A smart campus management system based on student safety

Abstract

The present application discloses a smart campus management system based on student safety, which relates to the technical field of campus safety management, and solves the technical problem that the existing campus safety management system only considers the abnormal analysis of students in a single way, resulting in low accuracy of analysis results. The application comprises: a data acquisition module: obtaining the positioning information and real-time monitoring data of students; a regional assessment module: obtaining the basic information of the campus and some historical positioning information, and constructing a corresponding three-dimensional campus model according to the construction information in the basic information; dividing the three-dimensional campus model into regions according to the historical positioning information to obtain several management areas and supervision coefficients; a data analysis module: generating the attention degree of the corresponding students according to the historical positioning information and the historical performance data; generating a warning signal according to the positioning information and the corresponding attention degree of each student; a danger warning module: obtaining the warning signal, and sending the warning signal to the supervision center for warning; improving the accuracy of the student safety analysis.

Description

Technical Field

[0001] The present application belongs to the field of campus safety management technology, and relates to campus safety management technology, specifically a smart campus management system based on student safety. Background Art

[0002] Campus safety issues are receiving increasing attention. Campus safety issues are a key issue of social concern. They involve the life safety and physical and mental health of students and need to be taken seriously.

[0003] The prior art (an invention patent with announcement number CN115099378B) discloses an intelligent early warning system and method for electronic student ID cards based on a safe campus, including: an information collection module, a data management center, a positioning information analysis module, a student ID card function management module and a security management module. The information collection module collects students' previous activity information and default function information of electronic ID students, and the data management center stores and manages all the collected information. The positioning information analysis module predicts the familiarity between students and analyzes the safety range of students participating in fixed collective activities. The student ID card function management module avoids inappropriate electronic student ID card update time and sets a fixed do not disturb time period. The security management module monitors students' activity trajectories and issues security warnings when abnormalities are detected, effectively avoiding inappropriate electronic student ID card update time and reducing the probability of security risks. Timely security warnings can prevent security accidents and protect students' safety.

[0004] The above-mentioned early warning management system uses the positioning function of the electronic student ID card to analyze the familiarity between classmates through the similarity of trajectories, and then determines whether there is any abnormality among students; however, in campus life, due to small conflicts between classmates or mental health problems of each student, some students may be tracked; simply analyzing the similarity of trajectories between classmates, analyzing the familiarity between classmates, and then analyzing the safety of students will obviously lead to a low accuracy of the final safety situation analysis results; therefore, a smart campus management system based on student safety is needed. Summary of the invention

[0005] The present application aims to solve at least one of the technical problems existing in the prior art; to this end, the present application proposes a smart campus management system based on student safety, which is used to solve the technical problem that the existing campus safety management system only considers a single situation in the abnormal analysis of students, resulting in low accuracy of the analysis results.

[0006] To achieve the above-mentioned purpose, the first aspect of the present application provides a smart campus management system based on student safety, including: a data collection module, a regional assessment module, a data analysis module, a danger warning module, a supervision center and a database;

[0007] The data acquisition module: obtains the student's location information through the electronic student certificate connected to it; and obtains real-time monitoring data through the monitoring acquisition device connected to it;

[0008] The regional assessment module: obtains basic information of the campus and several historical positioning information of each student through the database, and constructs a corresponding three-dimensional campus model according to the construction information in the basic information; divides the three-dimensional campus model into several management areas according to the historical positioning information, and generates a supervision coefficient for each management area according to the historical positioning information in the management area;

[0009] The data analysis module: obtains the recent historical location information and historical performance data of each student through the database; generates the attention level of the corresponding student according to the historical location information and historical performance data; and

[0010] Obtaining the real-time location information of each student, generating a preliminary warning signal based on the real-time location information and the corresponding attention level of each student; generating a warning signal based on the preliminary warning signal;

[0011] The danger warning module obtains warning signals and sends the warning signals to the supervision center for warning.

[0012] This application obtains basic information about the campus, builds a three-dimensional campus model based on the campus layout data and building data in the basic information; obtains the recent historical positioning information of each student, divides the three-dimensional campus model into several management areas based on the historical positioning information, and analyzes the management areas to obtain the supervision coefficient representing the characteristics of each management area; and generates the corresponding attention level of each student based on the recent performance and status of each student; obtains the real-time positioning information of each classmate, and generates early warning signals based on the real-time positioning information, attention level and supervision coefficient of each management area, as well as the monitoring video of each monitoring device; so that the relevant data on students' safety at school can be analyzed from multiple perspectives to improve the accuracy of student safety analysis, detect students' abnormal situations in a timely manner, and ensure students' safety at school.

[0013] Preferably, the step of constructing a corresponding three-dimensional campus model according to the construction information in the basic information includes:

[0014] The campus layout data and the architectural data of each building in the construction information are extracted, and a corresponding campus three-dimensional model is constructed based on the campus layout data and the architectural data of each building.

