An anti-riot gun anti-loss management system, method and platform based on Beidou positioning

Abstract

The present invention discloses an anti-riot gun anti-loss management system, method and platform based on Beidou positioning, which relates to the technical field of anti-riot device management, and includes the following steps: Set a clear starting activity range for each anti-riot gun, that is, the starting electronic fence range. When the anti-riot gun exceeds this starting activity range, an alarm is automatically triggered to remind the management personnel; The geographical location of the anti-riot gun is obtained in real time through the Beidou positioning unit installed on the anti-riot gun. By obtaining the position information of the anti-riot gun in real time and combining the analysis of historical trajectory data, the present invention can timely identify potential trajectory anomalies. Even if the anti-riot gun is still within the preset range, the machine learning model is used to deeply analyze and predict the trajectory data. When an anomaly is detected, the electronic fence range is dynamically adjusted and the monitoring density is enhanced, so as to quickly respond to potential risks, prevent the abuse or loss of anti-riot guns, improve the safety management ability of anti-riot guns, and achieve more intelligent and accurate real-time monitoring and risk prevention.

Description

Technical Field

[0001] The present invention relates to the technical field of riot control device management, and particularly to an anti-loss management system, method and platform for riot guns based on Beidou positioning. Background Art

[0002] Riot control devices are non-lethal equipment used to respond to and control violent incidents and are widely used in law enforcement and security fields. Common riot control equipment includes riot shields, riot guns, tear gas grenades, rubber bullets and pepper sprays, etc., aiming to effectively control riots or violent behaviors while minimizing permanent injuries. For example, riot shields can be used to block projectiles and close-range attacks, riot guns fire rubber bullets or beanbag rounds to stop lawbreakers, and tear gas grenades and pepper sprays are used to disperse crowds. The system realizes comprehensive tracking and management of the equipment by installing Beidou positioning terminals on riot control devices, obtaining their precise positions in real time and transmitting the data to the monitoring center.

[0003] A riot gun is a non-lethal law enforcement equipment mainly used to control violent behaviors and riots. By firing rubber bullets, beanbag rounds, tear gas grenades or other non-lethal ammunitions, it can effectively contain the violent behaviors or destructive activities of lawbreakers. The role of a riot gun is to quickly incapacitate or stop the target through precise shooting, thus avoiding the escalation of confrontation and harming the innocent. It is usually used by security personnel, law enforcement agencies and riot control forces to deal with violent conflicts, riots or illegal assemblies and other situations. Compared with traditional firearms, the design of riot guns pays more attention to reducing personnel injuries and achieving the purpose of controlling the situation in a non-lethal way.

[0004] The prior art has the following deficiencies:

[0005] The prior art conducts anti-loss management by presetting the activity range of riot guns. When the riot gun exceeds this range, the system will automatically trigger an alarm and remind the management personnel. However, when the riot gun is still within the preset range but its historical track shows abnormalities, the system will not automatically issue an alarm. In this case, the riot gun may be stolen or moved to other locations without authorization. Once the violent conflict escalates, once the riot gun, as a key piece of equipment, is misused, it will not only be unable to effectively control the situation, but may even exacerbate the violence, resulting in a large number of casualties and even causing social panic.

[0006] The above information disclosed in the background art section is only used to enhance the understanding of the background of the present disclosure, so it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0007] The object of the present invention is to provide an anti-riot gun anti-loss management system, method and platform based on Beidou positioning. By obtaining the position information of the anti-riot gun in real time and combining with the analysis of historical trajectory data, potential trajectory anomalies can be identified in a timely manner. Even if the anti-riot gun is still within the preset range, a machine learning model is used to deeply analyze and predict the trajectory data. When an anomaly is detected, the range of the electronic fence is dynamically adjusted, and the monitoring density is enhanced, so as to quickly respond to potential risks and prevent the abuse or loss of the anti-riot gun. Through this solution, the safety management ability of the anti-riot gun is improved, more intelligent and accurate real-time monitoring and risk prevention are realized, the management efficiency and emergency response ability of the anti-riot device are greatly enhanced, and the risk of illegal use or loss of the device is effectively reduced, so as to solve the problems in the above background technology.

[0008] To achieve the above object, the present invention provides the following technical solutions: An anti-riot gun anti-loss management method based on Beidou positioning, comprising the following steps:

[0009] Set a clear starting activity range for each anti-riot gun, that is, the starting electronic fence range. When the anti-riot gun exceeds this starting activity range, an alarm is automatically triggered to remind the management personnel;

[0010] The geographical location of the anti-riot gun is obtained in real time through the Beidou positioning unit installed on the anti-riot gun, and the obtained position information is transmitted to the monitoring center through the wireless communication network for real-time position monitoring of the anti-riot gun;

[0011] Record the position information of the anti-riot gun in chronological order to form complete historical trajectory data;

[0012] Divide the historical trajectory into several sub-segments according to the number of time points and match them with the preset path. For the historical trajectory of each sub-segment, extract the features reflecting potential position trajectory anomalies of the anti-riot gun, and deeply analyze the extracted features through feature engineering technology to evaluate their relevance to the abuse or loss of the anti-riot gun;

[0013] Use the analyzed features as feature vectors and input them into a pre-trained and in-use machine learning model for prediction. Determine whether the current trajectory of the anti-riot gun is abnormal through the results output by the machine learning model;

[0014] When the machine learning model identifies that the historical trajectory of the anti-riot gun is abnormal, according to the analysis results, dynamically narrow the activity range of the anti-riot gun, that is, adjust the electronic fence range, and respond to potential risks in a timely manner; at the same time, enhance the monitoring density to ensure real-time tracking of the anti-riot gun.

