A method for monitoring personnel safety behavior based on electronic fence technology

Abstract

The invention discloses a personnel safety behavior monitoring method based on electronic fence technology, comprising: when the dynamic adjustment of the electronic fence inclination is triggered, a fuzzy control algorithm is used to dynamically calculate the inclination adjustment value of the electronic fence according to the distance between the personnel and the electronic fence, the closer the distance is, the greater the inclination adjustment amplitude is, the inclination parameters of the electronic fence after dynamic adjustment are generated, and the electronic fence model is updated in real time; according to the optimal inclination parameters, the electronic fence model is updated and optimized, and at the same time, a mapping relationship between regional characteristics and inclination parameters is established according to the terrain characteristics and working characteristics of different regions, so as to form an adaptive dynamic adjustment strategy for the inclination of the electronic fence and improve the scene adaptability of the electronic fence.

Description

Technical Field

[0001] The present invention belongs to the technical field of safety protection, and in particular relates to a method for monitoring personnel safety behavior based on electronic fence technology. Background Art

[0002] When electronic fence technology is applied to personnel safety behavior monitoring, how to set the inclination of the fence is a key technical issue. If the inclination is set too large, it may lead to an increase in the false alarm rate, the system frequently issues invalid warnings, and interferes with normal working order; if the inclination is set too small, it may not be able to effectively identify the dangerous behavior of personnel, and there will be blind spots in safety monitoring. At the same time, the inclination setting also needs to take into account the terrain characteristics and working characteristics of different areas. For example, in areas with large slopes, the fence inclination should be appropriately adjusted, while a unified inclination standard can be used in flat areas. In addition, the dynamic adjustment of the fence inclination is also a technical challenge. It is necessary to adaptively adjust the inclination setting according to the real-time location and behavior trajectory of the personnel to improve the intelligence level of safety monitoring. In summary, the setting of the electronic fence inclination needs to weigh the false alarm rate, missed alarm rate, scene adaptability and intelligence level in many aspects, and find the optimal balance point to achieve efficient and reliable personnel safety behavior monitoring. However, this technology is still lacking. Summary of the invention

[0003] In order to solve the above technical problems, the present invention proposes a personnel safety behavior monitoring method based on electronic fence technology to solve the problems existing in the above prior art.

[0004] To achieve the above object, the present invention provides a method for monitoring personnel safety behavior based on electronic fence technology, comprising the following steps:

[0005] Acquire the terrain characteristics and working characteristic information of the target area through the pre-established electronic fence, set the initial inclination angle of the electronic fence based on the terrain characteristics and working characteristic information, and obtain the electronic fence adapted to the scene;

[0006] The behavior trajectory of the target person is obtained through the electronic fence adapted to the scene, and the movement direction and speed of the target person are obtained by analyzing the behavior trajectory. When the target person is detected moving toward the electronic fence, the inclination adjustment value of the electronic fence is dynamically calculated using a fuzzy control algorithm, and the electronic fence is updated in real time according to the inclination adjustment value;

[0007] After updating the electronic fence model, the latest location information of the target person is obtained. When the target person intersects the electronic fence and the inclination value of the electronic fence is greater than the warning threshold, it is determined that the target person has dangerous behavior.

[0008] Preferably, the target person safety behavior monitoring method further includes dynamically optimizing the inclination value according to the monitoring data, and the dynamic optimization method includes:

[0009] During the monitoring process, the false alarm rate and missed alarm rate data of the electronic fence are continuously obtained, and the false alarm rate and missed alarm rate data are statistically analyzed through the data analysis algorithm to obtain the false alarm rate and missed alarm rate curves under different inclination settings, and the inclination value at the balance point of the false alarm rate and missed alarm rate is selected as the optimal inclination parameter of the electronic fence;

[0010] According to the optimal inclination parameters, the electronic fence model is updated and optimized, and according to the terrain characteristics and working characteristics of different regions, a mapping relationship between regional characteristics and inclination parameters is established to form an adaptive dynamic adjustment strategy for the inclination of the electronic fence.

