Risk assessment method for public security industry

Abstract

The invention discloses a risk assessment method for the public security industry, and belongs to the technical field of public security management. The method is cyclically executed in a preset time period, and comprises the following steps of: constructing a multi-dimensional dynamic risk map: gridding an activity site, converging multi-source real-time data associated with geographic grids, and calculating a basic risk value of each grid through a multi-factor risk contribution model; a risk exposure degree deduction step: acquiring a real-time police force position, calculating a police force response delay index of each grid, and performing coupling calculation to obtain a risk exposure degree index of each grid; and a police scheduling suggestion generation step of generating a police scheduling suggestion aiming at minimizing the total risk exposure degree of the whole field through a resource optimization allocation algorithm based on the risk exposure degree index. The technical problem that in the prior art, risk assessment and police dispatching are separated from each other, and on-site dynamic changes cannot be handled is solved.

Description

Technical Field

[0001] This invention belongs to the field of public security management technology, and specifically relates to a risk assessment method for the public security industry. Background Technology

[0002] In security operations for large public events, such as music festivals or sporting events, existing technologies typically rely on static security plans developed before the event begins. These static plans divide the venue into several zones and assign fixed police posts and patrol routes to each zone. However, on-site risks during events are highly dynamic. For example, crowd density in specific areas may surge in a short period, or negative sentiment may accumulate on social media due to unforeseen events. These dynamic risk factors cannot be perceived in real time by static plans. Simultaneously, security resources, i.e., police manpower, are limited. Current technologies rely primarily on commanders' personal experience and delayed voice or text reports for on-site police deployment, lacking a technological means to correlate and optimize the matching of dynamically changing risks with limited police resources in real time. This disconnect between risk assessment and resource allocation, both in data and logic, prevents the dynamic and optimal deployment of police forces based on the real-time distribution of risks, resulting in a mismatch of security resources in time and space, and hindering effective responses to unforeseen circumstances. Summary of the Invention

[0003] This invention provides a risk assessment method for the public security industry to address the technical shortcomings of the prior art, which is that risk assessment and resource allocation are separated and cannot effectively respond to dynamic changes on site.

[0004] In view of the above problems, the present invention provides a risk assessment method for the public security industry, applied to security scenarios for large-scale public events. The method is executed cyclically within a preset time period and includes the following steps: The steps for constructing a multidimensional dynamic risk map are as follows: the activity site map is gridded to form multiple geographic grids; multi-source real-time data associated with the multiple geographic grids are aggregated; and based on the multi-source real-time data, a basic risk value is calculated and assigned to each geographic grid through a multi-factor risk contribution model to generate the multidimensional dynamic risk map. The steps for extrapolating risk exposure are as follows: Obtain real-time police location data for the current period; Based on the real-time police location data and the locations of the multiple geographic grids, calculate the police response delay index for each geographic grid; and derive the risk exposure index by coupling the basic risk value of the geographic grid with the police response delay index. Steps for generating police force dispatch recommendations: Based on the risk exposure index and combined with the available mobile police resources, a police force dispatch recommendation is generated through a resource optimization allocation algorithm, wherein the computational objective of the resource optimization allocation algorithm is to minimize the total risk exposure of the activity site.

[0005] The technical solution provided in this application establishes a closed-loop system, from multi-source real-time data perception and dynamic risk quantification assessment to optimized security resource scheduling, by cyclically executing the above steps within a time period. This method can dynamically calculate the risk exposure of each area based on real-time changes in risk factors such as crowd density, online sentiment, and device alarms at the event site, and further automatically generate data-driven police deployment recommendations aimed at minimizing overall risk. This allows the deployment of security resources to adaptively adjust to the dynamic distribution of risks, thereby improving the coverage efficiency and response capability for sudden risks at large public events without increasing the total number of security resources. Attached Figure Description

[0006] Figure 1 This is a flowchart illustrating a risk assessment method for the public security industry provided in an embodiment of the present invention. Detailed Implementation

[0007] The above technical solutions will now be described in detail with reference to the accompanying drawings and specific embodiments to provide a better understanding of them. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. It should be understood that the present invention is not limited to the exemplary embodiments used only to explain the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention. Furthermore, it should be noted that, for ease of description, only the parts related to the present invention are shown in the drawings, not all of them.