[0015] Preferably, setting a clear starting activity range for each anti-riot gun specifically includes the following steps:

[0016] Determine the working area of the riot gun according to the actual usage scenarios and requirements;

[0017] Through a geographic information system or a map tool, define the specific scope of the electronic fence according to the boundaries of the area to ensure that the riot gun moves within the preset starting activity range.

[0018] Preferably, for the historical trajectory of each sub-segment, extract the features that reflect potential anomalies in the position trajectory of the riot gun. Among them, the extracted features include the degree of deviation of the riot gun's movement trajectory from its preset task path and the density between the movement trajectory points of the riot gun. Through feature engineering techniques, deeply analyze the degree of deviation of the riot gun's movement trajectory from its preset task path and the density between the movement trajectory points of the riot gun, and generate a frequent deviation from the predetermined path index and a movement path sparsity index respectively. Evaluate its relevance to the abuse or loss of the riot gun through the frequent deviation from the predetermined path index and the movement path sparsity index.

[0019] Preferably, for the historical trajectory of each sub-segment, the specific steps to deeply analyze the degree of deviation of the riot gun's movement trajectory from its preset task path through feature engineering techniques to generate a frequent deviation from the predetermined path index are as follows:

[0020] For each historical trajectory point and preset path point of each sub-segment, first calculate the Euclidean distance from the historical trajectory point to the preset path point, and represent the degree of deviation of the historical trajectory point from the preset path point through the Euclidean distance. Based on the degree of deviation, judge whether it belongs to the "off-path" state. Define that when the degree of deviation is greater than the deviation threshold, it is regarded as off-path;

[0021] Calculate the number of deviation times and the total deviation time of each trajectory segment, and consider the time interval between trajectory points, that is, the time difference between each trajectory point update, and calculate the frequent deviation frequency. The calculation expression is:

[0022] ,

[0023] Where: is an indicator function, indicating that when the degree of deviation is greater than the deviation threshold it takes the value of 1, indicating that the trajectory point is off-path, is the time interval between trajectory points, represents the frequent deviation frequency, i represents the index of the trajectory point, N represents the total number of trajectory points on each sub-segment;

[0024] Comprehensively consider the frequent deviation frequency, the degree of deviation of the historical trajectory point from the preset path point, and the total time of the entire trajectory, and calculate the frequent deviation from the predetermined path index. The calculation expression is:

[0025] ,

[0026] Wherein: Indicates frequent deviation from the predetermined path index, Is the total time of the entire trajectory, Is the adjustment coefficient to adjust the degree of deviation Of the influence degree.

[0027] Preferably, for the historical trajectory of each sub-segment, the specific steps of generating the motion path sparsity index by deeply analyzing the density between the motion trajectory points of the riot gun through feature engineering technology are as follows:

[0028] Evaluate the time distribution and point density of the historical trajectory of the riot gun. For the trajectory data of each sub-segment, analyze the time interval between the timestamps of each motion trajectory point and the previous and subsequent trajectory points, and define "trajectory time sparsity" to represent the sparsity degree of trajectory points within a certain period of time. The formula is as follows:

[0029] ,

[0030] Wherein: And Are respectively the i th and the i +1th trajectory point timestamps, And Are respectively the maximum and minimum timestamps of the trajectory of this sub-segment, Represents the trajectory time sparsity;

[0031] Introduce a weighting coefficient, consider the distribution characteristics of trajectory points in space and their time sparsity, and comprehensively generate a path sparsity index by combining time sparsity and the distribution change between trajectory points. The generated expression is:

[0032] ,

[0033] Wherein: Represents the motion path sparsity index, And Are respectively the i th trajectory point spatial coordinates, And Are respectively the i +1th trajectory point spatial coordinates, Is the weighted attenuation coefficient of spatial distribution, controlling the influence of the spatial distribution between trajectory points.

[0034] Preferably, the analyzed frequent deviation from the predetermined path index and the motion path sparsity index are used as feature vectors and input into a pre-trained and in-use machine learning model. Based on the machine learning model, a trajectory anomaly risk index is generated, and whether there is an anomaly in the current riot gun's trajectory is judged through the trajectory anomaly risk index.

[0035] Preferably, when predicting the trajectory anomaly of the current riot gun through a pre-trained and in-use machine learning model, the generated trajectory anomaly risk index is compared and analyzed with a pre-set risk threshold to judge whether there is an anomaly in the current riot gun's trajectory. The specific steps are as follows:

[0036] If the trajectory anomaly risk index is greater than the risk threshold, it is determined that there is an anomaly in the current riot gun's trajectory; if the trajectory anomaly risk index is less than or equal to the risk threshold, it is determined that there is no anomaly in the current riot gun's trajectory.

[0037] Preferably, when the machine learning model identifies that there is an anomaly in the historical trajectory of the riot gun, according to the analysis results, the specific steps to dynamically narrow the activity range of the riot gun and enhance the monitoring density are as follows:

[0038] After identifying the trajectory anomaly, the activity range of the riot gun is dynamically adjusted according to the trajectory anomaly risk index, that is, the electronic fence range is adjusted. The initial activity range of the riot gun is According to the trajectory anomaly risk index and the risk threshold calculate the adjusted activity range The calculation expression is:

[0039] ,

[0040] where: is the preset maximum trajectory anomaly risk index, used to control the amplitude of the reduction of the activity range;

[0041] After adjusting the electronic fence, the real-time tracking of the riot gun is strengthened by enhancing the monitoring density. The enhancement of the monitoring density is achieved by increasing the position update frequency. The default position update frequency of the riot gun is After the trajectory anomaly detection, the enhanced update frequency is calculated by the following formula:

[0042] ,

[0043] where, is the adjustment coefficient, used to control the increase amplitude of the update frequency.