[0011] Preferably, the method for setting the initial inclination angle of the electronic fence includes:

[0012] The initial inclination angle of the electronic fence is set by the support vector machine algorithm. When the target area is a large slope area, the inclination angle value of the electronic fence is increased; when the target area is a flat area, a unified inclination angle standard is used for adjustment;

[0013] Through the decision tree algorithm, the terrain characteristics of the target area are judged to obtain the electronic fence inclination parameters suitable for the scene;

[0014] According to the working characteristic information in the electronic fence model, the random forest algorithm is used to optimize and adjust the inclination parameters of the electronic fence.

[0015] Preferably, the method for obtaining the moving direction and speed of the target person includes:

[0016] The location data of the target person is collected and stored in the database in chronological order to form the motion trajectory data of the target person; the motion trajectory data is smoothed using the Kalman filter algorithm, and the movement direction and average speed of the target person are determined by calculating the Euclidean distance and time difference between two adjacent position points based on the speed formula.

[0017] Preferably, the method for updating the electronic fence in real time includes:

[0018] According to the preset electronic fence area range, when the current position of the target person is within the preset electronic fence area range, obtaining the movement direction of the target person;

[0019] When the target person moves toward the electronic fence area, a dynamic adjustment mechanism of the electronic fence inclination is triggered;

[0020] According to the distance between the target person and the boundary of the electronic fence and the movement speed, dynamically calculate the adjustment angle value of the electronic fence inclination, and control the inclination angle of the electronic fence in real time based on the adjustment angle value;

[0021] When the moving direction of the target person changes or moves away from the electronic fence, the electronic fence inclination adjustment is stopped and the original inclination setting of the electronic fence is restored.

[0022] Preferably, the method for calculating the adjustment angle value of the electronic fence inclination includes:

[0023] A rule base of fuzzy control rules is preset, and the personnel distance is used as the input variable. The adjustment range of the electronic fence inclination is determined by the Mamdani fuzzy reasoning method, and the center of gravity defuzzification method is used to synthesize the reasoning results to obtain the specific inclination adjustment value.

[0024] Preferably, preprocessing is performed before statistical analysis of the false alarm rate and missed alarm rate data. The preprocessing method includes:

[0025] The 3σ principle is used to eliminate outliers beyond the normal range, and the Kalman filter algorithm is used to filter the noise data to obtain the preprocessed data set.

[0026] Preferably, the method for obtaining the false alarm rate and missed alarm rate curves under different tilt angle settings includes:

[0027] Based on the preprocessed data set, a regression analysis algorithm is used to perform modeling, wherein the regression analysis algorithm includes linear regression and polynomial regression;

[0028] By means of the regression analysis algorithm, a first functional relationship expression between the inclination angle setting and the false alarm rate and a second functional relationship expression between the inclination angle setting and the missed alarm rate are obtained;

[0029] A cross-validation method is used to evaluate the prediction performance of the first functional relationship expression and the second functional relationship expression under different inclination angle settings to obtain a regression model that meets preset requirements; the prediction performance includes the prediction accuracy of the false alarm rate and the prediction accuracy of the missed alarm rate;

[0030] The regression model is solved to obtain false alarm rate and missed alarm rate curves under different inclination angle settings.

[0031] Compared with the prior art, the present invention has the following advantages and technical effects:

[0032] The present invention discloses a method for monitoring personnel safety behaviors based on electronic fence technology. The method sets the initial inclination of the electronic fence according to the terrain characteristics and working characteristics of the target area through a machine learning algorithm, and uses a trajectory tracking algorithm to analyze the movement direction and speed of the personnel in real time. When a person is detected approaching the electronic fence, a fuzzy control algorithm is used to dynamically adjust the inclination of the fence, and the closer the distance, the greater the adjustment range. The method also judges dangerous behaviors by comparing the inclination value at the intersection with the warning threshold, and continuously analyzes the false alarm rate and missed alarm rate data, and selects the inclination value of the balance point as the optimal parameter. Finally, the present invention establishes a mapping relationship between regional characteristics and inclination parameters, forms an adaptive dynamic adjustment strategy, and improves the scene adaptability and monitoring accuracy of the electronic fence. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0034] Figure 1 The figure is a flow chart of a method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0035] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0036] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0037] Embodiment 1