[0015] Preferably, the campus three-dimensional model is divided into several management areas according to the historical positioning information, including:

[0016] Acquire positioning coordinates from a number of historical positioning information; integrate the positioning coordinates into a number of feature vectors; perform cluster analysis on a number of feature vectors to obtain a number of clusters; acquire the feature vector corresponding to the boundary of each cluster, extract the positioning coordinates in each feature vector as the boundary positioning coordinates, mark each positioning coordinate of the number of clusters into the three-dimensional model of the campus, and obtain a number of management areas.

[0017] Preferably, generating the supervision coefficient of each management area according to the historical positioning information in the management area includes:

[0018] Obtain the positioning ID corresponding to the positioning coordinates in several historical positioning information in each management area; obtain the number of positioning coordinates RSit, the number of positioning IDs ISit and the positioning duration SCit of each positioning ID in each set time period in the management area, i is the number of the management area, i=1, 2, ..., I; I is the total number of management areas; t is the number of the set time period, t=1, 2, ..., T; T is the total number of the set time periods;

[0019] By formula The supervision coefficient JXit of the management area numbered i in the set time period numbered t is calculated; among them, β1, β2 and β3 are adjustment coefficients, and the specific values ​​are set according to experience; α1, α2 and α3 are weight coefficients, and the specific values ​​are set according to experience; ZIS is the total number of location IDs of the corresponding schools; SDt is the time period length of the set time period numbered t.

[0020] This application generates a supervision coefficient for each management area in each time period according to the characteristics of each management area in each time period, thereby facilitating targeted analysis of each management area.

[0021] Preferably, generating the corresponding student's attention level according to the historical positioning information and the historical performance data includes:

[0022] Obtaining recent historical performance data of each student, and inputting the historical performance data into a supervision scoring model to obtain a supervision score of the corresponding student; the supervision scoring model is obtained by training an artificial intelligence model;

[0023] Obtain the student's recent historical location information, extract the location coordinates and corresponding location time in the historical location information; map each location coordinate to the campus 3D model, and combine the location time of each location coordinate to obtain the number of students appearing in each set time period in each management area CCit;

[0024] By formula The corresponding student's attention GZ is calculated; among them, γ1 and γ2 are proportional coefficients, and the specific values ​​are set according to experience; β4 is the adjustment coefficient, and the specific value is set according to experience; JP is the supervision score of the corresponding student.

[0025] Preferably, the supervision degree scoring model is obtained by training an artificial intelligence model, including:

[0026] Obtaining a number of historical performance data and corresponding supervision scores from a database; the supervision score is a score given by experts or relevant teachers on the degree to which a student needs supervision based on the performance data of the corresponding student. For example, if the corresponding student usually exhibits mania, depression, etc., the corresponding supervision score is high, and if the corresponding student usually exhibits normal performance, the corresponding supervision score is low; integrating the historical performance data and the corresponding supervision score into a number of sets of training data and test data;

[0027] The artificial intelligence model is trained using training data, and the trained artificial intelligence model is tested using test data. Specifically, after the artificial intelligence model is trained using a number of training data, the historical performance data in the test data is input into the trained artificial intelligence model to output a corresponding supervision score; when the difference between the output supervision score and the supervision score recorded in the corresponding test data is within an acceptable range, the next set of test data is tested; otherwise, the relevant parameters of the artificial intelligence model are adjusted, and the test is performed again using the current set of test data; until a set proportion of the test data passes the test; finally, a feature scoring model is obtained with the historical performance data as input and the supervision score as output; the artificial intelligence model includes a BP neural network model and an RBF neural network model.

[0028] Preferably, the generating of a preliminary warning signal according to the positioning information and the corresponding attention level of each student includes:

[0029] When there is a student's location information in a management area whose supervision coefficient is greater than a set supervision coefficient threshold; obtaining the student's attention degree, and determining whether the attention degree is less than the set attention degree threshold;

[0030] If yes, then the historical location information of the student within the set time is obtained and recorded as the target location information, and other historical location information within the set time is obtained and recorded as the reference location information; a low-risk warning signal is generated according to the attention of the target location information and the reference location information;

[0031] If not, a high-risk warning signal is generated; the preliminary warning signal includes a high-risk warning signal and a low-risk warning signal.