[0044] An anti-riot gun anti-loss management system based on Beidou positioning, including an initial activity range setting module, a real-time position monitoring module, a historical trajectory recording module, a historical trajectory analysis and feature extraction module, a machine learning prediction and anomaly judgment module, and a dynamic response and monitoring adjustment module;

[0045] The initial activity range setting module sets a clear starting activity range for each anti-riot gun, that is, the starting electronic fence range. When the anti-riot gun exceeds this starting activity range, an alarm is automatically triggered to alert the management personnel;

[0046] The real-time position monitoring module obtains the geographical location of the anti-riot gun in real time through the Beidou positioning unit installed on the anti-riot gun, and transmits the obtained position information to the monitoring center through the wireless communication network for real-time position monitoring of the anti-riot gun;

[0047] The historical trajectory recording module records the position information of the anti-riot gun in chronological order to form complete historical trajectory data;

[0048] The historical trajectory analysis and feature extraction module equally divides the historical trajectory into several sub-segments according to the number of time points and matches them with the preset path. For each sub-segment of the historical trajectory, features reflecting potential position trajectory anomalies of the anti-riot gun are extracted, and the extracted features are deeply analyzed through feature engineering techniques to evaluate their relevance to the abuse or loss of the anti-riot gun;

[0049] The machine learning prediction and anomaly judgment module takes the analyzed features as feature vectors and inputs them into a pre-trained and in-use machine learning model for prediction, and judges whether there are anomalies in the current trajectory of the anti-riot gun through the results output by the machine learning model;

[0050] The dynamic response and monitoring adjustment module, when the machine learning model identifies that there are anomalies in the historical trajectory of the anti-riot gun, dynamically reduces the activity range of the anti-riot gun according to the analysis results, that is, adjusts the electronic fence range, and responds to potential risks in a timely manner; at the same time, enhances the monitoring density to ensure real-time tracking of the anti-riot gun.

[0051] In the above technical solution, the technical effects and advantages provided by the present invention:

[0052] By obtaining the position information of the riot gun in real time and combining the analysis of historical trajectory data, the present invention can promptly identify potential trajectory anomalies. Even if the riot gun is still within the preset range, a machine learning model is used to deeply analyze and predict the trajectory data. When an anomaly is detected, the range of the electronic fence is dynamically adjusted, and the monitoring density is enhanced, so as to quickly respond to potential risks and prevent the abuse or loss of the riot gun. Through this solution, the safety management ability of the riot gun is improved, more intelligent and accurate real-time monitoring and risk prevention are achieved, the management efficiency and emergency response ability of the riot protection device are greatly enhanced, and the risk of illegal use or loss of the device is effectively reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required in the embodiments. Obviously, the drawings in the following description are only some embodiments recorded in the present invention. For those of ordinary skill in the art, other drawings can also be obtained according to these drawings.

[0054] Figure 1 It is a method flow chart of a method for preventing the loss of a riot gun based on Beidou positioning according to the present invention.

[0055] Figure 2 It is a module schematic diagram of a system for preventing the loss of a riot gun based on Beidou positioning according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0056] Example embodiments will now be described more fully with reference to the accompanying drawings. However, the example embodiments can be implemented in various forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that this disclosure will be more comprehensive and complete, and will fully convey the concept of the example embodiments to those skilled in the art.

[0057] The present invention provides a method for preventing the loss of a riot gun based on Beidou positioning as shown in Figure 1 the following, including the following steps:

[0058] Set a clear starting activity range for each riot gun, that is, the starting electronic fence range. When the riot gun exceeds this starting activity range, an alarm is automatically triggered to remind the management personnel, ensuring that the riot gun is always within the authorized area and preventing unauthorized movement or potential abuse risks;

[0059] To set a clear starting activity range for each riot gun, it is first necessary to determine the working area of the riot gun according to the actual usage scenarios and requirements. For example, when used by law enforcement officers, the starting activity range can be set around specific law enforcement areas, patrol routes, or public event areas. Then, through a Geographic Information System (GIS) or mapping tools, the specific range of the electronic fence is defined based on the boundaries of the area to ensure that the riot gun moves within the preset starting activity range. If the riot gun exceeds the set range, the system will automatically trigger an alarm.

[0060] The geographical location of the riot gun is obtained in real time through a Beidou positioning unit installed on the riot gun, and the acquired location information is transmitted to the monitoring center through a wireless communication network (such as GSM, Wi-Fi, etc.) to achieve real-time location monitoring of the riot gun and provide data support for subsequent trajectory analysis;

[0061] The location information of the riot gun is recorded in chronological order to form complete historical trajectory data;

[0062] These historical trajectory data include each location update, stay time, and movement path of the riot gun. The function of this step is to provide basic data for subsequent trajectory analysis and anomaly detection.

[0063] The historical trajectory is equally divided into several sub-segments according to the number of time points and matched with the preset path. For each sub-segment of the historical trajectory, features reflecting potential location trajectory anomalies of the riot gun are extracted, and the extracted features are deeply analyzed through feature engineering techniques to evaluate their relevance to the abuse or loss of the riot gun;

[0064] Equally dividing the historical trajectory into several sub-segments according to the number of time points and matching it with the preset path means dividing the historical movement path of the riot gun (including the timestamp and coordinates of each location) into several small segments, each containing a certain number of time points or location information. Each sub-segment represents the movement of the riot gun within a certain period of time. Then, these sub-segments are compared with the preset path of the riot gun (i.e., the ideal route that the riot gun should follow) to check whether the actual trajectory within each sub-segment conforms to the predetermined route or activity range. The function of this process is that by refining the historical trajectory and comparing it with the preset path, abnormal behaviors of the riot gun can be identified more accurately.