[0038] like Figure 1 As shown, this embodiment provides a method for monitoring personnel safety behavior based on electronic fence technology, including the following steps:

[0039] The terrain characteristics and working characteristic information of the target area are obtained through the pre-established electronic fence, and the initial inclination angle of the electronic fence is set based on the terrain characteristics and working characteristic information to obtain an electronic fence adapted to the scene;

[0040] The behavior trajectory of the target person is obtained through the electronic fence adapted to the scene, and the behavior trajectory is analyzed to obtain the movement direction and speed of the target person. When the target person is detected moving towards the electronic fence, the fuzzy control algorithm is used to dynamically calculate the inclination adjustment value of the electronic fence, and the electronic fence is updated in real time according to the inclination adjustment value;

[0041] After updating the electronic fence model, the latest location information of the target person is obtained. When the target person intersects the electronic fence and the inclination value of the electronic fence is greater than the warning threshold, it is determined that the target person has dangerous behavior.

[0042] Furthermore, the target person safety behavior monitoring method further includes dynamically optimizing the inclination value according to the monitoring data, and the dynamic optimization method includes:

[0043] During the monitoring process, the false alarm rate and missed alarm rate data of the electronic fence are continuously obtained, and the false alarm rate and missed alarm rate data are statistically analyzed through the data analysis algorithm to obtain the false alarm rate and missed alarm rate curves under different inclination settings, and the inclination value at the balance point of the false alarm rate and missed alarm rate is selected as the optimal inclination parameter of the electronic fence;

[0044] According to the optimal inclination parameters, the electronic fence model is updated and optimized, and according to the terrain characteristics and working characteristics of different regions, a mapping relationship between regional characteristics and inclination parameters is established to form an adaptive dynamic adjustment strategy for the inclination of the electronic fence.

[0045] The specific steps are as follows:

[0046] Step S101, based on the pre-established electronic fence model, obtain the terrain characteristics and working characteristics information of the target area, set the initial inclination angle of the electronic fence through a machine learning algorithm, and appropriately increase the inclination value for areas with larger slopes. For flat areas, adopt a unified inclination standard to obtain electronic fence inclination parameters suitable for different scenarios.

[0047] Specifically, according to the pre-established electronic fence model, the terrain characteristics and working characteristics information of the target area are obtained, and the initial inclination angle of the electronic fence is set through the support vector machine algorithm. If the target area is an area with a large slope, the inclination value of the electronic fence is appropriately increased; if the target area is a flat area, a unified inclination standard is adopted. The decision tree algorithm is used to judge the terrain characteristics of the target area and obtain the inclination parameters of the electronic fence adapted to different scenarios. According to the working characteristics information in the electronic fence model, the random forest algorithm is used to optimize and adjust the inclination parameters of the electronic fence.

[0048] Step S102, obtain the real-time location information of at least one person, continuously collect the location data of the person according to the preset time interval, analyze the behavior trajectory of the person through the trajectory tracking algorithm, determine the movement direction and speed of the person, and if it is detected that the person is moving in the direction of the electronic fence, the dynamic adjustment mechanism of the inclination angle of the electronic fence is triggered.

