A three-dimensional security linkage dispatching method and system based on an internet of things

Abstract

The application relates to the technical field of Internet of Things security and protection, and particularly discloses a three-dimensional security and protection linkage scheduling method and system based on the Internet of Things, which aims at the problems of resource mismatch and high-priority event disposal lag in a multi-event concurrent scene, collects security and protection event and idle resource data first, calculates the basic fitness of resources to events, obtains the weighted fitness by combining with the comprehensive priority of events, introduces the global benefit of resource allocation quantified by opportunity cost and solves the relative benefit, and allocates the optimal resource with positive relative benefit to events according to the comprehensive priority, triggers the dynamic prediction and guidance vector generation linked with the relative benefit when a mobile tracking event is allocated with a UAV and ground personnel at the same time, and realizes automatic air-ground cooperation. The application can realize global optimal configuration of security and protection resources, improve high-emergency event disposal efficiency, replace manual cooperation to reduce communication errors, improve the success rate of mobile target interception, and adapt to real-time low-delay scheduling requirements in security and protection scenes.

Description

Technical Field

[0001] This invention relates to the field of Internet of Things (IoT) security technology, and in particular to a three-dimensional security linkage scheduling method and system based on IoT. Background Technology

[0002] Currently, key security locations such as chemical industrial parks and major transportation hubs have widely deployed three-dimensional security systems consisting of patrol drones, ground security personnel, intelligent inspection robots, and IoT sensing devices to handle sudden security incidents such as intrusions, leaks, and fires.

[0003] Existing security dispatch systems generally employ a greedy resource allocation rule that combines fixed event priority with distance priority. Air-to-ground collaborative tracking between drones and ground personnel relies entirely on human voice communication, with guidance parameters set to fixed values. In real-world scenarios with multiple concurrent events, the current dispatch method fails to quantify the opportunity cost of resource allocation and cannot comprehensively assess the overall benefits. This easily leads to high-suitability resources like drones being occupied by lower-priority events that are closer, resulting in delays in handling high-priority events such as intrusions and fires due to a lack of suitable resources. Furthermore, fixed guidance parameters cannot match differences in task suitability, and human voice coordination suffers from communication delays and information transmission errors. This results in low efficiency and insufficient interception success rate for air-to-ground collaborative tracking of moving targets. Therefore, existing three-dimensional security dispatch systems cannot achieve globally optimal resource allocation and fail to meet the practical needs of rapid and accurate handling of security incidents in key locations.

[0004] Therefore, there is an urgent need for a three-dimensional security linkage scheduling method and system based on the Internet of Things to solve the above problems. Summary of the Invention

[0005] The purpose of this invention is to provide a three-dimensional security linkage scheduling method based on the Internet of Things, comprising the following steps: Obtain a list of incomplete events and available resources within the security area. The incomplete events include event type, event location, basic priority, and dynamic priority. The available resources include resource type, resource location, and remaining energy. For each idle resource and each incomplete event, calculate the basic fitness, which is a weighted sum of resource-event matching degree, distance proximity, and remaining energy; The base fitness is weighted according to the overall priority of unfinished events to obtain a weighted fitness, and the relative benefit of each idle resource relative to each unfinished event is calculated based on the weighted fitness. Unfinished events are processed in descending order of their overall priority. For the current event, an idle resource that maximizes its relative benefit and is greater than zero is selected for allocation. If there is no resource with a relative benefit greater than zero, the event is skipped. When the same event is assigned to both a drone and ground personnel and the event type is mobile tracking, collaborative guidance is triggered. The target's moving speed is estimated based on the drone's observations, and the target's future position is predicted using a prediction duration that is positively correlated with the drone's relative benefit from the event. A guidance vector is generated from the ground personnel's current position to the predicted position and sent to the ground personnel's terminal.

[0006] Furthermore, this invention also discloses a three-dimensional security linkage dispatch system based on the Internet of Things, comprising: The acquisition module is used to acquire a list of incomplete events and idle resources within the security area. The incomplete events include event type, event location, basic priority, and dynamic priority. The idle resources include resource type, resource location, and remaining energy. The first calculation module is used to calculate the basic fitness for each idle resource and each unfinished event. The basic fitness is a weighted sum of resource-event matching degree, distance proximity, and remaining energy. The second calculation module is used to weight the basic fitness according to the comprehensive priority of the unfinished events to obtain a weighted fitness, and to calculate the relative benefit of each idle resource relative to each unfinished event based on the weighted fitness. The processing module is used to process unfinished events in descending order of their overall priority. It selects the idle resources that maximize the relative benefit of the current event and have a relative benefit greater than zero for allocation. If there are no resources with a relative benefit greater than zero, the event is skipped. The sending module is used to trigger collaborative guidance when the same event is assigned to both a drone and a ground personnel and the event type is mobile tracking. It estimates the target's moving speed based on the drone's observations and predicts the target's future position with a prediction duration that is positively correlated with the drone's relative benefit from the event. It then generates a guidance vector from the ground personnel's current position to the predicted position and sends it to the ground personnel's terminal.

[0007] Furthermore, the processing module includes: The sorting unit is used to sort unfinished events in descending order of overall priority, and in order of occurrence when priorities are the same. The optimization unit is used to find the resource that maximizes the relative benefit of each sorted event from the available resources and record the maximum relative benefit. The first processing unit is used to allocate the resource to the current event when the maximum relative benefit is greater than zero, generate a scheduling instruction containing the resource identifier, event location and action mode, and remove the resource from the idle list. The second processing unit is used to temporarily not allocate the maximum relative return if it is not greater than zero, and to skip the event and continue processing the next event.

[0008] This application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described three-dimensional security linkage scheduling method based on the Internet of Things.

[0009] This application also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the above-described IoT-based three-dimensional security linkage scheduling method.