[0065] For the historical trajectory of each sub-segment, features reflecting the potential position trajectory anomalies of the riot gun are extracted, where the extracted features include the degree of deviation of the riot gun's motion trajectory from its preset task path and the density between the riot gun's motion trajectory points. The degree of deviation of the riot gun's motion trajectory from its preset task path and the density between the riot gun's motion trajectory points are deeply analyzed through feature engineering technology, and the frequent deviation from the preset path index and the motion path sparsity index are generated respectively. The frequent deviation from the preset path index and the motion path sparsity index are used to evaluate their correlation with the abuse or loss of the riot gun.

[0066] The higher the deviation of the riot gun's historical trajectory from its preset mission path, it usually indicates that the riot gun is at a higher risk of being misused or lost. This is because riot guns should follow predetermined paths when performing tasks, such as patrol routes, guard areas, or specific mission locations. If the historical trajectory shows that the riot gun frequently deviates from the normal path, especially when there is no clear mission requirement, this may indicate that the riot gun has been illegally moved or abused. For example, if a riot gun suddenly enters an unauthorized area or continuously deviates from the preset working range, it may mean that the equipment has been stolen or used improperly. Therefore, trajectory deviation not only reflects the physical movement of the equipment, but can also be an obvious sign of abuse, theft, or unauthorized use, indicating that the security of the equipment has been threatened.

[0067] For each sub-segment’s historical trajectory, the specific steps for generating frequent deviation indicators from the preset path by deeply analyzing the deviation degree between the anti-riot gun’s trajectory and its preset task path through feature engineering technology are as follows:

[0068] For each sub-segment's historical trajectory point and preset path point (the corresponding point of the preset path), first calculate the Euclidean distance from the historical trajectory point to the preset path point. The Euclidean distance is used to represent the deviation between the historical trajectory point and the preset path point. Based on the deviation, determine whether it belongs to the "deviation from path" state. When the deviation is greater than the deviation threshold, it is considered to be off-path.

[0069] The purpose of this step is to identify possible abnormal behaviors in the trajectory, such as device theft or abuse, and to discover potential risks in a timely manner.

[0070] Calculate the number of deviations and total deviation time of each trajectory segment, and consider the time interval between trajectory points, that is, the time difference between each trajectory point update, to calculate the frequent deviation frequency. The calculation expression is:

[0071] ,

[0072] in: is an indicator function, indicating that when the deviation Greater than the deviation threshold When the value is 1, it means that the trajectory point deviates from the path. is the time interval between trajectory points, represents the frequent deviation frequency, i represents the index of the trajectory point, N represents the total number of trajectory points on each sub-segment;

[0073] By calculating the severity of each deviation path point in the riot gun's historical trajectory and combining the time difference between each pair of trajectory points, the frequency and degree of trajectory deviation are measured.

[0074] Finally, considering the frequent deviation frequency, the deviation degree between the historical trajectory points and the preset path points, and the total time of the entire trajectory, the frequent deviation from the predetermined path index is calculated. The calculation expression is:

[0075] ,

[0076] where: represents the frequent deviation from the predetermined path index, is the total time of the entire trajectory, is the adjustment coefficient to adjust the influence degree of the deviation degree .

[0077] From the frequent deviation from the predetermined path index, it can be seen that for the historical trajectory of each sub-segment, the larger the performance value of the frequent deviation from the predetermined path index generated by in-depth analysis of the deviation degree between the riot gun's movement trajectory and its preset task path through feature engineering techniques, the higher the deviation degree between the riot gun's historical trajectory and its preset task path, and thus the higher the risk of abuse or loss of the riot gun. When the movement trajectory of the riot gun frequently deviates from the preset path, it usually means that the device is being illegally used, moved, or has deviated from its original task area. For example, the riot gun may be stolen and transferred to an unauthorized area, or be maliciously operated and used for non-predetermined purposes. Frequent deviations usually result in abnormal trajectories. The frequent deviation from the predetermined path index generated by the system through analyzing the frequency and severity of these deviations can effectively evaluate whether there is a high risk of abuse or loss of the riot gun. If the frequent deviation from the predetermined path index is high, it indicates that there are significant abnormalities in the trajectory of the riot gun, thus increasing the possibility of abuse, loss, or illegal use.

[0078] The sparser the density between the historical trajectory points of the riot gun, the higher the correlation between the historical trajectory of the riot gun and abuse or loss, and the greater the risk. Generally, when the riot gun is performing tasks, its movement trajectory should be continuous and relatively stable. If the interval between trajectory points is too large, or there is no position information update for a long time in certain areas, it may mean that the riot gun has not been operating normally during this period, or its signal has been lost, resulting in incomplete trajectory records. Such a sparse trajectory may indicate that the riot gun has been illegally parked, disassembled, or transferred to an unmonitored location, or even stolen. Therefore, the sparsity of the trajectory density is a warning signal indicating that the riot gun may be being illegally used or has been lost, which requires high attention and further security checks.