[0049] Specifically, real-time location information of at least one target person is obtained, the real-time location information is obtained by global positioning system positioning or indoor positioning technology, and the location data of the target person is continuously collected at preset time intervals;

[0050] The collected position data is stored in chronological order to obtain the movement trajectory data of the target person;

[0051] The motion trajectory data is analyzed using a pre-established trajectory tracking algorithm, and the motion direction and average speed of the target person in the time period are determined by calculating the distance and time difference between two adjacent position points;

[0052] According to the preset electronic fence area range, determine whether the current position of the target person is within the warning area of ​​the electronic fence boundary, and if so, obtain the movement direction of the target person;

[0053] If the target person's movement direction is toward the electronic fence area, the electronic fence inclination angle dynamic adjustment mechanism is triggered;

[0054] According to the distance between the target person and the boundary of the electronic fence and the movement speed, the angle value of the electronic fence inclination adjustment is dynamically calculated, and the inclination angle of the electronic fence is changed in real time by controlling the actuator;

[0055] If the moving direction of the target person changes or moves away from the electronic fence, the electronic fence inclination adjustment is stopped and the original inclination setting of the electronic fence is restored.

[0056] Step S103, when the dynamic adjustment of the electronic fence inclination is triggered, the fuzzy control algorithm is used to dynamically calculate the inclination adjustment value of the electronic fence according to the distance between the person and the electronic fence. The closer the distance, the greater the inclination adjustment range. The dynamically adjusted electronic fence inclination parameters are generated, and the electronic fence model is updated in real time.

[0057] Specifically, the coordinates of the current position of the person are obtained, and the distance between the person and the electronic fence is calculated; according to the preset fuzzy control rules, the relationship between the adjustment range of the electronic fence inclination and the distance between the person and the person is determined; if the distance between the person and the person is less than the preset threshold, the dynamic adjustment of the electronic fence inclination is triggered; the fuzzy control algorithm is used to dynamically calculate the electronic fence inclination adjustment value according to the distance between the person and the person; the adjusted electronic fence inclination parameters are generated, and the electronic fence three-dimensional model data is updated; the updated electronic fence model data is transmitted to the display terminal in real time for display; the changes in the position of the person are continuously tracked, and the inclination of the electronic fence is dynamically adjusted according to the real-time distance between the person and the electronic fence, so as to achieve the real-time deformation effect of the electronic fence.

[0058] Step S104, after updating the electronic fence model, obtain the latest location information of the person and determine whether the person intersects the electronic fence. If an intersection occurs, compare the inclination value of the electronic fence at the intersection with the preset warning threshold. If the inclination value is greater than the warning threshold, trigger an alarm and determine it as a dangerous behavior of the person.

[0059] Specifically, according to the pre-established electronic fence model, the real-time location information of the person is obtained, and the spatial relationship between the location information and the electronic fence model is calculated to determine whether the person's position intersects with the electronic fence. If the person's position intersects with the electronic fence, the inclination value of the electronic fence at the intersection is obtained, and the inclination value is compared with the preset warning threshold to determine whether the inclination value is greater than the warning threshold. If the inclination value is greater than the warning threshold, an alarm is triggered, the crossing behavior is determined as a dangerous behavior of the person, and an alarm message is generated. According to the alarm information, the historical behavior data of the person is trained using the support vector machine algorithm to obtain a dangerous behavior recognition model, which is used to subsequently identify the potential dangerous behavior of the person.

[0060] Step S105, during the monitoring process, continuously obtain the false alarm rate and missed alarm rate data of the electronic fence, perform statistical analysis on the false alarm rate and missed alarm rate through a data analysis algorithm, obtain the false alarm rate and missed alarm rate curves under different inclination settings, and select the inclination value at the balance point of the false alarm rate and missed alarm rate as the optimal inclination parameter of the electronic fence.

[0061] Specifically, the historical data of the false alarm rate and missed alarm rate of the electronic fence under different inclination settings are obtained to construct a data set. The data set is preprocessed to remove outliers and noise data to ensure data quality. According to the preprocessed data set, a regression analysis algorithm is used to establish a mathematical model between the inclination setting and the false alarm rate and missed alarm rate. By solving the mathematical model, the false alarm rate curve and missed alarm rate curve under different inclination settings are obtained. The intersection of the false alarm rate curve and the missed alarm rate curve is determined to obtain the inclination value at the balance point. The inclination value at the balance point is determined as the optimal inclination parameter of the electronic fence. The optimal inclination parameter is applied to the electronic fence system, the operating status of the electronic fence is monitored in real time, and the inclination parameter is continuously optimized.