[0032] Preferably, generating a low-risk warning signal according to the target positioning information and the reference positioning information attention level includes:

[0033] Extract the positioning coordinates and positioning time from the target positioning information and each reference positioning information, and fit the positioning coordinates into the target time trajectory curve MG (xmj, ymj, zmj, tj) and the reference time trajectory curve CG (xcj, ycj, zcj, tj) in the order of positioning time;

[0034] By formula The correlation degree GL between the corresponding reference time trajectory curve and the target time trajectory curve is calculated; wherein, j is the number of the set time node, j=1, 2, ..., J; J is the total number of the set time nodes; tj is the set time node numbered j; xmj, ymj and zmj are respectively the x coordinate, y coordinate and z coordinate in the positioning coordinate corresponding to the set time node numbered j in the target time trajectory area; xcj, ycj and zcj are respectively the x coordinate, y coordinate and z coordinate in the positioning coordinate corresponding to the set time node numbered j in the reference time trajectory area; δj is the weight coefficient of the set time node numbered j, and the specific value is set according to experience; β5 is the adjustment coefficient, and the specific value is set according to experience;

[0035] Determine whether the correlation between the target time trajectory curve and each reference time trajectory curve is greater than a set correlation threshold;

[0036] If yes, then the attention degree of the reference students corresponding to the reference positioning information corresponding to each of the reference time trajectory curves is obtained; when there is an attention degree greater than the attention degree threshold, a tracking warning signal is generated;

[0037] If not, a separate warning signal is generated; the low-risk warning signal includes a tracking warning signal and a separate warning signal.

[0038] Preferably, generating a warning signal according to the preliminary warning signal comprises the following steps:

[0039] Step 1: Determine whether the preliminary warning signal is a high-risk warning signal; if yes, proceed to step 3; if no, proceed to step 2;

[0040] Step 2: Determine whether the preliminary warning signal is a separate warning signal; if yes, proceed to step 3; if no, proceed to step 4;

[0041] Step 3: Obtain the historical monitoring data corresponding to each location coordinate in the historical positioning information of the target student within the set time, as well as the real-time monitoring data corresponding to the current positioning information; proceed to step 5;

[0042] Step 4: Obtain the historical monitoring data of the positions corresponding to the positioning coordinates in the historical positioning information of the target student and the reference student whose attention degree is greater than the attention threshold within the set time, as well as the real-time monitoring data corresponding to the current positioning information; proceed to step 5;

[0043] Step 5: Input each historical monitoring data and real-time monitoring data into the anomaly analysis model to obtain the anomaly scores corresponding to the historical monitoring data and the real-time monitoring data; add up each anomaly score to obtain the total anomaly score, and when the total anomaly score is greater than the set total anomaly score threshold, generate a warning signal; the anomaly analysis model is obtained through artificial intelligence model training.

[0044] Preferably, the abnormality analysis model is obtained by training an artificial intelligence model, including:

[0045] Obtaining a number of monitoring data and their corresponding abnormal scores from a database; the abnormal scores are scores obtained by experts based on their experience in abnormal analysis of the target students' movements, behaviors and demeanor in the monitoring data. The larger the abnormal score, the more abnormal the corresponding student's movement and behavior, and the greater the degree of attention required; integrating the monitoring data and the abnormal scores into several groups of training data and test data;

[0046] The artificial intelligence model is trained using training data, and the trained artificial intelligence model is tested using test data, and finally an anomaly analysis model is obtained whose input is monitoring data and output is an anomaly score; wherein, when in use, historical monitoring data and real-time monitoring data are input as monitoring data into the anomaly analysis model, and the artificial intelligence model includes a BP neural network model and an RBF neural network model.

[0047] Compared with the prior art, the beneficial effects of this application are:

[0048] 1. This application obtains basic information about the campus, builds a three-dimensional campus model based on the campus layout data and building data in the basic information; obtains the recent historical positioning information of each student, divides the three-dimensional campus model into several management areas based on the historical positioning information, and analyzes the management areas to obtain the supervision coefficient representing the characteristics of each management area; and generates the corresponding attention level of each student based on the recent performance and status of each student; obtains the real-time positioning information of each student, and generates early warning signals based on the real-time positioning information, attention level and supervision coefficient of each management area, as well as the monitoring video of each monitoring device; enables the relevant data on students' safety at school to be analyzed from multiple perspectives to improve the accuracy of student safety analysis, timely discover students' abnormal situations, and ensure students' safety at school.

[0049] 2. This application generates the supervision coefficient of each management area in each time period according to the characteristics of each management area in each time period, so as to facilitate targeted analysis of each management area. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0051] Figure 1 This is a schematic diagram of the principle of the smart campus management system in this application;

[0052] Figure 2 This is a schematic diagram of the method steps of the smart campus management system in this application;

[0053] Figure 3 This is a partial flow chart for generating warning signals in this application. DETAILED DESCRIPTION

[0054] The technical solution of the present application will be described clearly and completely below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0055] See also Figure 1-Figure 2 , the first aspect of the present application provides a smart campus management system based on student safety, including: a data collection module, a regional assessment module, a data analysis module, a danger warning module, a supervision center and a database;

[0056] Data acquisition module: obtains the location information of students in real time through the electronic student ID card connected to it; the location information includes the location coordinates of the corresponding students. It can be understood that the electronic student ID card has a location function; and obtains real-time monitoring data through the monitoring and acquisition equipment connected to it. The real-time monitoring data is the monitoring data collected by each monitoring and acquisition equipment; the monitoring and acquisition equipment includes a monitoring camera;