[0079] For the historical trajectory of each sub-segment, the specific steps to generate the movement path sparsity index by deeply analyzing the density between the movement trajectory points of the riot gun through feature engineering techniques are as follows:

[0080] Evaluate the time distribution and point density of the historical trajectory of the riot gun. For the trajectory data of each sub-segment, analyze the time stamps of each movement trajectory point and the time intervals with the previous and subsequent trajectory points. By evaluating the time differences, deduce the movement frequency of the riot gun within this sub-segment, and define "trajectory time sparsity" to represent the sparsity degree of trajectory points within a certain period. The formula is as follows:

[0081] ,

[0082] Where: and are the time stamps of the i th and the i +1th trajectory points respectively, and are the maximum and minimum time stamps of the trajectory of this sub-segment respectively, represents the trajectory time sparsity;

[0083] This step evaluates the sparsity of the trajectory by calculating the ratio of the time interval between every two trajectory points to the total time period. The greater the time sparsity, the longer the time interval between trajectory points, indicating a higher sparsity degree of the movement trajectory, which may reflect that the device has stalled or lost contact at certain stages. Through the evaluation in terms of time distribution, it provides a basic framework for subsequent sparsity calculation.

[0084] Introduce a weighting coefficient, consider the distribution characteristics of trajectory points in space and their time sparsity, and comprehensively generate the path sparsity index by combining the time sparsity and the distribution changes between trajectory points (such as the relationship between the time sparsity within each sub-segment and the spatial distribution of trajectory points). The generated expression is:

[0085] ,

[0086] Wherein: represents the motion path sparsity index, and are respectively the spatial coordinates of the i th trajectory point, and are respectively the spatial coordinates of the i th + 1 trajectory point, is the weighted attenuation coefficient of the spatial distribution, controlling the influence of the spatial distribution between trajectory points;

[0087] By introducing the attenuation of the spatial distance and considering the actual distribution of the trajectory points, the farther the trajectory points are, the greater the influence. The finally calculated motion path sparsity index reflects the combined effect of the temporal sparsity and spatial distribution of the trajectory points. The larger the value, the higher the trajectory sparsity and the greater the risk of riot gun abuse or loss. The function of this step is to comprehensively consider the temporal and spatial factors and improve the sensitivity of the motion path sparsity index through the weighted attenuation function to ensure that abnormal trajectories can be identified in a timely manner.

[0088] It can be seen from the motion path sparsity index that for the historical trajectory of each sub - segment, the larger the performance value of the motion path sparsity index generated by deeply analyzing the density between the motion trajectory points of the riot gun through feature engineering technology, the higher the correlation between the historical trajectory of the riot gun and abuse or loss, and the greater the risk. Specifically, a larger motion path sparsity index means that there are obvious sparse trajectory points of the riot gun in some time periods, that is, in some regions or time periods, the riot gun does not move or update normally, which may lead to the loss of the device or illegal parking. This sparsity reflects that the riot gun may lose contact or be illegally moved within a period of time, thus reducing its traceability in the normal area and increasing the risk of abuse or loss. Therefore, the motion path sparsity index can be used as an effective index to help detect potential security hazards in a timely manner.

[0089] Taking the analyzed features as feature vectors, input them into a pre - trained and in - use machine learning model for prediction, and judge whether the current trajectory of the riot gun is abnormal through the results output by the machine learning model;

[0090] Taking the analyzed frequent deviation from the predetermined path index and motion path sparsity index as feature vectors and inputting them into a pre - trained and in - use machine learning model, generating a trajectory anomaly risk index based on the machine learning model, and judging whether the current trajectory of the riot gun is abnormal through the trajectory anomaly risk index.

[0091] A pre-trained and put-into-use machine learning model refers to a model that is trained, learned, and optimized through historical data, aiming to predict and judge whether there are abnormalities in the riot gun trajectory. The key feature of this model is that it has been extensively trained on a historical dataset to obtain effective parameters and weights, so as to be able to identify patterns and features related to trajectory abnormalities. The machine learning model learns how to judge the abnormality of a trajectory based on the characteristics of the trajectory (such as frequent deviation from the predetermined path index and the sparsity index of the movement path) by training on the labeled trajectory data. This model usually uses supervised learning methods. Specifically, it can be classification algorithms (such as support vector machines, random forests, or neural networks, etc.). By learning the normal and abnormal patterns of historical trajectories, a classifier is established that can automatically predict new riot gun trajectory data. The training process of the model includes extracting features from a large amount of historical trajectory data, marking which trajectories are normal and which are abnormal, and then adjusting the parameters in the model so that it can accurately judge the abnormality of the riot gun trajectory when new data is input. Finally, the trained and verified machine learning model is put into use, which can automatically process real-time data and trigger an alarm when the riot gun trajectory shows an abnormality.

[0092] The core advantages of the put-into-use machine learning model lie in its adaptability and efficiency. Through training, the model can identify complex patterns in the trajectory that are difficult to detect by simple rules. For example, traditional methods may only rely on preset trajectory deviation ranges or thresholds to judge whether a trajectory is normal, while the machine learning model can discover more complex abnormal patterns by analyzing historical trajectory data, such as subtle changes in the sparse distribution of trajectory points, frequent deviation from the normal path, especially the combination of multi-dimensional data (such as time, space, movement patterns, etc.). By inputting based on feature vectors (such as the frequent deviation from the predetermined path index and the sparsity index of the movement path), the machine learning model can comprehensively consider the influence of multiple factors, thereby making more accurate predictions and classifications. This intelligent judgment not only improves the accuracy of detection but also can timely discover potential abuse or loss risks, enhancing the overall security and emergency response capabilities of the riot gun management system.