[0008] Please see Figure 1 A risk assessment method for the public security industry, applied to security scenarios for large-scale public events, is described. The method is executed cyclically within a preset time period and includes the following steps: The steps for constructing a multidimensional dynamic risk map are as follows: the activity site map is gridded to form multiple geographic grids; multi-source real-time data associated with the multiple geographic grids are aggregated; and based on the multi-source real-time data, a basic risk value is calculated and assigned to each geographic grid through a multi-factor risk contribution model to generate the multidimensional dynamic risk map. The steps for extrapolating risk exposure are as follows: Obtain real-time police location data for the current period; Based on the real-time police location data and the locations of the multiple geographic grids, calculate the police response delay index for each geographic grid; and derive the risk exposure index by coupling the basic risk value of the geographic grid with the police response delay index. Steps for generating police force dispatch recommendations: Based on the risk exposure index and combined with the available mobile police resources, a police force dispatch recommendation is generated through a resource optimization allocation algorithm, wherein the computational objective of the resource optimization allocation algorithm is to minimize the total risk exposure of the activity site.

[0009] This invention provides a risk assessment method for the public security industry, which, in one embodiment, is implemented through a smart security decision-making system deployed in a security command center. The core of this smart security decision-making system is a risk assessment and scheduling engine, which automatically executes the risk assessment and scheduling process once every preset time period, for example, every 300 seconds.

[0010] A single-cycle workflow is initiated by the risk assessment and scheduling engine, which performs the construction and updating of a dynamic risk map. The risk assessment and scheduling engine first performs gridding processing on the event site map. It loads pre-stored Geographic Information System (GIS) data describing the physical layout of the event site and invokes spatial partitioning algorithms, such as quadtree partitioning, to divide the two-dimensional map of the event site into multiple equally sized geographic grid cells, such as 10-meter by 10-meter square grids, assigning a unique identifier to each generated geographic grid cell.

[0011] After completing the map gridding process, the risk assessment and scheduling engine then aggregates multi-source heterogeneous real-time data. Through its data interface module, at the beginning of each time period, the risk assessment and scheduling engine initiates data requests in parallel to multiple independent external data sources. In one embodiment, the risk assessment and scheduling engine requests real-time crowd density values ​​within each geographic grid cell from the video analytics server via a Hypertext Transfer Protocol (HTTP) interface; it also receives social media negative sentiment indices with geographic location tags from the public opinion monitoring cloud platform via a message queue telemetry transmission protocol subscription mechanism; and furthermore, it polls IoT sensors deployed on key physical facilities within the site via the Modbus transmission control protocol to obtain status information of equipment anomaly alarms. Upon receiving this data, the risk assessment and scheduling engine parses and cleans it, associates all data with the corresponding geographic grid cell identifiers, and stores the data in an in-memory database.

[0012] After aggregating real-time data associated with geographic grids, the risk assessment and scheduling engine proceeds to calculate the basic risk values ​​for each grid. To eliminate differences in units and numerical ranges between data from different sources, the engine first performs normalization on each piece of real-time data before weighted calculation. In one embodiment, this normalization method uses min-max normalization to map different ranges of values, such as population density values ​​from 0 to 1000 and negative sentiment indices from 0 to 1, to a dimensionless interval of 0 to 1. After normalization, the engine iterates through each geographic grid cell in the in-memory database and applies a pre-defined multi-factor risk contribution model to calculate the various real-time data associated with that grid cell. In a specific embodiment, this model is a linear weighted summation formula: Here, α, β, and γ are preset weighting coefficients. In one embodiment, these weighting coefficients are empirically set by security experts through a human-computer interaction interface, based on the risk emphasis of different event types. For example, for music festivals, the values ​​of α and β can be increased to focus on crowd and emotional risks, while for equipment exhibitions, the value of γ can be increased to focus on facility security risks. In another embodiment, these weighting coefficients can also be automatically optimized and determined through regression training using historical risk event data from similar events via a machine learning module. After the risk assessment and scheduling engine calculates the basic risk value for each geographic grid unit, it completes the construction of a multi-dimensional dynamic risk map reflecting the risk distribution within the current period.