[0010] The beneficial effects of this application are as follows: This invention breaks down data silos between systems by constructing a unified real-time data warehouse, enabling seamless real-time data sharing and providing accurate data support for subsequent scheduling. Machine learning predicts palletizing table load and generates priority task sequences based on sorting plans, avoiding situations where vehicles are needed but not available, or vehicles are available but not loaded, thus improving task foresight. The scheduling algorithm considers the structure of the automated freight station, optimizing task allocation and paths through multiple objectives, reducing the total empty driving distance of TV vehicles and task delays, and improving resource utilization. Real-time instruction distribution and status feedback enable controllable task execution and visualized traffic within the station, ensuring effective scheduling. Closed-loop control dynamically adjusts parameters, triggering rescheduling for significant deviations, continuously adapting to dynamic changes in the freight station, ultimately improving TV vehicle scheduling efficiency and optimizing key indicators such as empty driving rate and on-time performance. Attached Figure Description

[0011] Figure 1 This is a schematic diagram of a method flow proposed in an embodiment of this application.

[0012] Figure 2 This is a schematic diagram of the system structure proposed in an embodiment of the present invention.

[0013] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0014] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0015] like Figure 1 As shown, this application provides a three-dimensional security linkage scheduling method based on the Internet of Things, including the following steps: S1: Real-time acquisition of the list of incomplete events and the list of idle resources within the security area; the events include event type, event location, basic priority, and dynamic priority; the resources include resource type, resource location, and remaining energy; S2: For each pair of resources and events, calculate the basic fitness, which is a weighted sum of resource-event matching degree, distance proximity, and remaining energy; S3: The basic fitness is weighted according to the comprehensive priority of the events to obtain the weighted fitness. Then, the relative benefit of each resource relative to each event is calculated. The relative benefit is equal to the weighted fitness of the resource for the current event minus the maximum weighted fitness that the resource can obtain for all other events. S4: Process events in descending order of overall priority. Select the idle resource that maximizes the relative benefit of the current event and has a relative benefit greater than 0 for allocation. If there is no resource with a relative benefit greater than 0, skip the event. S5: When the same event is assigned to both a drone and ground personnel and the event type is mobile tracking, trigger collaborative guidance: estimate the target's moving speed based on the drone's continuous observations, predict the target's future position with a prediction duration that is positively correlated with the drone's relative benefit from the event, generate a guidance vector from the ground personnel's current position to the predicted position, and send the guidance vector to the ground personnel's terminal.

[0016] As described in steps S1-S5 above, in multi-event concurrent, three-dimensional security scenarios such as industrial parks and large transportation hubs, existing scheduling methods have some technical shortcomings. Specifically, existing scheduling methods only employ a single greedy allocation rule based on fixed event priorities or distance priority, failing to quantify the opportunity cost of resource allocation. When high-priority events such as intrusions and fires occur simultaneously with low-priority events such as leaks, highly adaptable resources like drones are easily occupied by closer, low-priority events, directly leading to delays in handling high-priority events and extremely low global security resource allocation efficiency. Furthermore, the air-to-ground collaborative tracking between drones and ground personnel relies entirely on human voice communication, with guidance parameters set using fixed values ​​and not linked to scheduling decisions. This results in significant communication delays, high information transmission errors, and low collaborative interception accuracy, failing to meet the practical security needs for rapid handling of moving targets.

[0017] This invention constructs a complete technical framework encompassing data acquisition, adaptation quantification, opportunity cost decision-making, resource allocation, and air-ground collaborative guidance. Through these core processes, it achieves opportunity cost-aware scheduling and automated air-ground collaboration for mobile targets in multi-event concurrent scenarios, fundamentally solving the core problems of resource mismatch in traditional security scheduling, delayed handling of high-priority events, and inefficient manual collaboration in mobile tracking.

[0018] This invention is based on real-time IoT sensing data. First, it completes the standardized digital collection of security events and idle resources, quantifies the comprehensive adaptability of resources to events as the basic fitness, and then combines the comprehensive priority of dynamically updated events to obtain the weighted fitness. The concept of opportunity cost is introduced to calculate relative benefits, and relative benefits are used as the core decision-making basis. Subsequently, according to the comprehensive priority of events from high to low, only the optimal resources with positive relative benefits are allocated to events. When the same mobile tracking event is simultaneously allocated to drones and ground personnel, dynamic collaborative guidance deeply linked with relative benefits is automatically triggered. The entire process does not require manual voice coordination, realizing the global optimal scheduling decision and the automated and precise execution of air-ground collaboration.

[0019] In one embodiment, step S1 specifically includes the following sub-steps: S11: Read event data from the IoT message bus and represent each event as structured data containing event identifier, event type, event location, basic priority and dynamic priority. The event type includes intrusion, leakage and fire, the event location is a 2D coordinate, the basic priority is a preset fixed value, and the initial value of the dynamic priority increment is 0. S12: Every fixed time period, the system accumulates the dynamic priority increment for each unprocessed event according to the increment step size corresponding to its event type. Specifically, for intrusion events, the dynamic priority increment is increased by a first value every period; for leakage events, the dynamic priority increment is increased by a second value every period, where the second value is less than the first value; and for fire events, the dynamic priority increment is increased by a third value every period, where the third value is greater than the first value. S13: Read idle resource data from the IoT message bus, and represent each resource as structured data containing resource identifier, resource type, resource location, normalized remaining energy and idle status. The resource type includes drones, ground personnel and patrol robots, the resource location is the current GPS coordinates, and the normalized remaining energy ranges from 0 to 1. The system only reads resources marked as "idle" and adds them to the list to be allocated. S14: Sort all unfinished events in real time according to the current comprehensive priority, and generate a priority event queue. The current comprehensive priority is equal to the sum of the base priority and the current dynamic priority increment. At the same time, group and cache all idle resources according to resource type for quick retrieval in subsequent steps.

[0020] As described in steps S11-S14 above, standardized data collection of unfinished events and idle resources within the security area can be completed, event dynamic priority can be updated in real time, data can be structured, sorted, and grouped for caching. This provides accurate, orderly, and real-time basic data for subsequent opportunity cost-aware scheduling and air-ground collaborative guidance based on relative benefits, ensuring the accuracy and timeliness of scheduling decisions in scenarios with multiple concurrent events.