[0093] The machine learning model is not specifically limited here. Any machine learning model that can realize comprehensive analysis of the frequent deviation from the predetermined path index and the sparsity index of the movement path to generate a trajectory abnormality risk index is acceptable. To implement the technical solution of the present invention, the present invention provides a specific implementation method; the expression for generating the trajectory abnormality risk index is: , in the formula, , are respectively the frequent deviation from the predetermined path index and the preset proportionality coefficient of the movement path sparsity index , and , are both greater than 0. The preset proportionality coefficient refers to the parameter used to weight the contributions of different input features to the final result in a machine learning model or algorithm. Specifically, in the formula, the preset proportionality coefficient is used to adjust the influence degree of each feature (such as the index of frequent deviation from the predetermined path and the movement path sparsity index ) on the final output (i.e., the trajectory anomaly risk index ). These coefficients determine the importance of each input feature and are set as constants in the model to balance the relative contributions of each feature. For example, represents the weight of the index of frequent deviation from the path , represents the weight of the movement path sparsity index , and usually . By adjusting these proportionality coefficients, the model can be made more sensitive or more balanced to a certain feature in different situations, thus affecting the final prediction effect.

[0094] It can be seen from the trajectory anomaly risk index that for the historical trajectory of each sub-segment, the larger the performance value of the index of frequent deviation from the predetermined path generated by in-depth analysis of the deviation degree between the movement trajectory of the riot gun and its preset task path through feature engineering techniques, and the larger the performance value of the movement path sparsity index generated by in-depth analysis of the density between the movement trajectory points of the riot gun through feature engineering techniques, that is, the larger the performance value of the trajectory anomaly risk index generated by predicting the trajectory anomaly of the current riot gun through a pre-trained and put-into-use machine learning model, the greater the risk of trajectory anomaly of the riot gun, and vice versa, the smaller the risk of trajectory anomaly of the riot gun.

[0095] Compare and analyze the trajectory anomaly risk index generated by predicting the trajectory anomaly of the current riot gun through a pre-trained and put-into-use machine learning model with a pre-set risk threshold to determine whether there is an anomaly in the trajectory of the current riot gun. The specific steps are as follows:

[0096] If the trajectory anomaly risk index is greater than the risk threshold, it is determined that there is an anomaly in the trajectory of the current riot gun; if the trajectory anomaly risk index is less than or equal to the risk threshold, it is determined that there is no anomaly in the trajectory of the current riot gun.

[0097] When the machine learning model identifies an anomaly in the historical trajectory of the riot gun, according to the analysis results, dynamically narrow the activity range of the riot gun, that is, adjust the range of the electronic fence, and respond to potential risks in a timely manner; at the same time, increase the monitoring density to ensure real-time tracking of the riot gun;

[0098] When the machine learning model identifies an anomaly in the historical trajectory of the riot gun, according to the analysis results, the specific steps to dynamically narrow the activity range of the riot gun and enhance the monitoring density are as follows:

[0099] After identifying the trajectory anomaly, dynamically adjust the activity range of the riot gun according to the trajectory anomaly risk index, that is, adjust the electronic fence range. Set the initial activity range of the riot gun as , according to the trajectory anomaly risk index and the risk threshold Calculate the adjusted activity range , and the calculation expression is:

[0100] ,

[0101] where: is the preset maximum trajectory anomaly risk index (the maximum value indicating a completely abnormal trajectory), used to control the reduction amplitude of the activity range;

[0102] In this step, by taking the difference between the trajectory anomaly risk index and the risk threshold as the adjustment factor, narrow the activity range of the riot gun. When the trajectory anomaly risk index is higher, it indicates that the trajectory anomaly is more serious, the reduction amplitude of the electronic fence is larger, and the activity range of the riot gun is smaller, thus restricting its activities in potential risk areas. The function of this step is to reduce the risk of the riot gun entering unsafe or unauthorized areas by dynamically adjusting the electronic fence.

[0103] After adjusting the electronic fence, strengthen the real-time tracking of the riot gun by enhancing the monitoring density. The enhancement of the monitoring density is achieved by increasing the position update frequency. Set the default position update frequency of the riot gun as , after the trajectory anomaly detection, the enhanced update frequency is calculated by the following formula:

[0104] ,

[0105] where, is the adjustment coefficient, used to control the increase amplitude of the update frequency;

[0106] When the trajectory anomaly risk index is relatively high, the enhanced update frequency Enable the position of the riot gun to be updated at a higher frequency, ensuring that the system can monitor its position changes in real time and respond promptly to abnormal situations. By enhancing the monitoring density, the system can track the position of the riot gun more precisely, ensuring that any potential danger or abuse can be quickly detected and handled. The role of this step is to improve the real-time tracking ability of the riot gun, ensuring a quick response and accurate judgment of trajectory anomalies.

[0107] By dynamically adjusting the activity range of the riot gun and enhancing the monitoring density, promptly respond to trajectory anomalies and reduce potential risks. By narrowing the electronic fence range, the activities of the riot gun are restricted to a smaller and safer area, reducing the possibility of entering risk areas or unauthorized areas. At the same time, enhancing the monitoring density ensures higher-precision real-time tracking of the dynamic changes of the riot gun by increasing the position update frequency, thus quickly detecting any abnormal behavior or potential threat. This comprehensive response mechanism helps to enhance the security of the riot gun, ensuring that rapid response measures can be taken in case of trajectory anomalies and preventing serious consequences such as abuse or loss.