[0062] Step S106, based on the optimal inclination parameters, the electronic fence model is updated and optimized, and at the same time, a mapping relationship between regional characteristics and inclination parameters is established according to the terrain characteristics and working characteristics of different regions, forming an adaptive electronic fence inclination dynamic adjustment strategy to improve the scene adaptability of the electronic fence.

[0063] Specifically, the terrain characteristic data and working characteristic data of the target area are obtained, and the optimal inclination parameter of the area is determined according to the preset mapping relationship between the regional characteristics and the inclination parameter. The optimal inclination parameter is used to optimize and update the various components of the electronic fence model. If a certain component cannot be applied to the optimal inclination parameter, the adaptive inclination parameter is obtained from the mapping relationship according to its characteristics for updating. After the optimization and update of the electronic fence model is completed, the environmental characteristic data of the current area is obtained in real time to determine whether the preset adaptive adjustment conditions are met. If satisfied, the inclination parameter of the electronic fence is dynamically adjusted according to the mapping relationship. A mapping relationship model between regional characteristics and inclination parameters is established through a machine learning algorithm, and the mapping relationship model is optimized according to the historical data of each area to improve the accuracy of the mapping relationship. The inclination parameter change data and event data during the operation of the electronic fence are obtained, and the association rule mining algorithm is used to find the association rules between the inclination parameter and the fence event, and the mapping relationship model is optimized. The optimized mapping relationship model is applied to the dynamic inclination adjustment strategy of the electronic fence, and the optimal inclination parameter is recommended in real time according to the regional characteristics to improve the scene adaptability of the electronic fence. The optimized inclination parameters are applied to each component of the electronic fence for detection and evaluation. If there is an abnormality, the component is marked as a special component and targeted processing is performed when the inclination parameters are dynamically adjusted.

[0064] The specific implementation is as follows:

[0065] First, according to the pre-established electronic fence model, the terrain characteristics and working characteristics information of the target area are obtained. The initial inclination angle of the electronic fence is set by the support vector machine algorithm. For example, in a mountainous area with a slope of 15°, the inclination value of the electronic fence is set to 25°; in a plain area with an average altitude of 50 meters, a unified inclination standard of 10° is adopted. Then, the terrain characteristics of the target area are judged by the decision tree algorithm. When the terrain undulation exceeds 30%, the inclination angle of the electronic fence is increased by 5°; when the terrain undulation is less than 10%, it is reduced by 3°. According to the working characteristics information in the electronic fence model, the random forest algorithm is used to comprehensively consider factors such as terrain, altitude, and vegetation through 500 decision trees to obtain the optimal value of the inclination parameter of the electronic fence.

[0066] By wearing a GPS positioning device on the target person, the position data is continuously collected at intervals of 5 seconds, and the collected position data is stored in the database in chronological order to form the motion trajectory data of the target person. The Kalman filter algorithm is used to smooth the motion trajectory data. By calculating the Euclidean distance and time difference between two adjacent position points, the speed formula v=s / t is used to determine the movement direction and average speed of the target person in this time period. According to the preset circular electronic fence area with a radius of 50 meters, the ray method is used to determine whether the current position of the target person is within the 10-meter warning zone of the electronic fence boundary. If so, the movement direction of the target person is obtained. If the angle between the movement direction of the target person and the normal direction of the electronic fence boundary is less than 90°, the dynamic adjustment mechanism of the electronic fence inclination is triggered. According to the distance d between the target person and the electronic fence boundary and the movement speed v, the angle value θ of the electronic fence inclination adjustment is dynamically calculated using the formula tanθ=v / (10-d), and the inclination angle of the electronic fence is changed in real time by controlling the servo motor. If the target person's movement direction changes and the angle with the normal direction of the electronic fence boundary is greater than 90° or the distance is greater than 10 meters, the electronic fence inclination adjustment is stopped and the servo motor is controlled to restore the original inclination setting of the electronic fence.