[0057] Regional assessment module: obtain the basic information of the campus and several historical positioning information of each student through the database, and build the corresponding campus three-dimensional model according to the construction information in the basic information; the campus three-dimensional model is a three-dimensional model of the school constructed according to the basic information of the corresponding school; the understandable three-dimensional model and the student positioning share a three-dimensional coordinate system; the basic information is the basic data of the campus, including building information and the location of each management area; the building information is the layout size and other data corresponding to the campus buildings; the management area is the various areas in the campus; the campus three-dimensional model is divided into several management areas according to the historical positioning information, and the supervision coefficient of each management area is generated according to the positioning information in the management area; the historical positioning information is a collection of the historical positioning information of the students, including positioning coordinates, etc.; the management area is the various areas obtained by dividing the campus three-dimensional model according to the historical positioning information; the supervision coefficient represents the degree of supervision of each management area in each time period, and the larger the supervision coefficient, the higher the degree of supervision.

[0058] Data analysis module: obtain the recent historical location information and historical performance data of each student through the database; the historical performance data is the performance data recorded by the corresponding student recently, including test scores, learning attitudes, and punishments, etc.; generate the attention level of the corresponding student based on the historical location information and historical performance data; the attention level indicates the attention level that the corresponding student needs to receive. The greater the attention level, the greater the probability that the corresponding student will be abnormal, and the greater the attention level required; and,

[0059] Acquire the real-time location information of each student, and generate a preliminary warning signal based on the location information and the corresponding attention level of each student; generate a warning signal based on the preliminary warning signal; the preliminary warning signal indicates a warning that the corresponding student may be abnormal or dangerous, which is obtained based on the location information of the corresponding student; the warning signal is a warning that the corresponding student may be abnormal or dangerous, which is obtained by combining the preliminary warning signal with the recent monitoring status of the corresponding student;

[0060] Danger warning module: obtains warning signals and sends them to the supervision center for warning; the supervision center is a supervision platform used by parents and teachers, as well as school leaders and staff, which can display students' trajectories, locations and surveillance videos, as well as alarm signals.

[0061] This embodiment obtains basic information of the campus, builds a three-dimensional model of the campus according to the campus layout data and building data in the basic information; obtains the historical positioning information of each student in recent time, divides the three-dimensional model of the campus into several management areas according to the historical positioning information, and analyzes the management areas to obtain the supervision coefficient representing the characteristics of each management area; and generates the corresponding attention degree of each student according to the performance and status of each student in recent time; obtains the real-time positioning information of each classmate, and generates an early warning signal according to the real-time positioning information, the attention degree and the supervision coefficient of each management area, and the monitoring video of each monitoring device; so that the relevant data of students' safety in school can be analyzed from multiple perspectives to improve the accuracy of student safety analysis, timely discover students' abnormal conditions, and ensure students' safety in school.

[0062] Construct a corresponding three-dimensional campus model according to the construction information in the basic information, including: extracting the campus layout data and the architectural data of each building in the construction information, and constructing the corresponding three-dimensional campus model based on the campus layout data and the architectural data of each building; at the same time, establishing a corresponding three-dimensional coordinate system; and projecting the three-dimensional campus model into the three-dimensional coordinate system; such as taking the central axis of the campus as the y-axis, the ground to the air as the z-axis, and the direction perpendicular to the y-axis and the z-axis as the x-axis; or taking the exact center of the campus as the origin, selecting the x, y, and z axes to establish a three-dimensional coordinate system. It can be understood that the three-dimensional coordinate system here is the same as the three-dimensional coordinate system of the positioning coordinates of the electronic student card parameters, or can be converted.

[0063] According to the historical positioning information, the campus 3D model is divided into several management areas, including:

[0064] Obtain positioning coordinates from several historical positioning information; integrate the positioning coordinates into several feature vectors; for example, the feature vector is {x, y, z}; perform cluster analysis on several feature vectors to obtain several clusters; the cluster analysis can use K-means clustering algorithm and density cluster analysis method, etc.; obtain the feature vector corresponding to the boundary of each cluster, extract the positioning coordinates in each feature vector as the boundary positioning coordinates, mark each positioning coordinate of the several clusters into the three-dimensional model of the campus, and obtain several management areas; this embodiment divides the areas where students or employees often appear in the campus and the areas where students and employees do not often appear through the above method, and management can be carried out for each area later.

[0065] In this embodiment, the campus is divided into several management areas by the positioning coordinates of several students or on-campus employees, and targeted management can be performed on each management area, so as to improve the efficiency of student safety management.