[0013] After constructing the multi-dimensional dynamic risk map, the risk assessment and scheduling engine then performs the task of extrapolating and quantifying risk exposure. The engine first collects and maps real-time police location data. This is achieved by accessing a real-time updated police status database via an internal data bus. This database stores the timestamped latitude and longitude coordinates reported by each on-site security personnel via their handheld terminals through the GPS module. After obtaining the latest coordinates of all security personnel, the engine calls a coordinate transformation function to convert these coordinates into two-dimensional planar coordinates corresponding to the aforementioned geographic grid units, and records the identifier of each security personnel's current geographic grid unit.

[0014] After collecting and mapping all police locations, the risk assessment and dispatch engine proceeds to calculate the police response delay index for each grid cell. The engine iterates through each geographic grid cell in the multi-dimensional dynamic risk map. For each target geographic grid cell, the engine performs the following calculation steps: First, it retrieves the location of the nearest security personnel to the target grid cell; then, it loads pre-stored path network map data containing information on all walkable paths within the site; finally, on this path network map, starting from the nearest security personnel's location and ending at the center point of the target grid cell, it runs a shortest path algorithm, such as the A* (A-Star) algorithm, to calculate the length of the shortest patrol path required from the start point to the end point. The engine divides the calculated path length by a preset average movement speed of security personnel (e.g., 1.5 meters per second) to obtain a time value in seconds, which is defined as the police response delay index for that target geographic grid cell.

[0015] After calculating the police response delay index for each geographic grid cell, the risk assessment and dispatch engine finally generates the risk exposure index for each grid cell. The engine again traverses each geographic grid cell in the multidimensional dynamic risk map, and for each grid cell, substitutes the basic risk value and police response delay index calculated in the preceding steps into a preset coupling formula to generate the final risk exposure index.

[0016] It should be noted that this risk exposure index is designed with a response-sensitive technical characteristic. Response sensitivity is defined as follows: when the base risk value of a geographic grid is fixed, its final risk exposure index will increase disproportionately with the increase of the police response delay index.

[0017] To achieve the above characteristics, in a specific embodiment, the coupling calculation formula used for the i-th geographic grid is: ,in, The final risk exposure value generated for the i-th geographic grid is... The base risk value for the i-th geographic grid. Let i be the police response delay index for the i-th geographic grid. For example, a reference time used to achieve dimensional normalization. It can be set to 60 seconds. The time will be measured in seconds. Divide by This yields a dimensionless relative delay ratio to satisfy the input requirements of the logarithmic function. This is a dimensionless delay impact coefficient used to adjust the degree to which response delay amplifies the underlying risk. For example, It can be set to 0.5. According to the formula, when the response latency exponent... When it is 0, the risk exposure is Equal to the basic risk value This indicates that the risk is fully covered by police force, but its inherent risk still exists. As the response latency index increases relative to the baseline time, the risk exposure will increase disproportionately with a logarithmic curve. The engine associates the calculated risk exposure value with the corresponding geographic grid cell identifier and updates it to the in-memory database to complete a comprehensive quantification of the risk exposure of the entire event site.

[0018] After quantifying the risk exposure of all geographic grid cells, the risk assessment and scheduling engine finally executes the task of generating and outputting police force dispatch recommendations. The risk assessment and scheduling engine first identifies risk exposure hotspot grids. This is achieved by sorting the risk exposure values ​​of all geographic grid cells stored in the in-memory database in descending order, and extracting all geographic grid cells with risk exposure values ​​higher than a preset threshold (e.g., values ​​greater than 8.0). The set of these extracted geographic grid cells is defined as the hotspot grid set for the current period.

[0019] After identifying the hotspot grid set, the risk assessment and scheduling engine then executes a scheduling scheme optimization solution based on the objective of minimizing total risk. The engine first retrieves a list of currently "mobile" and available security personnel or teams from the police force status database. Then, the engine initiates a resource optimization allocation algorithm. In one specific embodiment, this algorithm is a greedy algorithm, and its execution logic is as follows: In each iteration, the algorithm selects the hotspot grid with the highest current risk exposure value from the unassigned hotspot grids as the target; simultaneously, it selects the mobile police force with the shortest dispatch time to the target from the unassigned mobile police forces. The algorithm virtually assigns this mobile police force to the target hotspot grid and recalculates the expected risk exposure of the hotspot grid based on the new, shortened response delay. The algorithm repeats this iterative process until all mobile police forces have been assigned, or all hotspot grids have been covered.