[0108] Through the above-mentioned anti-loss management method for riot guns based on Beidou positioning, it can effectively solve the limitation of existing technologies that riot guns rely only on preset activity ranges for monitoring. This method can timely identify potential trajectory anomalies by obtaining the position information of the riot gun in real time and combining historical trajectory data analysis, even if the riot gun is still within the preset range. Using machine learning models to deeply analyze and predict the trajectory data, dynamically adjust the electronic fence range when anomalies are detected, and enhance the monitoring density, so as to quickly respond to potential risks and prevent the abuse or loss of riot guns. This solution improves the safety management ability of riot guns, realizes more intelligent and precise real-time monitoring and risk prevention, greatly enhances the management efficiency and emergency response ability of riot guns, and effectively reduces the risk of illegal use or loss of equipment.

[0109] The present invention provides a Figure 2 riot gun anti-loss management system based on Beidou positioning as shown, including an initial activity range setting module, a real-time position monitoring module, a historical trajectory recording module, a historical trajectory analysis and feature extraction module, a machine learning prediction and anomaly judgment module, and a dynamic response and monitoring adjustment module;

[0110] The initial activity range setting module sets a clear starting activity range for each riot gun, that is, the starting electronic fence range. When the riot gun exceeds this starting activity range, an alarm is automatically triggered to remind the management personnel;

[0111] The real-time position monitoring module obtains the geographical location of the riot gun in real time through the Beidou positioning unit installed on the riot gun, and transmits the obtained position information to the monitoring center through a wireless communication network for real-time position monitoring of the riot gun;

[0112] A historical trajectory recording module that records the position information of the riot gun in chronological order to form complete historical trajectory data;

[0113] A historical trajectory analysis and feature extraction module that equally divides the historical trajectory into several sub - segments according to the number of time points and matches them with a preset path. For each sub - segment of the historical trajectory, it extracts features reflecting potential abnormal position trajectories of the riot gun, and deeply analyzes the extracted features through feature engineering techniques to evaluate their relevance to the abuse or loss of the riot gun;

[0114] A machine learning prediction and anomaly judgment module that takes the analyzed features as feature vectors and inputs them into a pre - trained and in - use machine learning model for prediction, and judges whether the current trajectory of the riot gun is abnormal based on the results output by the machine learning model;

[0115] A dynamic response and monitoring adjustment module that, when the machine learning model identifies an abnormality in the historical trajectory of the riot gun, dynamically reduces the activity range of the riot gun according to the analysis results, that is, adjusts the electronic fence range, and responds to potential risks in a timely manner; at the same time, it enhances the monitoring density to ensure real - time tracking of the riot gun.

[0116] The riot gun anti - loss management method based on Beidou positioning provided by the embodiments of the present invention is implemented through the above - mentioned riot gun anti - loss management system based on Beidou positioning. The specific methods and processes of the riot gun anti - loss management system based on Beidou positioning are detailed in the embodiments of the above - mentioned riot gun anti - loss management method based on Beidou positioning, and will not be elaborated here.

[0117] The above formulas are all dimensionless and take their numerical values for calculation. The formula is obtained by collecting a large amount of data for software simulation to get a formula closest to the real situation. The preset parameters in the formula are set by those skilled in the art according to the actual situation.

[0118] The above - mentioned is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in the present application, and all should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claimed rights.

[0119] Only some exemplary embodiments of the present invention have been described by way of illustration above. Undoubtedly, for those of ordinary skill in the art, the described embodiments can be modified in various different ways without departing from the spirit and scope of the present invention. Therefore, the above - mentioned drawings and descriptions are illustrative in nature and should not be construed as limiting the protection scope of the claims of the present invention.

[0120] Only certain exemplary embodiments of the present invention have been described above by way of illustration. Without doubt, for those of ordinary skill in the art, the described embodiments can be modified in various different ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and description are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A method for preventing the loss of riot guns based on Beidou positioning, characterized in that: The following steps are involved: Set a clear starting range for each anti-riot gun, that is, the starting electronic fence range. When the anti-riot gun exceeds the starting range, an alarm will be automatically triggered to remind the management personnel; The Beidou positioning unit installed on the anti-riot gun obtains its geographic location in real time, and transmits the obtained location information to the monitoring center through the wireless communication network to monitor the real-time location of the anti-riot gun; The location information of the riot gun is recorded in chronological order to form complete historical trajectory data; The historical trajectory is divided into several sub-segments according to the number of time points, and matched with the preset path. For the historical trajectory of each sub-segment, the features reflecting the potential position trajectory anomalies of the riot gun are extracted. The extracted features are deeply analyzed through feature engineering technology to evaluate their correlation with the abuse or loss of the riot gun. The analyzed features are used as feature vectors and input into a pre-trained and used machine learning model for prediction. The results output by the machine learning model are used to determine whether there is an abnormality in the trajectory of the current riot gun. When the machine learning model identifies anomalies in the historical trajectory of the riot gun, it dynamically reduces the activity range of the riot gun based on the analysis results, that is, adjusts the range of the electronic fence to respond to potential risks in a timely manner; at the same time, it increases the monitoring density to ensure real-time tracking of the riot gun; For each sub-segment of the historical trajectory, features reflecting the potential position trajectory anomalies of the riot gun are extracted, where the extracted features include the degree of deviation of the riot gun's motion trajectory from its preset task path and the density between the riot gun's motion trajectory points. The degree of deviation of the riot gun's motion trajectory from its preset task path and the density between the riot gun's motion trajectory points are deeply analyzed through feature engineering technology, and the frequent deviation from the preset path index and the motion path sparsity index are generated respectively. The frequent deviation from the preset path index and the motion path sparsity index are used to evaluate their correlation with the abuse or loss of the riot gun; For each sub-segment’s historical trajectory, the specific steps for generating frequent deviation indicators from the preset path by deeply analyzing the deviation degree between the anti-riot gun’s trajectory and its preset task path through feature engineering technology are as follows: For each sub-segment's historical trajectory points and preset path points, the Euclidean distance from the historical trajectory points to the preset path points is first calculated. The Euclidean distance is used to represent the deviation between the historical trajectory points and the preset path points. Based on the deviation, it is determined whether it belongs to the "deviation from path" state. When the deviation is greater than the deviation threshold, it is considered to be off-path. Calculate the number of deviations and total deviation time of each trajectory segment, and consider the time interval between trajectory points, that is, the time difference between each trajectory point update, to calculate the frequent deviation frequency. The calculation expression is: ， in: is an indicator function, indicating that when the deviation Greater than the deviation threshold When the value is 1, it means that the trajectory point deviates from the path. is the time interval between trajectory points, Indicates frequent deviation frequency, i represents the index of the trajectory point, N represents the total number of trajectory points on each subsegment; Taking into account the frequent deviation frequency, the deviation between the historical trajectory points and the preset path points, and the total time of the entire trajectory, the frequent deviation index of the preset path is calculated. The calculation expression is: ， in: Indicates frequent deviations from the intended path indicator, is the total time of the entire trajectory, To adjust the deviation degree of impact.