[0066] Generate the supervision coefficient of each management area according to the positioning information in the management area, including: obtaining the positioning ID corresponding to the positioning coordinates in several historical positioning information in each management area; the positioning ID includes the positioning ID of students, teachers, employees and other people who enter the campus through formal channels, obtaining the number of positioning coordinates RSit, the number of positioning IDs ISit and the positioning duration SCit of each positioning ID in each set time period in the management area, the positioning duration is the length of time that the positioning of the corresponding positioning ID appears in the corresponding management area within the set time period; i is the number of the management area, i=1, 2, ..., I; I is the total number of management areas; t is the number of the set time period, t=1, 2, ..., T; T is the total number of the set time periods; the set time period is set by relevant management personnel, and can be set as a time period every fixed time period, such as setting a time period every hour; it can also be divided according to class time and get out of class time, and this division method will make the subsequent analysis of student positioning better;

[0067] By formula The supervision coefficient JXit of the management area numbered i in the set time period numbered t is calculated; among them, β1, β2 and β3 are adjustment coefficients, the specific values ​​are set according to experience, and are mainly used to adjust the distribution of the values ​​of different items; α1, α2 and α3 are weight coefficients, the specific values ​​are set according to experience; ZIS is the total number of location IDs of the corresponding school; SDt is the time period length of the set time period numbered t. The larger the supervision coefficient, the greater the intensity of supervision required for the corresponding area;

[0068] In this embodiment, the supervision coefficient of each set time period in each management area is calculated through the above formula. The more students or related personnel appear in a set time period in the management area, the more times they appear, and the longer the time when there are students or related personnel in the set time period; that is, the personnel in the corresponding management areas in the corresponding time period can supervise each other, and the probability of accidents is small; the degree of supervision is also small, and the corresponding supervision coefficient is set small; such as the playground during the break, the classroom during the class time, and the cafeteria at the meal time; on the contrary, if there are few students or personnel in a management area, the probability of accidents in the corresponding area is large, and if an accident occurs, it is difficult to be discovered, resulting in missing the time for timely processing, such as the school rooftop.

[0069] This embodiment generates the supervision coefficient of each management area in each time period according to the characteristics of each management area in each time period, so as to facilitate targeted analysis of each management area.

[0070] Generating the attention level of the corresponding student according to the historical positioning information and the historical performance data, including: obtaining the recent historical performance data of each student, and inputting the historical performance data into a supervision scoring model to obtain the supervision score of the corresponding student; the supervision scoring model is obtained by training an artificial intelligence model;

[0071] Obtain the student's recent historical location information, extract the location coordinates and corresponding location time in the historical location information; map each location coordinate to the campus 3D model, and combine the location time of each location coordinate to obtain the number of students appearing in each set time period in each management area CCit;

[0072] By formula The corresponding student's attention GZ is calculated; among them, γ1 and γ2 are proportional coefficients, and the specific values ​​are set according to experience; β4 is the adjustment coefficient, and the specific value is set according to experience; JP is the supervision score of the corresponding student.

[0073] This embodiment calculates the attention level of each student through the above formula. The more corresponding students appear in the management area with a larger supervision coefficient, the greater the possibility that the corresponding students will be abnormal; the larger the value of their attention level is; the worse the corresponding students' recent performance is, the greater the possibility that they will be abnormal, and the larger the value of their attention level is.

[0074] The supervision degree scoring model is obtained through artificial intelligence model training, including: obtaining a number of historical performance data and the corresponding supervision degree scores through a database; the supervision degree scores are scores given by experts or relevant teachers on the degree to which students need supervision based on the performance data of the corresponding students. For example, if the corresponding students usually show mania, depression, etc., the corresponding supervision degree scores are high, and if the corresponding students usually show normal performance, the corresponding supervision degree scores are low; the historical performance data and the corresponding supervision degree scores are integrated into several groups of training data and test data;

[0075] The artificial intelligence model is trained using training data, and the trained artificial intelligence model is tested using test data. Specifically, after the artificial intelligence model is trained using a number of training data, the historical performance data in the test data is input into the trained artificial intelligence model to output a corresponding supervision score; when the difference between the output supervision score and the supervision score recorded in the corresponding test data is within an acceptable range, the next set of test data is tested; otherwise, the relevant parameters of the artificial intelligence model are adjusted, and the test is performed again using the current set of test data; until a set proportion of the test data passes the test; finally, a feature scoring model is obtained with the historical performance data as input and the supervision score as output; the artificial intelligence model includes a BP neural network model and an RBF neural network model.