[0067] The system obtains the current position coordinates (x, y) of the person in real time through the positioning device, and compares it with the boundary coordinates (x1, y1), (x2, y2), (x3, y3), (x4, y4) of the electronic fence, and uses the Euclidean distance formula to calculate the distance d1, d2, d3, d4 between the person and each boundary of the electronic fence. The system presets a rule base containing 25 fuzzy control rules, takes the person distance di (i = 1, 2, 3, 4) as the input variable, and determines the adjustment amplitude θ of the electronic fence inclination angle through the Mamdani fuzzy reasoning method. Among them, the fuzzy control rule is in the form of "IF diis near THENθis large". The system uses the center of gravity defuzzification method to synthesize the reasoning results and obtain the specific inclination adjustment value. For example, when the person is 2 meters away from the boundary of the electronic fence, the inclination angle of the electronic fence is adjusted by 5 degrees. Based on the calculated inclination value, the system transmits the coordinates (xi', yi') (i = 1, 2, 3, 4) of each boundary point of the electronic fence and its corresponding inclination value θ to the actuator, and controls the actual inclination angle of the electronic fence through the servo motor. At the same time, the system continuously obtains the coordinates of the person's position through the infrared sensor with a sampling period of 1 second to determine whether the person has crossed the boundary of the electronic fence.

[0068] In order to obtain the historical data of false alarm rate and missed alarm rate of the electronic fence under different inclination settings, relevant data can be extracted from the log database of the electronic fence system, including inclination settings, alarm records, actual intrusion events, etc., to build a data set containing 1000 data. The data set is preprocessed, and the 3σ principle is used to eliminate outliers beyond the normal range, and the Kalman filter algorithm is used to filter the noise data to ensure that the data quality reaches more than 95%. Based on the preprocessed data set, the least squares method is used for regression analysis to establish a quadratic polynomial model between the inclination setting and the false alarm rate and missed alarm rate. The model fit goodness R 2 Reach more than 98. By solving the mathematical model, the false alarm rate curve and missed alarm rate curve under different inclination settings are obtained. The false alarm rate curve is y1=-05x 2 +8x+2, the underreporting rate curve is y2=1x 2 -5x+8, where x is the inclination value. Determine the intersection of the false alarm rate curve and the missed alarm rate curve, and by solving the equation group y1=y2, obtain the inclination value at the equilibrium point to be 60°. 60° is determined as the optimal inclination parameter of the electronic fence and applied to the electronic fence.

[0069] The above are only preferred specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for monitoring personnel safety behavior based on electronic fence technology, characterized in that: The following steps are involved: Acquire the terrain characteristics and working characteristic information of the target area through the pre-established electronic fence, set the initial inclination angle of the electronic fence based on the terrain characteristics and working characteristic information, and obtain the electronic fence adapted to the scene; The behavior trajectory of the target person is obtained through the electronic fence adapted to the scene, and the movement direction and speed of the target person are obtained by analyzing the behavior trajectory. When the target person is detected moving toward the electronic fence, the inclination adjustment value of the electronic fence is dynamically calculated using a fuzzy control algorithm, and the electronic fence is updated in real time according to the inclination adjustment value; After updating the electronic fence model, the latest location information of the target person is obtained. When the target person intersects the electronic fence and the inclination value of the electronic fence is greater than the warning threshold, it is determined that the target person has dangerous behavior.

2. The method for monitoring personnel safety behavior based on electronic fence technology according to claim 1 is characterized in that: The target personnel safety behavior monitoring method further includes dynamically optimizing the inclination value according to the monitoring data, and the dynamic optimization method includes: During the monitoring process, the false alarm rate and missed alarm rate data of the electronic fence are continuously obtained, and the false alarm rate and missed alarm rate data are statistically analyzed through the data analysis algorithm to obtain the false alarm rate and missed alarm rate curves under different inclination settings, and the inclination value at the balance point of the false alarm rate and missed alarm rate is selected as the optimal inclination parameter of the electronic fence; According to the optimal inclination parameters, the electronic fence model is updated and optimized, and according to the terrain characteristics and working characteristics of different regions, a mapping relationship between regional characteristics and inclination parameters is established to form an adaptive dynamic adjustment strategy for the inclination of the electronic fence.