[0021] In multi-dimensional security scenarios such as chemical industrial parks and large transportation hubs, multiple incidents such as intrusion, leakage, and fire may occur simultaneously. The urgency of handling these incidents increases with the waiting time, and the location, energy, and operational status of security resources such as drones, ground personnel, and patrol robots are constantly changing in real time. If a unified standard for event and resource data cannot be established, and the urgency cannot be dynamically adjusted based on event type and waiting time, it will directly lead to data chaos and inaccurate priority determination when multiple events occur concurrently. Ultimately, this results in high-capability resources being occupied by low-priority events and delays in handling high-priority events. Therefore, it is necessary to complete the digital modeling of events and resources, dynamic priority updates, and data standardization to solve the problems of missing basic data and fixed priorities in real-time security scheduling.

[0022] Traditional three-dimensional security dispatching relies solely on fixed event priorities, failing to adjust urgency based on waiting time. Furthermore, resource data collection formats are fragmented and lack standardized formats, hindering rapid resource-event matching in multi-event concurrent scenarios. This easily leads to mismatches, such as allocating high-priority resources like drones (suitable for intrusion events) to low-priority events like data leaks. By structurally collecting comprehensive event and resource data, dynamically updating priorities based on event hazard levels, and sorting, grouping, and caching the data, this approach addresses the shortcomings of traditional solutions, such as chaotic data management and rigid priority settings, laying a solid data foundation for subsequent opportunity cost-aware dispatching.

[0023] Step S11 reads event data from the IoT message bus, representing each event as structured data containing event identifier, event type, event location, basic priority, and dynamic priority increment. Event types are categorized into intrusion, leakage, and fire; event locations use two-dimensional coordinates; the basic priority is a preset fixed value; and the initial value of the dynamic priority increment is zero. This process transforms physical security events into standardized digital information, eliminating computational obstacles caused by differences in event data formats and establishing a unified data benchmark for subsequent comprehensive priority calculation and scheduling decisions.

[0024] Step S12, using fixed time periods, accumulates a dynamic priority increment for unhandled events according to their type and corresponding increment step size. Intrusion events increase by a first value each period, leakage events increase by a second value less than the first value, and fire events increase by a third value greater than the first value. Specifically, the first value for intrusion events can be set to 0.01, the second value for leakage events to 0.005, and the third value for fire events to 0.02. By setting differentiated priority growth rates based on the event's hazard level, the urgency of events increases reasonably with waiting time, preventing high-risk events such as fires and intrusions from being delayed due to waiting, and ensuring that high-priority events receive appropriate resources first.

[0025] Step S13 reads idle resource data from the IoT message bus, representing each resource as structured data containing resource identifier, resource type, resource location, normalized remaining energy, and idle status. Resource types are categorized into three types: drones, ground personnel, and patrol robots. Resource locations use current GPS coordinates, and normalized remaining energy ranges from 0 to 1. The system only includes idle resources in the allocation list. This process accurately filters schedulable idle resources, eliminates interference from non-idle resources, and quantifies resource status into standardized data, providing accurate input for subsequent resource and event fitness calculations.

[0026] Step S14 involves real-time sorting of all incomplete events according to a comprehensive priority obtained by adding the base priority and the dynamic priority increment, generating a priority event queue. Simultaneously, all idle resources are grouped and cached according to resource type. This comprehensive priority sorting clarifies the order of event handling, while resource grouping and caching simplify subsequent data retrieval processes, reduce the computation time of the scheduling algorithm, adapt to the low-latency requirements of real-time scheduling in security scenarios, and provide a well-organized data processing foundation for subsequent resource allocation based on relative benefits.

[0027] In one embodiment, step S2 specifically includes the following sub-steps: S21: For each combination of resource type and event type, the system pre-stores a matching degree value. The value of the matching degree value ranges from 0 to 1. The matching degree value represents the degree of matching between the resource type and the event type. The higher the matching degree, the more suitable the resource is for handling the event type. The matching degree value can be updated periodically based on the offline statistical results of historical handling success rate. S22: Calculate the proximity between resources and events: First, calculate the Euclidean distance between the current location of the resource and the location of the event. Then, subtract this distance from the maximum span of the security area. Finally, divide the difference by the maximum span to obtain the proximity. When the resource and event locations coincide, the proximity is equal to 1. When the distance is equal to the maximum span, the proximity is equal to 0. S23: Directly use the normalized residual energy of resources as the energy factor, with a value range of 0 to 1; S24: Calculate the basic fitness using the following weighted summation method: multiply the resource-event matching degree by the first weight coefficient, multiply the proximity degree by the second weight coefficient, multiply the energy factor by the third weight coefficient, and then add the three products together to obtain the basic fitness. The sum of the first, second, and third weight coefficients is 1, with the first weight coefficient having the largest value, the second weight coefficient being the second largest, and the third weight coefficient being the smallest. The basic fitness value ranges from 0 to 1, and the larger the value, the higher the immediate adaptability of the resource to the event.

[0028] As described in steps S21-S24 above, the quantitative calculation and weighted fusion of resource and event matching degree, distance proximity, and energy factor can be completed to obtain a standardized basic fitness value. This provides a unified and accurate quantitative evaluation basis for subsequent weighted fitness calculation and relative benefit solution, and solves the problem of unreasonable resource and event matching in multi-event concurrent scenarios from the perspective of resource adaptability.

[0029] In multi-dimensional security scenarios such as chemical industrial parks, different types of resources have inherent differences in their ability to handle different types of events. The distance between resources and events directly determines the response speed, while the remaining energy of resources affects the continuity of task execution. If there is a lack of a unified indicator that comprehensively quantifies the adaptability of resources and events, scheduling decisions can only rely on a single dimension for judgment, which cannot comprehensively measure the overall adaptability of resources to events. This leads to technical problems such as resource-event mismatch and low scheduling efficiency. Therefore, it is necessary to build a multi-dimensional integrated basic fitness calculation system to achieve accurate quantification of resource-event adaptability.