[0020] After optimizing the dispatch plan, the risk assessment and dispatch engine finally generates and issues structured dispatch instructions. The engine transforms the police force allocation plan determined in the previous step into one or more sets of structured dispatch instructions. In one embodiment, each set of dispatch instructions is a data object containing the following fields: {“Dispatch Unit ID”: “Mobile Group 2”, “Target Grid ID”: “A034”, “Instruction Type”: “Go to Support”, “Suggested Route”: [Coordinate Sequence], “Expected Arrival Time”: “3 minutes”}. The risk assessment and dispatch engine pushes these generated structured dispatch instructions to the front-end interface of the geographic information system deployed on the command center's large screen for visualization via an internal application programming interface (API). Simultaneously, it converts the instruction content into voice information via a text-to-speech module, which is then automatically broadcast to the communication equipment of the corresponding security unit via a wireless communication gateway.

[0021] The embodiments described above are only some embodiments of the present invention, and not all embodiments. It should be noted that the present invention is not limited to the embodiments described above. Any modifications, equivalent substitutions, variations, or improvements made by those skilled in the art based on the above embodiments without creative effort, within the scope of the concept and principles of the present invention, should fall within the protection scope claimed by the claims of the present invention.

Claims

1. A risk assessment method for the public security industry, applied to security scenarios for large-scale public events, characterized in that, The method is executed cyclically within a preset time period, including the following steps: The steps for constructing a multidimensional dynamic risk map are as follows: the activity site map is gridded to form multiple geographic grids; multi-source real-time data associated with the multiple geographic grids are aggregated; and based on the multi-source real-time data, a basic risk value is calculated and assigned to each geographic grid through a multi-factor risk contribution model to generate the multidimensional dynamic risk map. The steps for extrapolating risk exposure are as follows: Obtain real-time police location data for the current period; Based on the real-time police location data and the locations of the multiple geographic grids, calculate the police response delay index for each geographic grid; and derive the risk exposure index by coupling the basic risk value of the geographic grid with the police response delay index. Steps for generating police force dispatch recommendations: Based on the risk exposure index and combined with the available mobile police resources, a police force dispatch recommendation is generated through a resource optimization allocation algorithm, wherein the computational objective of the resource optimization allocation algorithm is to minimize the total risk exposure of the activity site.

2. The risk assessment method for the public security industry according to claim 1, characterized in that, The multi-source real-time data includes at least one of the following: real-time crowd density data from a video analytics system; negative sentiment index from a social media monitoring system; and device anomaly alarms from IoT sensors.

3. The risk assessment method for the public security industry according to claim 1, characterized in that, The multi-factor risk contribution model calculates the basic risk value by assigning different weights to the multi-source real-time data from different sources and performing weighted aggregation.

4. The risk assessment method for the public security industry according to claim 1, characterized in that, The calculation of the police response delay index includes determining the shortest patrol path time required to reach the geographic grid from the nearest police location.

5. The risk assessment method for the public security industry according to claim 1, characterized in that, The risk exposure index has a response sensitivity, which means that when the basic risk value is the same, the risk exposure index increases disproportionately with the increase of the police response delay index.

6. The risk assessment method for the public security industry according to claim 5, characterized in that, The risk exposure is calculated by multiplying the base risk value by a correction factor, which includes a base value of 1 and a delay gain term that increases logarithmically with the increase of the police response delay index.

7. The risk assessment method for the public security industry according to claim 6, characterized in that, For the i-th geographic grid, the risk exposure is calculated using the following formula: in, The final risk exposure value generated for the i-th geographic grid is... The base risk value for the i-th geographic grid. Let i be the police response delay index for the i-th geographic grid. As a reference time for achieving dimensional normalization, It is a dimensionless delay effect coefficient used to adjust the degree to which response delay amplifies the underlying risk.

8. The risk assessment method for the public security industry according to claim 1, characterized in that, The step of generating police force dispatch suggestions specifically includes: Among the multiple geographic grids, hotspot grids with risk exposure levels higher than a preset threshold are identified; The resource optimization allocation algorithm is then activated to allocate the mobile police force resources to the hotspot grid, so as to minimize the total risk exposure of the entire activity area after allocation.

Images