[0076] A preliminary warning signal is generated based on the positioning information and the attention level of each student, including: when the positioning information of a student is in a management area where the supervision coefficient is greater than a set supervision coefficient threshold; the supervision coefficient threshold is set according to expert experience; then the attention level of the student is obtained, and it is determined whether the attention level is less than the set attention threshold;

[0077] If yes, then the historical location information of the student within the set time is obtained and recorded as the target location information, and other historical location information within the set time is obtained, the historical location information includes the location information of the student, the location information of the teacher and related personnel, and recorded as the reference location information; a low-risk warning signal is generated according to the attention of the target location information and the reference location information; the low-risk warning signal is a warning signal with a low probability of abnormality;

[0078] If not, a high-risk warning signal is generated; the preliminary warning signal includes a high-risk warning signal and a low-risk warning signal; a high-risk warning signal is a warning signal with a high probability of abnormality; students with high attention levels appear in the danger zone, and the probability of danger is high;

[0079] Generate a preliminary warning signal according to the attention of the target positioning information and the reference positioning information, including: extracting the positioning coordinates and positioning time in the target positioning information and each reference positioning information, fitting the positioning coordinates into a target time trajectory curve in the order of positioning time, which can be expressed as MG (xmj, ymj, zmj, tj); the reference time trajectory curve can be expressed as CG (xcj, ycj, zcj, tj);

[0080] By formula The correlation degree GL between the corresponding reference time trajectory curve and the target time trajectory curve is calculated; wherein, j is the number of the set time node, j=1, 2, ..., J; J is the total number of the set time nodes; tj is the set time node numbered j; the set time node can be evenly set in the considered time period; xmj, ymj and zmj are respectively the x coordinate, y coordinate and z coordinate in the positioning coordinates corresponding to the set time node numbered j in the target time trajectory area; xcj, ycj and zcj are respectively the x coordinate, y coordinate and z coordinate in the positioning coordinates corresponding to the set time node numbered j in the reference time trajectory area; δj is the weight coefficient of the set time node numbered j, and the closer the data is to the current time, the more it can reflect the actual situation; in this embodiment, the closer to the current set time node, the larger the corresponding weight coefficient is set; the specific value is set according to experience; β5 is the adjustment coefficient, and the specific value is set according to experience; β5 can be a dimensioned adjustment coefficient used for the unit of regional coordinates;

[0081] This embodiment uses the above formula to determine whether the trajectories of different students are related, as well as the correlation situation. When the physical distances of different students at the same time point are closer, it means that the correlation between the two students is greater, and the corresponding correlation degree is set to be greater; when the physical distances at multiple time points are close, it means that the correlation between the two students is greater, and the corresponding correlation degree is set to be greater.

[0082] Determine whether the correlation between the target time trajectory curve and each reference time trajectory curve is greater than a set correlation threshold; the specific value of the correlation threshold is set based on experience;

[0083] If yes, then obtain the attention level of the reference students corresponding to the reference positioning information corresponding to each of the reference time trajectory curves; when the attention level is greater than the attention level threshold, the attention level threshold is set according to expert experience; then generate a tracking warning signal; the tracking warning signal indicates that the corresponding student may be tracked by other students with higher attention levels, which may easily lead to danger;

[0084] If not, a separate warning signal is generated; the separate warning signal indicates that the student is alone in a dangerous area and is prone to danger; the low-risk warning signal includes a tracking warning signal and a separate warning signal.

[0085] The lower the corresponding attention level of students appearing in the management area with a high supervision coefficient, the smaller the probability of abnormality. If other students with lower attention levels appear in the corresponding management area with the current student at this time, the probability of accidents will be further reduced; such as several classmates going out for a walk together; if no other people appear in the corresponding area at this time, it means that the probability of abnormality of the student will increase; further attention needs to be paid to the corresponding students; if students with higher attention levels appear in the corresponding area at the same time, it means that the probability of abnormality of the student will increase; tracking may occur; further attention needs to be paid to the corresponding students.

[0086] See also Figure 3 , generating an early warning signal according to the preliminary early warning signal, including the following steps:

[0087] Step 1: Determine whether the preliminary warning signal is a high-risk warning signal; if yes, proceed to step 3; if no, proceed to step 2;

[0088] Step 2: Determine whether the preliminary warning signal is a separate warning signal; if yes, proceed to step 3; if no, proceed to step 4;

[0089] Step 3: Obtain the historical monitoring data corresponding to each location coordinate in the historical positioning information of the target student within the set time, as well as the real-time monitoring data corresponding to the current positioning information; proceed to step 5;

[0090] Step 4: Obtain the historical monitoring data of the positions corresponding to the positioning coordinates in the historical positioning information of the target student and the reference student whose attention degree is greater than the attention threshold within the set time, as well as the real-time monitoring data corresponding to the current positioning information; proceed to step 5;

[0091] Step 5: Input each historical monitoring data and real-time monitoring data into the anomaly analysis model to obtain the anomaly scores corresponding to the historical monitoring data and the real-time monitoring data; add up each anomaly score to obtain the total anomaly score. When the total anomaly score is greater than the set total anomaly score threshold, the total anomaly score threshold is set according to expert experience; then a warning signal is generated; the anomaly analysis model is obtained through artificial intelligence model training.