3. The method for monitoring personnel safety behavior based on electronic fence technology according to claim 1 is characterized in that: The method for setting the initial inclination angle of the electronic fence includes: The initial inclination angle of the electronic fence is set by the support vector machine algorithm. When the target area is a large slope area, the inclination angle value of the electronic fence is increased; when the target area is a flat area, a unified inclination angle standard is used for adjustment; Through the decision tree algorithm, the terrain characteristics of the target area are judged to obtain the electronic fence inclination parameters suitable for the scene; According to the working characteristic information in the electronic fence model, the random forest algorithm is used to optimize and adjust the inclination parameters of the electronic fence.

4. The method for monitoring personnel safety behavior based on electronic fence technology according to claim 1 is characterized in that: Methods for obtaining the target person's movement direction and speed include: The location data of the target person is collected and stored in the database in chronological order to form the motion trajectory data of the target person; the motion trajectory data is smoothed using the Kalman filter algorithm, and the movement direction and average speed of the target person are determined by calculating the Euclidean distance and time difference between two adjacent position points based on the speed formula.

5. The method for monitoring personnel safety behavior based on electronic fence technology according to claim 1 is characterized in that: The method for real-time updating of the electronic fence includes: According to the preset electronic fence area range, when the current position of the target person is within the preset electronic fence area range, obtaining the movement direction of the target person; When the target person moves toward the electronic fence area, a dynamic adjustment mechanism of the electronic fence inclination is triggered; According to the distance between the target person and the boundary of the electronic fence and the movement speed, dynamically calculate the adjustment angle value of the electronic fence inclination, and control the inclination angle of the electronic fence in real time based on the adjustment angle value; When the moving direction of the target person changes or moves away from the electronic fence, the electronic fence inclination adjustment is stopped and the original inclination setting of the electronic fence is restored.

6. The method for monitoring personnel safety behavior based on electronic fence technology according to claim 5 is characterized in that: Methods for calculating the adjustment angle value of the electronic fence inclination include: A rule base of fuzzy control rules is preset, and the personnel distance is used as the input variable. The adjustment range of the electronic fence inclination is determined by the Mamdani fuzzy reasoning method, and the center of gravity defuzzification method is used to synthesize the reasoning results to obtain the specific inclination adjustment value.

7. The method for monitoring personnel safety behavior based on electronic fence technology according to claim 2 is characterized in that: Before statistical analysis of false alarm rate and missed alarm rate data, preprocessing is performed. The preprocessing methods include: The 3σ principle is used to eliminate outliers beyond the normal range, and the Kalman filter algorithm is used to filter the noise data to obtain the preprocessed data set.

8. The method for monitoring personnel safety behavior based on electronic fence technology according to claim 7 is characterized in that: Methods for obtaining false alarm rate and missed alarm rate curves under different tilt angle settings include: Based on the preprocessed data set, a regression analysis algorithm is used to perform modeling, wherein the regression analysis algorithm includes linear regression and polynomial regression; By means of the regression analysis algorithm, a first functional relationship expression between the inclination angle setting and the false alarm rate and a second functional relationship expression between the inclination angle setting and the missed alarm rate are obtained; A cross-validation method is used to evaluate the prediction performance of the first functional relationship expression and the second functional relationship expression under different inclination angle settings to obtain a regression model that meets preset requirements; the prediction performance includes the prediction accuracy of the false alarm rate and the prediction accuracy of the missed alarm rate; The regression model is solved to obtain false alarm rate and missed alarm rate curves under different inclination angle settings.

Images