[0030] Traditional security dispatching methods rely solely on single rules such as distance priority or fixed type matching, failing to comprehensively consider three core factors: resource type adaptability, spatial distance, and operational capacity. This one-sided adaptability assessment fails to identify the optimal resources in multi-event concurrent scenarios, easily leading to a mismatch between resource handling capabilities and event demands. By pre-storing standardized matching degrees, normalizing distance proximity calculations, introducing energy factors, and employing multi-dimensional weighted calculations, a comprehensive and quantitative basic adaptability model is constructed. This model specifically addresses the shortcomings of traditional solutions, such as singular adaptability assessments and low resource matching accuracy.

[0031] Step S21 involves pre-storing a matching degree value for each resource type and each event type combination in the system. The value ranges from 0 to 1. This value characterizes the degree of matching between the resource's inherent ability to handle the event and can be periodically updated based on offline statistical results of historical handling success rates. Specifically, the matching degree value can be set as follows: 0.9 for drones to intrusion events, 0.3 for drones to leak events, 0.6 for ground personnel to intrusion events, and 0.8 for ground personnel to leak events. This quantifies the adaptability attributes based on the inherent capabilities of the resources, distinguishes the handling expertise of different resources, provides a core evaluation dimension for basic adaptability, and avoids mismatches between resource types and event types from the source.

[0032] Step S22 calculates the proximity between the resource and the event. First, the Euclidean distance between the current location of the resource and the event location is calculated. Then, this Euclidean distance is subtracted from the maximum span of the security area, and the difference is divided by the maximum span of the security area to obtain the final proximity score. The proximity score is 1 when the resource and event locations coincide, and 0 when the Euclidean distance equals the maximum span of the security area. For example, if the maximum span of the security area is set to 2000 meters, the proximity score is 1 when the distance between the resource and the event is 0 meters, and 0 when the distance is 2000 meters. This converts the spatial distance into a normalized value, eliminating scale differences between different security area sizes, quantifying the spatial response efficiency of the resource, and providing a spatial dimension basis for adaptability assessment.

[0033] Step S23 directly uses the normalized remaining energy of the resource as the energy factor, with a value range limited to between zero and one. No additional conversion calculations are required; standardized energy data is directly used to characterize the resource's continuous operational capability, eliminating resource interference due to insufficient remaining energy. This ensures that the scheduled resource has the basic energy conditions to complete event handling, guaranteeing the effective execution of the scheduling task.

[0034] Step S24 calculates the basic fitness using a weighted summation method. The resource-event matching degree is multiplied by a first weight coefficient, the proximity degree by a second weight coefficient, and the energy factor by a third weight coefficient. The products of these three are then added together to obtain the basic fitness. The sum of the three weight coefficients is 1, and the first weight coefficient is greater than the second, which in turn is greater than the third. Specifically, the first weight coefficient can be set to 0.5, the second to 0.3, and the third to 0.2. The basic fitness value ranges from zero to one; a higher value indicates a higher degree of immediate adaptability of the resource to the event. By integrating these three core evaluation dimensions, the core position of resource type matching degree is highlighted, resulting in a standardized comprehensive fitness value. This provides a unified benchmark for subsequent weighted fitness and relative benefit calculations, improving the accuracy of opportunity cost-aware scheduling and reducing resource mismatch problems from an adaptability perspective.

[0035] In one embodiment, step S3 specifically includes the following sub-steps: S31: Calculate the overall priority of each event, which is to add the base priority of the event to the current dynamic priority increment. The base priority is a fixed base number. Different event types have different base priorities. Fire events have the highest base priority, followed by intrusion events, and leakage events have the lowest base priority. S32: For each pair of resources and events, calculate the weighted fitness: multiply the basic fitness obtained in step S2 by the overall priority of the event to obtain the weighted fitness. The physical meaning of the weighted fitness is the benefit value after weighting the basic fitness with the event priority. The higher the priority of the event, the greater the weighted fitness it produces for the same resource. S33: For each resource and the current event, calculate the maximum weighted fitness that the resource could obtain if it were not allocated to the current event but to other events: that is, traverse the weighted fitness of the resource with all other events, find the maximum value among them, which reflects the opportunity cost of the resource and the best benefit that can be obtained from other events after giving up the current event; S34: Calculate relative benefit: Subtract the maximum weighted fitness that the resource can obtain for all other events from the weighted fitness of the resource for the current event. The difference is the relative benefit. When the relative benefit is greater than 0, it means that allocating the resource to the current event has a positive relative advantage compared to allocating it to other best events. When the relative benefit is less than or equal to 0, it means that allocating it to other events is at least no worse.

[0036] As described in steps S31-S34 above, by completing the comprehensive priority calculation of events, weighted fitness solution, resource opportunity cost quantification and relative benefit determination, an evaluation system with opportunity cost as the core is constructed, providing core decision-making basis for subsequent resource allocation, and fundamentally solving the core technical problem of high-adaptability resources being occupied by low-priority events in multi-event concurrent scenarios.

[0037] In multi-dimensional security scenarios such as chemical industrial parks, the urgency of different events varies significantly. Relying solely on basic fitness cannot reflect the weighted impact of event priority on resource allocation. Furthermore, allocating resources to a particular event incurs the opportunity cost of forgoing other events. Failure to quantify opportunity costs leads to a lack of a global perspective in scheduling decisions, making it impossible to guarantee that high-priority events receive optimal resources. Therefore, it is necessary to combine event priority weighting with opportunity cost calculation to form a relative benefit indicator that can directly guide allocation.

[0038] Traditional security dispatching relies solely on a greedy allocation method based on basic suitability or fixed priority, neglecting to consider the opportunity cost of resource allocation. When high-priority and low-priority events occur concurrently, valuable resources that are geographically close may be allocated to low-priority events, while high-priority events suffer delays due to a lack of suitable resources. By comprehensively weighting priorities, quantifying opportunity costs, and solving for relative benefits, the global opportunity cost is integrated into the dispatching evaluation, specifically addressing the shortcomings of traditional dispatching methods that lack a holistic perspective and suffer from resource misallocation.