[0092] The abnormal analysis model is obtained through artificial intelligence model training, including: obtaining a number of monitoring data and its corresponding abnormal scores through a database; the abnormal scores are scores obtained by experts based on experience to analyze the actions, behaviors and demeanor of target students in the monitoring data. The larger the abnormal score, the more abnormal the corresponding student's actions and behaviors are, and the greater the degree of attention required; integrating the monitoring data and abnormal scores into several groups of training data and test data;

[0093] The artificial intelligence model is trained using training data, and the trained artificial intelligence model is tested using test data, ultimately obtaining an anomaly analysis model whose input is monitoring data and output is an anomaly score; wherein, when in use, historical monitoring data and real-time monitoring data are input as monitoring data into the anomaly analysis model, and the artificial intelligence model includes a BP neural network model and an RBF neural network model; the training and testing process of the anomaly analysis model is similar to that of the supervision degree scoring model.

[0094] Part of the data in the above formula is calculated by removing the dimension and taking its numerical value. The formula is a formula closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0095] How this application works:

[0096] This application obtains basic information about the campus, builds a three-dimensional campus model based on the campus layout data and building data in the basic information; obtains the recent historical positioning information of each student, divides the three-dimensional campus model into several management areas based on the historical positioning information, and analyzes the management areas to obtain the supervision coefficient representing the characteristics of each management area; and generates the corresponding attention level of each student based on the recent performance and status of each student; obtains the real-time positioning information of each classmate, and generates early warning signals based on the real-time positioning information, attention level and supervision coefficient of each management area, as well as the monitoring video of each monitoring device; so that the relevant data on students' safety at school can be analyzed from multiple perspectives to improve the accuracy of student safety analysis, detect students' abnormal situations in a timely manner, and ensure students' safety at school.

[0097] The above embodiments are only used to illustrate the technical method of the present application and are not intended to limit it. Although the present application has been described in detail with reference to the preferred embodiments, a person of ordinary skill in the art should understand that the technical method of the present application may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present application.

Claims

1. A smart campus management system based on student safety, comprising: Data collection module, regional assessment module, data analysis module, danger warning module and database; characterized in that, The data acquisition module: obtains the student's location information through the electronic student certificate connected to it; and obtains real-time monitoring data through the monitoring acquisition device connected to it; The regional assessment module: obtains basic information of the campus and several historical positioning information of each student through the database, and constructs a corresponding three-dimensional campus model according to the construction information in the basic information; divides the three-dimensional campus model into several management areas according to the historical positioning information, and generates a supervision coefficient for each management area according to the historical positioning information in the management area; The step of generating the supervision coefficient of each management area according to the historical positioning information in the management area includes: Obtain the positioning ID corresponding to the positioning coordinates in several historical positioning information in each management area; obtain the number of positioning coordinates RSit, the number of positioning IDs ISit and the positioning duration SCit of each positioning ID in each set time period in the management area, i is the number of the management area, i=1, 2, ..., I; I is the total number of management areas; t is the number of the set time period, t=1, 2, ..., T; T is the total number of the set time periods; The data analysis module: obtains the recent historical location information and historical performance data of each student through the database; generates the attention level of the corresponding student according to the historical location information and historical performance data; and Obtaining the real-time location information of each student, generating a preliminary warning signal based on the real-time location information and the corresponding attention level of each student; generating a warning signal based on the preliminary warning signal; The generating of the corresponding student's attention degree according to the historical positioning information and the historical performance data includes: Obtaining recent historical performance data of each student, and inputting the historical performance data into a supervision scoring model to obtain a supervision score of the corresponding student; the supervision scoring model is obtained by training an artificial intelligence model; Obtain the student's recent historical location information, extract the location coordinates and corresponding location time in the historical location information; map each location coordinate to the campus 3D model, and combine the location time of each location coordinate to obtain the number of students appearing in each set time period in each management area CCit; By formula The corresponding student's attention GZ is calculated; where γ1 and γ2 are proportional coefficients; β4 is the adjustment coefficient; JP is the corresponding student's supervision score; The generating of a preliminary warning signal according to the real-time positioning information and the corresponding attention level of each student includes: When there is a student's location information in a management area whose supervision coefficient is greater than a set supervision coefficient threshold, the student's attention is obtained, and it is determined whether the attention is less than the set attention threshold; If yes, then the historical location information of the student within the set time is obtained and recorded as the target location information, and other historical location information within the set time is obtained and recorded as the reference location information; a low-risk warning signal is generated according to the attention of the target location information and the reference location information; If not, a high-risk warning signal is generated; the preliminary warning signal includes a high-risk warning signal and a low-risk warning signal; The danger warning module obtains warning signals and sends the warning signals to the supervision center for warning.

2. According to claim 1, a smart campus management system based on student safety is characterized in that: The step of constructing a corresponding campus three-dimensional model according to the construction information in the basic information includes: The campus layout data and the architectural data of each building in the construction information are extracted, and a corresponding campus three-dimensional model is constructed based on the campus layout data and the architectural data of each building.

