CN-122022092-A - Park security multi-target path planning system and method

CN122022092ACN 122022092 ACN122022092 ACN 122022092ACN-122022092-A

Abstract

The application discloses a park security multi-target path planning system and method, and belongs to the technical field of security. The data layer builds a park security space-time database and realizes multidimensional data fusion, and data support is provided for security dispatching and risk assessment. The multi-objective path optimizing engine fuses coverage, time efficiency and personnel fatigue to construct a three-dimensional dynamic optimizing model, and realizes the dynamic adjustment of patrol paths by improving NSGA-III algorithm. The application layer provides a six-dimensional analysis grid layer and assists the command center and security inspection personnel in making decisions.

Inventors

SONG ZEMING
LIU XINWEI
LI ZIXUAN
ZHANG XIN
LIU XIAOYUAN
JIANG QI
ZENG FENG
WU SHUFAN

Assignees

中国铁塔股份有限公司
铁塔智联技术有限公司

Dates

Publication Date: 20260512
Application Date: 20260107

Claims (10)

1. The park security multi-target path planning system is characterized by comprising a data layer, an algorithm layer and an application layer; The data layer comprises a park space-time database and a multidimensional data fusion module, wherein the park space-time database comprises basic attributes and dynamic attributes, the basic attributes comprise inspection priority, surface object types and passing difficulty, the dynamic attributes comprise real-time event heat, historical scene weight and fatigue coefficient, and the multidimensional data fusion module comprises a historical data interface and a real-time data interface; the algorithm layer comprises a dynamic acting range model and a multi-objective optimization engine, wherein the dynamic acting range model is used for calculating a dynamic acting range based on the data of the data layer; the application layer comprises a six-dimensional analysis tool and a mobile terminal APP, wherein the six-dimensional analysis tool is used for generating a corresponding thermodynamic diagram according to requirements, and the mobile terminal APP is used for displaying real-time path navigation and fatigue early warning.
2. The campus security multi-objective path planning system of claim 1, wherein the construction of the data layer comprises: Dividing the park into square grid units with preset side lengths, wherein each grid unit stores corresponding basic attributes and dynamic attributes and is used for calculating basic features and dynamic state scenerization weights of quantized grids; And obtaining a weight matrix associated with the scene and the time period through historical event training, and determining the importance duty ratio of each factor under different scenes.
3. The campus security multi-objective path planning system of claim 1, wherein the dynamic range model calculates a dynamic range based on data of the data layer, comprising: The method comprises the steps of combining the real-time event density and the historical event frequency of a grid, quantifying the event activity degree of the grid, and calculating a thermodynamic coefficient; And dynamically adjusting the basic action radius according to the thermal coefficient based on a preset adjustment rule so as to adapt to areas with different risk levels.
4. A campus security multi-objective path planning system according to claim 3, wherein the preset adjustment rules are: if the thermal coefficient is greater than a first preset threshold value, wherein the action radius=the basic radius×the first preset coefficient, the first preset threshold value corresponds to a high risk area, and the first preset coefficient is smaller than 1; If the thermal coefficient is less than a second preset threshold (low risk region) that the action radius=the base radius×the second preset coefficient, the second preset threshold corresponds to the low risk region, and the second preset coefficient is greater than 1; if the thermal coefficient is not in the above range, the medium risk area is the action radius=the preset basic radius.
5. The campus security multi-objective path planning system of claim 1, wherein the multi-objective optimization engine generates an optimal path set comprising: constructing an objective function, wherein the objective function of path optimization comprises three core dimensions, and determining the path quality through comprehensive evaluation comprises the following steps: Objective function (path) = [ coverage, time efficiency, fatigue ]; the coverage rate represents the proportion of the path to cover the key grid, the time efficiency represents the reciprocal of the path completion time, and the fatigue degree represents the comprehensive consumption of the path; Initializing a path population to generate a batch of initial paths; The method comprises the steps of carrying out iterative optimization on an initial path, wherein the initial path comprises the steps of adaptively adjusting the weights of coverage rate, time efficiency and fatigue degree according to the current iterative progress and total iterative times, screening a high-quality path based on target weights, reserving a better solution, and extracting the pareto optimal front edge after iteration is finished, namely weighing an optimal path set of three targets.
6. The campus security multi-objective path planning system of claim 1, wherein the six-dimensional analysis tool generating the corresponding thermodynamic diagram according to demand comprises: If the layer type is real-time heating power, generating a thermodynamic diagram based on real-time event data; if the layer type is history contemporaneous, a thermodynamic diagram is generated based on the history contemporaneous event data.
7. The campus security multi-objective path planning system of claim 1, wherein the six-dimensional analysis tool further calculates a dead zone density, comprising: The number of grids with the statistical coverage rate lower than a third preset threshold is calculated according to the following formula: Blind area density = coverage < total number of meshes of preset coverage.
8. A method for planning a multi-objective path for park security, which is applicable to a multi-objective path planning system for park security as claimed in any one of claims 1-7, and is characterized by comprising: acquiring real-time event data through the data layer; Calculating a thermodynamic coefficient through the dynamic action range model of the algorithm layer, and calculating a dynamic action range based on the thermodynamic coefficient; and displaying the dynamic blind area diagram through an application layer, pushing the optimized patrol route by the mobile terminal APP, and carrying out risk prompt.
9. The electronic equipment is characterized by comprising a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory are communicated with each other through the communication bus; a memory for storing a computer program; A processor for executing the program stored in the memory to implement the campus security multi-objective path planning system of claim 8.
10. A computer readable storage medium storing a computer program, which when executed by a processor implements the campus security multi-objective path planning system of claim 8.