[0039] Step S31 calculates the comprehensive priority of each event by adding the event's base priority to the current dynamic priority increment. The base priority is a preset fixed base number, and different event types have clearly defined hierarchical base priorities: fire events have the highest base priority, followed by intrusion events, and leakage events have the lowest. Specifically, the base priority for fire events can be set to 0.9, for intrusion events to 0.8, and for leakage events to 0.3. By combining a fixed urgency level with dynamic waiting time, a complete quantitative index of event priority is formed, ensuring that the urgency of events can be accurately quantified, providing a weighting basis for weighted fitness calculation.

[0040] Step S32 calculates the weighted fitness for each resource and event pair. The base fitness obtained in step S2 is multiplied by the overall priority of the event to obtain the weighted fitness. The weighted fitness represents the weighted benefit of resource allocation for an event; the higher the overall priority of the event, the greater the weighted fitness value for the same resource. Integrating event priority into fitness assessment amplifies the resource allocation value of high-priority events, giving them a natural advantage in resource competition and preventing low-priority events from crowding out high-quality resources.

[0041] Step S33 calculates the opportunity cost of resources by iterating through the weighted fitness of the current resource and all other events, extracting the maximum value. This maximum value represents the best benefit that could be obtained if the resource were not allocated to the current event, which is the opportunity cost of resource allocation. Introducing the concept of opportunity cost into security scheduling quantifies the global cost of resource allocation, allowing scheduling decisions to move beyond the suitability of a single event and instead maximize the benefits of all events.

[0042] Step S34 calculates the relative benefit by subtracting the resource's maximum weighted fitness for other events from its weighted fitness for the current event. A relative benefit greater than zero indicates that allocating the resource to the current event has a positive relative advantage; a relative benefit less than or equal to zero indicates that allocating the resource to other events will not result in a benefit loss. This forms the final core evaluation metric for scheduling, providing a clear decision-making standard for subsequent resource allocation, ensuring that every resource allocation generates a positive global benefit, and fundamentally preventing high-fitness resources from being occupied by low-priority events.

[0043] In one embodiment, step S4 specifically includes the following sub-steps: S41: Obtain the list of currently unfinished events, sort them in descending order according to the comprehensive priority calculated in step S31, and obtain the event processing queue. For multiple events with the same comprehensive priority, further sort them according to the order of their occurrence time, and process the events that occurred earlier. S42: For each event in the sorted event queue, perform the following operations in sequence: traverse each idle resource from the current list of idle resources, find the resource that maximizes the relative benefit based on the relative benefit value calculated in step S34, and record the maximum relative benefit value. S43: Determine if the maximum relative benefit value is greater than 0: If it is greater than 0, then formally allocate the resource to the current event and generate a scheduling instruction. The scheduling instruction includes at least the resource identifier, event location, and action mode. The action mode is determined according to the resource type and event type. For example, when a drone is allocated to an intrusion event, the action mode is "tracking". When ground personnel are allocated to a leak event, the action mode is "patrol". Then, remove the resource from the idle resource list and mark the current event as allocated. S44: If the maximum relative benefit value is less than or equal to 0, no resource allocation will be made for the current event, its unallocated state will be maintained, and the event will be skipped to continue processing the next event in the event queue; the event will re-enter step S41 in the next scheduling cycle. Since its dynamic priority increment may have increased and the overall priority has improved, the relative benefit may turn into a positive value.

[0044] As described in steps S41-S44 above, by completing event processing queue sorting, optimal resource retrieval and matching, allocation decision judgment, and temporary storage and loop processing of unqualified events, opportunity cost-aware resource allocation based on relative benefits is realized. This ensures that security resources are allocated only to high-priority events with positive relative benefits, and solves the core technical problems of resource mismatch and delayed processing of high-priority events in multi-event concurrent scenarios from the scheduling and execution level.

[0045] In multi-dimensional security scenarios such as chemical industrial parks and large transportation hubs, the number of idle security resources is always limited. When multiple events such as intrusion, fire, and leakage occur simultaneously, if resource allocation lacks a global benefit judgment standard and is assigned according to a single rule, it will directly lead to high-suit resources being occupied by low-priority events, and high-priority events being delayed in handling because they cannot obtain the best resources. Therefore, it is necessary to use relative benefit as the core judgment criterion and allocate resources in an orderly manner according to the urgency of the event to ensure that every resource allocation can generate positive global benefits.

[0046] Traditional security dispatch systems employ greedy allocation methods based on fixed priorities or distance, completely disregarding the opportunity cost of resource allocation. This results in a high probability of resource mismatch in multi-event concurrent scenarios, failing to achieve optimal resource-event matching. By prioritizing events, retrieving resources with maximum relative benefit, determining positive benefit allocation, and cyclically reprocessing unmet events, a complete opportunity cost-aware allocation process is constructed. This specifically addresses the shortcomings of traditional dispatch systems, such as lack of a global perspective and unreasonable resource assignment.

[0047] Step S41 retrieves the list of currently incomplete events, sorts them in descending order of overall priority, and generates an event processing queue. Events with the same overall priority are further sorted according to the order of their occurrence time, with earlier events processed first. For example, fire events have the highest overall priority and are ranked first in the event queue, followed by intrusion events, with leakage events at the end of the queue. If two intrusion events have the same overall priority, the earlier intrusion event is ranked first. This clearly defines the order of event handling, ensuring that high-priority, first-occurring events receive priority in resource allocation, aligning with the core requirement of prioritizing emergency event handling in security scenarios.

[0048] Step S42 involves iterating through all available resources in the current list of available resources for each sorted event, identifying the resource that maximizes the relative benefit of the current event based on its relative benefit value, and recording this maximum relative benefit value. For example, if the current event is an intrusion event, the system iterates through available drones, ground personnel, and patrol robots, calculates the relative benefit of each, and finally selects the drone with the highest relative benefit value as the candidate resource. This precise retrieval identifies the resource that best suits the current event and offers the highest benefit, providing a unique and accurate candidate for subsequent allocation decisions.

