reserving one piece of data with earlier time stamps and higher data quality; if the device association fingerprint of the two data is completely matched with the core attribute fingerprint, but the time stamp is deviated from the geographic coordinates, the device association fingerprint of the two data is judged to be continuous collected data of the same target, and the first data is reserved and the subsequent data are combined.

Description

Personnel perception early warning system based on data fusion analysis Technical Field The invention relates to the technical field of perception early warning, in particular to a personnel perception early warning system based on data fusion analysis. Background In public security basic level work, basic data acquisition and key personnel management and control are core links for maintaining social security and striking illegal crimes, the prior art relies on actual living data such as face snapshot files, vehicle snapshot files, house property registration, water consumption of residents and the like, and cell flow data formed by frequent statistics of cell access personnel and vehicle drivers, and personnel management and risk early warning work is carried out by constructing a related data model and combining manual checking and other modes, so that actual combat demands such as security control and case investigation are supported through data integration; On the one hand, the data source dispersion causes incomplete information of personnel involved, the basic data of the community is not new, it is difficult to form a complete and accurate personnel related data system, on the other hand, the data processing is dependent on static record and manual operation, there are limitations of passive registration, static data dependence and low efficiency of manual check, real-time perception and accurate quantification of important personnel risks cannot be realized, and further the acquisition quality of basic data of public security, the management and control effect of important personnel and the detection efficiency of cases are affected, and the actual requirements of public security work on technological energization and efficient prevention and control are difficult to meet; therefore, the invention provides a personnel perception early warning system based on data fusion analysis, which solves the defects in the prior art. Disclosure of Invention The invention aims to provide a personnel perception early warning system based on data fusion analysis, which is used for solving the defects in the background technology. The aim of the invention can be achieved by the following technical scheme: a personnel perception early warning system based on data fusion analysis is characterized by comprising the following modules: The data acquisition and preprocessing module is used for acquiring actual residence data and cell flow data of key personnel, preprocessing the acquired actual residence data and cell flow data, removing invalid and abnormal data, and adding a time stamp and geographic coordinates; the model construction module is used for constructing a man-machine ground data model and a man-machine vehicle data model according to the preprocessed actual living data and the cell flow data; the data analysis module is used for respectively acquiring the abnormal risk value and the associated early warning value of the key personnel according to the real living data and the district flow data updated in real time; And the early warning notification module is used for comparing the acquired abnormal risk value with the associated early warning value and the set safety threshold value and carrying out early warning notification according to the comparison result. Preferably, the method for working by the data analysis module comprises the following steps: Extracting key data in the man-machine ground data model and the man-machine vehicle data model; Carrying out normalization processing on key data in the man-machine ground data model and the man-machine vehicle data model; And respectively calculating and acquiring an abnormal risk value and an associated early warning value according to the normalized man-machine ground data model and key data in the man-machine vehicle data model. Preferably, the method for obtaining the abnormal risk value comprises the following steps: acquiring the water and gas use matching degree, the frequency deviation degree of entering and exiting and the technology detection data association degree according to the normalized key data; according to the acquired water and gas use matching degree, calculating an acquired water and gas abnormal amplification value, according to the frequency deviation degree, calculating an acquired frequency attenuation coefficient, and according to the acquired technical detection data association degree, calculating an acquired technical detection association coefficient; And calculating and obtaining the living abnormal risk value of the key personnel according to the preset distribution weight coefficient by the obtained water-air abnormal amplification value, the frequency attenuation coefficient and the technical detection correlation coefficient. Preferably, the method for acquiring the association early warning value comprises the following steps: Acquiring the coincidence degree of personnel, mobile phones and vehicles in the same tim