Description

Park security multi-target path planning system and method Technical Field The application belongs to the technical field of security protection, and particularly relates to a system and a method for planning a multi-target path of park security protection. Background With the deep advancement of smart city construction, the field of park security and protection is undergoing a deep revolution, and the intelligent development trend is more remarkable. The park is used as a region with high concentration of personnel, equipment and information, the requirement for safety guarantee is continuously improved, and the traditional security mode is difficult to meet the increasing safety management requirement. Under the background, the dynamic path planning technology becomes a core component of intelligent transformation of park security and protection, and the importance of the dynamic path planning technology is increasingly highlighted. According to MARKETSANDMARKETS reports, the 2025 global smart security market size is expected to reach 1840 billion dollars, with annual compound growth rates of dynamic path planning technology exceeding 15%. This data fully shows the great potential and rapid development situation of the technology in the market, and attracts the attention and investment of numerous enterprises and scientific research institutions. Currently, the existing technologies of park security dynamic path planning are mainly divided into three categories, and in practical application, some key defects exist in each category: 1. based on regular static scheduling, the patrol path of security personnel is planned according to preset rules, such as patrol is performed according to fixed time intervals and fixed routes. The method has the advantages of simplicity and easiness in operation, and can maintain basic security inspection work when the environment of a park is relatively stable and no emergency occurs. But the dynamic adaptability of the method is seriously insufficient when facing complex and changeable park environments and emergencies. It relies on a preset path and cannot respond to the thermal change of an event in real time. For example, in emergency situations such as fire breaks out in a campus, people gathering, etc., the statically planned patrol route cannot be adjusted in time, resulting in a response time delay. Experimental data show that when an emergency occurs in the static path, the response time delay reaches 40-60%, so that the emergency processing capacity of the security system is greatly reduced, and the expansion of the security accident can be possibly caused. 2. Single-objective optimization algorithm many existing systems plan paths with coverage or time efficiency as a single objective. Taking the maximum coverage area as an example, the algorithm calculates a patrol route which can cover as many areas as possible, but the influence of the fatigue degree of the personnel on the continuous combat ability can be ignored. The security personnel work with high intensity for a long time, and the fatigue degree can be continuously increased, so that the inspection efficiency and the judgment capability are affected. The actual measurement data show that the inspection efficiency is reduced by more than 35% in the fatigue state, which means that even if the coverage area is wide, the potential safety hazard can not be found in time due to poor personnel state. Likewise, if only the time efficiency is targeted, the route with the shortest patrol time is selected, so that some high risk areas may be omitted, and the safety of the park cannot be fully ensured. 3. The simple dynamic adjustment mechanism can adjust the patrol path to a certain extent according to real-time conditions, and if an abnormal condition occurs in a certain area, security personnel are temporarily arranged to go to the process. But such adjustments are often based on simple decisions, lacking comprehensive analysis of global conditions and predictions of future trends. It does not fully take into account the associations between different areas in the campus and the changes in overall security situation. For example, when a monitoring point in a campus monitors the aggregation of people, only nearby security personnel are arranged to go to and view, the overall security situation and the continuity of subsequent patrol tasks are not considered, the security force of other areas is possibly left empty, or the subsequent patrol routes are disordered, and an effective security closed loop cannot be formed. Meanwhile, the prior art also faces a plurality of technical development bottlenecks: 1. The modeling of the action range is rough, namely the existing system adopts a fixed radius (such as 50 meters) to divide the prevention area, and the mode does not establish the quantization relation between the thermodynamic coefficient and the action range. The risk degree and the event heat of different areas a