[0049] Step S43 determines whether the maximum relative benefit value is greater than 0. If the maximum relative benefit is greater than 0, the candidate resource is formally allocated to the current event, generating a scheduling instruction containing the resource identifier, event location, and action mode. The action mode is determined based on the resource type and event type. Simultaneously, the resource is removed from the idle resource list, and the current event is marked as allocated. For example, the maximum relative benefit of a drone for an intrusion event is 0.23, which is greater than 0. The system allocates the drone to perform the intrusion handling task, and the generated scheduling instruction includes the drone identifier, intrusion event location, and tracking action mode. By strictly enforcing the positive benefit allocation principle, it ensures that each resource allocation improves the overall scheduling benefit, fundamentally avoiding meaningless resource assignments.

[0050] Step S44 determines if the maximum relative benefit is less than or equal to 0. In this case, no resource allocation is performed on the current event; the event remains unallocated and is skipped. The event will re-enter the sorting process in the next scheduling cycle. As the dynamic priority increment accumulates, the overall priority of the event increases, and the relative benefit can turn positive. For example, if the current maximum relative benefit of an intrusion event is -0.05, the system will not allocate resources. In the next scheduling cycle, after the dynamic priority of the intrusion event accumulates, the overall priority increases, and the relative benefit turns to 0.18, at which point resource allocation can be completed. This avoids resource waste caused by negative benefit allocation and ensures that high-priority events eventually obtain appropriate resources through a cyclic reprocessing mechanism, maintaining the rationality and continuity of the scheduling system.

[0051] In one embodiment, step S5 specifically includes the following sub-steps: S51: After the allocation is completed in step S4, check if there is a case where the same event is simultaneously allocated to both drone resources and ground personnel resources, and if the event type is intrusion, i.e., it belongs to the moving target tracking type. If this condition is met, activate the cooperative guidance mode. If the conditions are not met, this step will terminate. S52: Obtain the data reported by the UAV at fixed time intervals. The data includes the UAV's own GPS coordinates, gimbal orientation angle, and the pixel offset of the target in the image plane. Use the camera imaging model to perform back projection calculation, and combine the UAV's flight altitude and gimbal angle to calculate the absolute coordinates of the target in the global coordinate system. Then, use the difference method of two consecutive measurements to estimate the instantaneous velocity vector of the target. To improve the stability of the velocity estimation, perform a moving average filter on the velocity vectors obtained from the most recent measurements. S53: Extract the relative benefit value of the UAV to the current event from step S34, and calculate the dynamic prediction time according to the following rules: take the baseline prediction time as the basis, add the product of the relative benefit value and the preset ratio coefficient to obtain the dynamic prediction time, and then limit the dynamic prediction time between the preset minimum prediction time and the maximum prediction time. The larger the relative benefit value, the greater the advantage of the UAV to this task is compared with other tasks. Therefore, a longer prediction time is adopted to enable ground personnel to move to a more distant interception point in advance. S54: Calculate the predicted target position: Add the target's absolute coordinates at the current moment to the product of the target's moving speed vector and the dynamic prediction time to obtain the predicted position. Then generate a guidance vector, which is equal to the predicted position minus the current GPS coordinates of the ground personnel. Convert the guidance vector into a azimuth angle and scalar distance relative to true north, and push it to the handheld terminal of the ground personnel through the IoT message queue telemetry transmission protocol. The terminal displays the guidance information in the form of dynamic arrows and remaining distance text. S55: Simultaneously generate gimbal servo commands for the UAV: ​​Calculate the desired gimbal horizontal rotation angle and desired pitch angle based on the horizontal and vertical differences between the target and the UAV, so that the target is always in the center of the UAV image screen. Write the desired horizontal rotation angle and pitch angle into the UAV's gimbal control interface. Repeat steps S52 to S55 at fixed time intervals until the event is handled or the resources are reallocated.

[0052] As described in steps S51-S55 above, by completing the collaborative guidance trigger determination, target motion parameter estimation, dynamic prediction duration calculation, guidance vector generation and push and UAV gimbal synchronous control, automated air-ground collaborative guidance based on relative benefits is achieved, solving the technical problems of low efficiency of human voice coordination, disjointed cooperation and high interception failure rate in mobile target tracking scenarios.

[0053] In security scenarios such as tracking moving targets intrusions in chemical industrial parks, when drones and ground personnel coordinate their operations, traditional methods rely entirely on walkie-talkie voice communication for route planning. This results in significant communication delays and high information transmission errors. Furthermore, the guidance parameters are set with fixed values, making it impossible to adjust the guidance intensity based on resource suitability for the task. This can easily lead to situations where ground personnel are slow to intercept and drone targets are lost. Therefore, it is necessary to establish an automated guidance mechanism that is linked to the relative benefits of scheduling to achieve precise air-ground coordination without human intervention.

[0054] Traditional air-to-ground collaborative guidance relies solely on fixed-duration geometric position predictions, failing to link with relative benefit indicators for scheduling decisions. This results in a mismatch between guidance intensity and task adaptability, and a lack of synchronous collaborative control from the UAV gimbal, leaving air-to-ground coordination entirely dependent on human experience. By precisely determining trigger conditions, stably estimating target motion states, dynamically adjusting prediction duration, generating guidance vectors in real time, and synchronously controlling gimbal angles, a collaborative guidance process deeply bound to relative benefits is constructed. This effectively addresses the shortcomings of traditional guidance, such as high reliance on human intervention, rigid parameters, and low collaborative accuracy.

[0055] Step S51, after resource allocation in step S4, checks whether the same event is simultaneously allocated both UAV and ground personnel resources, and whether the event type is intrusion-related mobile target tracking. If the conditions are met, the collaborative guidance mode is activated; otherwise, this step is terminated. For example, if an intrusion event is simultaneously allocated both UAV and ground personnel resources, the system immediately activates collaborative guidance; if only a single resource is allocated, the guidance process is not triggered. This allows for precise limitation of guidance triggering scenarios, avoiding invalid guidance calculations, and ensuring that guidance is only initiated in necessary scenarios for air-to-ground collaborative tracking, thus guaranteeing the rational use of system computing resources.