2. According to claim 1, a Beidou positioning-based anti-riot gun loss prevention management method is characterized in that: Set a clear starting range for each riot gun. The specific steps are as follows: Determine the working area of ​​the riot gun based on actual usage scenarios and needs; Through geographic information systems or mapping tools, the specific range of the electronic fence is defined according to the boundaries of the area to ensure that the riot gun moves within the preset starting activity range.

3. The method for preventing loss of riot guns based on Beidou positioning according to claim 1 is characterized in that: The analyzed frequent deviation from the predetermined path index and motion path sparsity index are input as feature vectors into a pre-trained and put into use machine learning model. A trajectory anomaly risk index is generated based on the machine learning model. The trajectory anomaly risk index is used to determine whether there is an abnormality in the current trajectory of the riot control gun.

4. The method for preventing loss of riot guns based on Beidou positioning according to claim 3 is characterized in that: The trajectory anomaly risk index generated when predicting the current anti-riot gun's trajectory anomaly through the pre-trained and put into use machine learning model is compared and analyzed with the pre-set risk threshold to determine whether the current anti-riot gun's trajectory is abnormal. The specific steps are as follows: If the trajectory anomaly risk index is greater than the risk threshold, it is determined that the trajectory of the current riot gun is abnormal; If the trajectory anomaly risk index is less than or equal to the risk threshold, it is determined that there is no anomaly in the trajectory of the current riot gun.

5. The method for preventing loss of riot guns based on Beidou positioning according to claim 4 is characterized in that: When the machine learning model identifies anomalies in the historical trajectory of the riot gun, the specific steps for dynamically narrowing the activity range of the riot gun and increasing the monitoring density are as follows based on the analysis results: After identifying the abnormal trajectory, the activity range of the riot gun is dynamically adjusted according to the trajectory abnormality risk index, that is, the range of the electronic fence is adjusted. The initial activity range of the riot gun is set to , according to the trajectory anomaly risk index and risk thresholds Calculate adjusted range of motion , the calculation expression is: ， in: It is the preset maximum trajectory abnormality risk index, which is used to control the extent of the reduction of the activity range; After adjusting the electronic fence, the real-time tracking of the riot gun is strengthened by increasing the monitoring density. The enhancement of monitoring density is achieved by increasing the location update frequency. The default location update frequency of the riot gun is set to , after trajectory anomaly detection, the enhanced update frequency Calculated by the following formula: ， in, is the adjustment factor used to control the increase in the update frequency.

6. A Beidou-based anti-riot gun loss prevention management system, used to implement the Beidou-based anti-riot gun loss prevention management method described in any one of claims 1 to 5, characterized in that: It includes the initial activity range setting module, real-time location monitoring module, historical trajectory recording module, historical trajectory analysis and feature extraction module, machine learning prediction and abnormality judgment module, and dynamic response and monitoring adjustment module; The initial activity range setting module sets a clear starting activity range for each riot gun, that is, the starting electronic fence range. When the riot gun exceeds the starting activity range, an alarm is automatically triggered to remind the management personnel; The real-time location monitoring module obtains the geographic location of the riot gun in real time through the Beidou positioning unit installed on the riot gun, and transmits the obtained location information to the monitoring center through the wireless communication network to monitor the real-time location of the riot gun; The historical trajectory recording module records the location information of the riot gun in chronological order to form complete historical trajectory data; The historical trajectory analysis and feature extraction module divides the historical trajectory into several sub-segments according to the number of time points and matches them with the preset path. For each sub-segment of the historical trajectory, it extracts features that reflect the potential position trajectory anomalies of the riot gun. The extracted features are deeply analyzed through feature engineering technology to evaluate their correlation with the abuse or loss of the riot gun. The machine learning prediction and anomaly judgment module uses the analyzed features as feature vectors and inputs them into a pre-trained and used machine learning model for prediction. The output of the machine learning model is used to determine whether there is an anomaly in the trajectory of the current riot gun. Dynamic response and monitoring adjustment module: When the machine learning model identifies that there are anomalies in the historical trajectory of the riot gun, the activity range of the riot gun is dynamically reduced according to the analysis results, that is, the range of the electronic fence is adjusted to respond to potential risks in a timely manner; at the same time, the monitoring density is enhanced to ensure real-time tracking of the riot gun.

7. A Beidou-based anti-riot gun loss prevention management platform, comprising the Beidou-based anti-riot gun loss prevention management system according to claim 6.

Images