3. A smart campus management system based on student safety according to claim 2, characterized in that: The campus three-dimensional model is divided into several management areas according to the historical positioning information, including: Acquire positioning coordinates from a number of historical positioning information; integrate the positioning coordinates into a number of feature vectors; perform cluster analysis on a number of feature vectors to obtain a number of clusters; acquire the feature vector corresponding to the boundary of each cluster, extract the positioning coordinates in each feature vector as the boundary positioning coordinates, mark each positioning coordinate of the number of clusters into the three-dimensional model of the campus, and obtain a number of management areas.

4. A smart campus management system based on student safety according to claim 2, characterized in that: The calculation method of the regulatory coefficient includes: By formula The supervision coefficient JXit of the management area numbered i in the set time period numbered t is calculated; among them, β1, β2 and β3 are adjustment coefficients; α1, α2 and α3 are weight coefficients; ZIS is the total number of location IDs of the corresponding schools; SDt is the time period length of the set time period numbered t.

5. According to claim 1, a smart campus management system based on student safety is characterized in that: The supervision degree scoring model is obtained through artificial intelligence model training, including: Acquire a number of historical performance data and corresponding supervision scores through a database; the supervision score is a score given by an expert or a relevant teacher on the degree to which a student needs supervision based on the performance data of the corresponding student, and integrate the historical performance data and the corresponding supervision score into a number of sets of training data and test data; The artificial intelligence model is trained using training data, and the trained artificial intelligence model is tested using test data, ultimately obtaining a feature scoring model whose input is historical performance data and whose output is a regulatory score; the artificial intelligence model includes a BP neural network model and an RBF neural network model.

6. A smart campus management system based on student safety according to claim 1, characterized in that: The generating of a low-risk warning signal according to the target positioning information and the reference positioning information attention level includes: Extract the positioning coordinates and positioning time from the target positioning information and each reference positioning information, and fit the positioning coordinates into the target time trajectory curve MG (xmj, ymj, zmj, tj) and the reference time trajectory curve CG (xcj, ycj, zcj, tj) in the order of positioning time; By formula The correlation degree GL between the corresponding reference time trajectory curve and the target time trajectory curve is calculated; wherein j is the number of the set time node, j=1, 2, ..., J; J is the total number of the set time nodes; tj is the set time node numbered j; xmj, ymj and zmj are respectively the x coordinate, y coordinate and z coordinate in the positioning coordinate corresponding to the set time node numbered j in the target time trajectory area; xcj, ycj and zcj are respectively the x coordinate, y coordinate and z coordinate in the positioning coordinate corresponding to the set time node numbered j in the reference time trajectory area; δj is the weight coefficient of the set time node numbered j; β5 is the adjustment coefficient; Determine whether the correlation between the target time trajectory curve and each reference time trajectory curve is greater than a set correlation threshold; If yes, then the attention degree of the reference students corresponding to the reference positioning information corresponding to each of the reference time trajectory curves is obtained; when there is an attention degree greater than the attention degree threshold, a tracking warning signal is generated; If not, a separate warning signal is generated; the low-risk warning signal includes a tracking warning signal and a separate warning signal.

7. A smart campus management system based on student safety according to claim 1, characterized in that: The step of generating a warning signal according to the preliminary warning signal comprises the following steps: Step 1: Determine whether the preliminary warning signal is a high-risk warning signal; if yes, proceed to step 3; if no, proceed to step 2; Step 2: Determine whether the preliminary warning signal is a separate warning signal; if yes, proceed to step 3; if no, proceed to step 4; Step 3: Obtain the historical monitoring data corresponding to each location coordinate in the historical positioning information of the target student within the set time, as well as the real-time monitoring data corresponding to the current positioning information; proceed to step 5; Step 4: Obtain the historical monitoring data of the positions corresponding to the positioning coordinates in the historical positioning information of the target student and the reference student whose attention degree is greater than the attention threshold within the set time, as well as the real-time monitoring data corresponding to the current positioning information; proceed to step 5; Step 5: Input each historical monitoring data and real-time monitoring data into the anomaly analysis model to obtain the anomaly scores corresponding to the historical monitoring data and the real-time monitoring data; add up each anomaly score to obtain the total anomaly score, and when the total anomaly score is greater than the set total anomaly score threshold, generate a warning signal; the anomaly analysis model is obtained through artificial intelligence model training.

8. A smart campus management system based on student safety according to claim 7, characterized in that: The abnormal analysis model is obtained through artificial intelligence model training, including: Obtaining a number of monitoring data and their corresponding abnormal scores from a database; the abnormal scores are scores obtained by experts through abnormal analysis of the target students' actions, behaviors and demeanor in the monitoring data based on their experience; integrating the monitoring data and the abnormal scores into a number of sets of training data and test data; The artificial intelligence model is trained using training data, and the trained artificial intelligence model is tested using test data, and finally an anomaly analysis model is obtained whose input is monitoring data and output is an anomaly score; wherein the artificial intelligence model includes a BP neural network model and an RBF neural network model.

Images