[0056] Step S52 acquires the UAV's own GPS coordinates, gimbal orientation angle, and target pixel offset reported at fixed time intervals. The target's global absolute coordinates are calculated through back projection using the camera imaging model. The instantaneous velocity vector of the target is estimated using a two-times measurement difference method, and a moving average filter is applied to the most recent velocity measurements to improve stability. The fixed time interval is set to 0.5 seconds, and the moving average filter using three velocity measurements effectively eliminates velocity estimation errors caused by measurement noise. This accurately obtains the target's true motion state, providing a stable and reliable data foundation for subsequent position prediction and solving the core problem of inaccurate motion estimation of moving targets.

[0057] Step S53 extracts the relative benefit value of the drone for the current event. Based on the baseline prediction duration, the product of the relative benefit and a preset proportional coefficient is added to obtain the dynamic prediction duration, which is then limited to a preset minimum prediction duration and a preset maximum prediction duration. The specific formula for calculating the dynamic prediction duration is as follows: ; Among them, the Indicates the duration of dynamic prediction. Indicates the baseline forecast duration. This represents the proportionality coefficient. This indicates the relative benefit of drones to the current event. Indicates the minimum prediction duration. This represents the maximum prediction duration. The baseline prediction duration is set to 2 seconds, the scaling factor is set to 2 seconds, the minimum prediction duration is 1 second, and the maximum prediction duration is 5 seconds. When the relative benefit of the drone is 0.23, the dynamic prediction duration is calculated to be 2.46 seconds. The larger the relative benefit value, the longer the prediction duration, and the more aggressive the guidance strategy. Linking the core guidance parameters with the scheduling relative benefit allows the guidance intensity to match the drone's adaptability to the mission, improving the foresight of collaborative interception.

[0058] Step S54 multiplies the target's current absolute coordinates with its velocity vector and dynamic prediction duration, then adds the results to obtain the target's predicted position. The predicted position is then subtracted from the ground personnel's current coordinates to generate a guidance vector. This guidance vector is converted into a azimuth angle relative to true north and a scalar distance, and pushed to the ground personnel's handheld terminal via a message queue telemetry transmission protocol, displayed with dynamic arrows and text. For example, if the target's predicted position is (100, 200) and the ground personnel's current coordinates are (80, 180), the system generates the corresponding guidance vector and pushes it to the terminal. This provides ground personnel with real-time and accurate interception guidance, completely replacing voice communication and significantly reducing coordination delays.

[0059] Step S55 calculates the desired horizontal rotation angle and pitch angle of the UAV gimbal based on the horizontal and vertical differences between the target and the UAV, and writes the angle parameters into the UAV gimbal control interface to keep the target centered in the frame. This process is repeated at fixed time intervals until the event is resolved. This enables synchronous and coordinated control of the UAV gimbal, ensuring the UAV continuously and stably locks onto the target, providing uninterrupted observation data for the guidance process, forming a complete air-ground coordinated guidance closed loop, and ensuring the efficient completion of the moving target tracking task.

[0060] like Figure 2 As shown, this invention also discloses a three-dimensional security linkage dispatch system based on the Internet of Things, comprising: The acquisition module is used to acquire a list of incomplete events and idle resources within the security area. The incomplete events include event type, event location, basic priority, and dynamic priority. The idle resources include resource type, resource location, and remaining energy. The first calculation module is used to calculate the basic fitness for each idle resource and each unfinished event. The basic fitness is a weighted sum of resource-event matching degree, distance proximity, and remaining energy. The second calculation module is used to weight the basic fitness according to the comprehensive priority of the unfinished events to obtain a weighted fitness, and to calculate the relative benefit of each idle resource relative to each unfinished event based on the weighted fitness. The processing module is used to process unfinished events in descending order of their overall priority. It selects the idle resources that maximize the relative benefit of the current event and have a relative benefit greater than zero for allocation. If there are no resources with a relative benefit greater than zero, the event is skipped. The sending module is used to trigger collaborative guidance when the same event is assigned to both a drone and a ground personnel and the event type is mobile tracking. It estimates the target's moving speed based on the drone's observations and predicts the target's future position with a prediction duration that is positively correlated with the drone's relative benefit from the event. It then generates a guidance vector from the ground personnel's current position to the predicted position and sends it to the ground personnel's terminal.

[0061] In one embodiment, the processing module includes: The sorting unit is used to sort unfinished events in descending order of overall priority, and in order of occurrence when priorities are the same. The optimization unit is used to find the resource that maximizes the relative benefit of each sorted event from the available resources and record the maximum relative benefit. The first processing unit is used to allocate the resource to the current event when the maximum relative benefit is greater than zero, generate a scheduling instruction containing the resource identifier, event location and action mode, and remove the resource from the idle list. The second processing unit is used to temporarily not allocate the maximum relative return if it is not greater than zero, and to skip the event and continue processing the next event.

[0062] This application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described three-dimensional security linkage scheduling method based on the Internet of Things.

[0063] This application also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the above-described IoT-based three-dimensional security linkage scheduling method.

[0064] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in this application and in the embodiments can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0065] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, apparatus, article, or method. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, apparatus, article, or method that includes that element.

[0066] The above description is merely a preferred embodiment of the present invention and does not limit the scope of this application. Any equivalent results or equivalent process transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the scope of protection of this application.

Claims

1. A three-dimensional security linkage scheduling method based on the Internet of Things, characterized in that, Includes the following steps: Obtain a list of incomplete events and available resources within the security area. The incomplete events include event type, event location, basic priority, and dynamic priority. The available resources include resource type, resource location, and remaining energy. For each idle resource and each incomplete event, calculate the basic fitness, which is a weighted sum of resource-event matching degree, distance proximity, and remaining energy; The base fitness is weighted according to the overall priority of unfinished events to obtain a weighted fitness, and the relative benefit of each idle resource relative to each unfinished event is calculated based on the weighted fitness. Unfinished events are processed in descending order of their overall priority. For the current event, an idle resource that maximizes its relative benefit and is greater than zero is selected for allocation. If there is no resource with a relative benefit greater than zero, the event is skipped. When the same event is assigned to both a drone and ground personnel and the event type is mobile tracking, collaborative guidance is triggered. The target's moving speed is estimated based on the drone's observations, and the target's future position is predicted using a prediction duration that is positively correlated with the drone's relative benefit from the event. A guidance vector is generated from the ground personnel's current position to the predicted position and sent to the ground personnel's terminal.

2. The IoT-based three-dimensional security linkage scheduling method according to claim 1, characterized in that, The steps for obtaining the list of incomplete events and available resources within the security area specifically include: Read event data from the IoT message bus and represent each event as structured data containing event identifier, event type, event location, basic priority, and dynamic priority; At fixed time intervals, the dynamic priority of unprocessed events is accumulated according to the incremental step size corresponding to their event type. Read idle resource data from the IoT message bus, represent each resource as structured data containing resource identifier, resource type, resource location, normalized remaining energy and idle status, and only read resources marked as idle in the idle status into the list to be allocated; All unfinished events are sorted in real time according to their current overall priority to generate an event queue, and idle resources are grouped and cached by type.

3. The IoT-based three-dimensional security linkage scheduling method according to claim 1, characterized in that, The step of calculating the basic fitness for each idle resource and each incomplete event specifically includes: For each combination of resource type and event type, a matching degree value is pre-stored to represent the degree to which the resource's ability to handle the event matches. The proximity between resources and events is calculated and normalized based on the spatial distance between the resource location and the event location, as well as the maximum span of the security area, so that the proximity is negatively correlated with the spatial distance. The normalized residual energy of the resources is used as the energy factor; The basic fitness is obtained by multiplying the matching degree, distance proximity, and energy factor by their respective preset weight coefficients and then summing them.

4. The IoT-based three-dimensional security linkage scheduling method according to claim 1, characterized in that, The step of calculating the relative benefit of each idle resource relative to each incomplete event based on the weighted fitness specifically includes: Calculate the overall priority of each event, which is the sum of the base priority and the current dynamic priority. The base priority is different for different event types. For each resource and event pair, the base fitness is multiplied by the overall priority of the event to obtain the weighted fitness. For each resource, calculate the maximum weighted fitness that it could obtain by being assigned to other events instead of the current event, and use that as the opportunity cost of the resource. The relative benefit is obtained by subtracting the opportunity cost from the weighted fitness of the current event.

5. The IoT-based three-dimensional security linkage scheduling method according to claim 1, characterized in that, The step of selecting idle resources that maximize the relative benefit of the current event and have a relative benefit greater than zero for allocation specifically includes: Unfinished events are sorted in descending order of overall priority, and if the priorities are the same, they are sorted in order of occurrence. For each sorted event, find the resource that maximizes the relative benefit of that event from the available resources, and record the maximum relative benefit. If the maximum relative benefit is greater than zero, the resource is allocated to the current event, a scheduling instruction containing the resource identifier, event location and action mode is generated, and the resource is removed from the idle list. If the maximum relative return is not greater than zero, then no allocation will be made, and the event will be skipped to continue processing the next event.

6. The IoT-based three-dimensional security linkage scheduling method according to claim 1, characterized in that, The step of generating a guidance vector pointing from the current location of the ground personnel to the predicted location and sending it to the ground personnel's terminal specifically includes: After resource allocation is completed, check if the same event is simultaneously assigned to both drones and ground personnel and if the event is moving target tracking. If so, activate the collaborative guidance mode. The system acquires the UAV's periodically reported self-position, gimbal orientation, and target image offset data, calculates the target's absolute coordinates, and estimates the target's instantaneous velocity vector based on the coordinate differences at consecutive time points. The prediction duration is dynamically determined based on the relative benefit value of the drone to the event, so that the prediction duration is positively correlated with the relative benefit. The predicted position is calculated based on the target's current position, velocity vector, and the predicted duration. A guiding vector is generated from the ground personnel's current position to the predicted position and sent to the ground personnel's terminal. Simultaneously, gimbal servo commands are generated based on the target's position relative to the drone to keep the target centered in the drone's view; the above guidance process is repeated until the event is resolved or resources are reallocated.

7. A three-dimensional security linkage dispatch system based on the Internet of Things, characterized in that, include: The acquisition module is used to acquire a list of incomplete events and idle resources within the security area. The incomplete events include event type, event location, basic priority, and dynamic priority. The idle resources include resource type, resource location, and remaining energy. The first calculation module is used to calculate the basic fitness for each idle resource and each unfinished event. The basic fitness is a weighted sum of resource-event matching degree, distance proximity, and remaining energy. The second calculation module is used to weight the basic fitness according to the comprehensive priority of the unfinished events to obtain a weighted fitness, and to calculate the relative benefit of each idle resource relative to each unfinished event based on the weighted fitness. The processing module is used to process unfinished events in descending order of their overall priority. It selects the idle resources that maximize the relative benefit of the current event and have a relative benefit greater than zero for allocation. If there are no resources with a relative benefit greater than zero, the event is skipped. The sending module is used to trigger collaborative guidance when the same event is assigned to both a drone and a ground personnel and the event type is mobile tracking. It estimates the target's moving speed based on the drone's observations and predicts the target's future position with a prediction duration that is positively correlated with the drone's relative benefit from the event. It then generates a guidance vector from the ground personnel's current position to the predicted position and sends it to the ground personnel's terminal.

8. The IoT-based three-dimensional security linkage dispatch system according to claim 7, characterized in that, The processing module includes: The sorting unit is used to sort unfinished events in descending order of overall priority, and in order of occurrence when priorities are the same. The optimization unit is used to find the resource that maximizes the relative benefit of each sorted event from the available resources and record the maximum relative benefit. The first processing unit is used to allocate the resource to the current event when the maximum relative benefit is greater than zero, generate a scheduling instruction containing the resource identifier, event location and action mode, and remove the resource from the idle list. The second processing unit is used to temporarily not allocate the maximum relative return if it is not greater than zero, and to skip the event and continue processing the next event.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images