Blockchain-game theory-based dynamic permission allocation method and system for cooperative scheduling of unmanned aerial vehicles

Abstract

The application discloses a kind of based on block chain-game theory's unmanned vehicle cooperative scheduling dynamic permission allocation method and system, the method includes steps: collection real-time field observation data and is converted into disaster environment and personnel characteristic vector;Disaster environment and personnel characteristic vector are quantitatively processed, and output comprehensive priority score;When detecting the comprehensive priority score is greater than preset critical judgment threshold value, generate dynamic preemption instruction;The dynamic preemption instruction is loaded into the game theory cooperative scheduling model in blockchain network architecture, and the corresponding global resource allocation result and local optimal path planning result are output.The technical scheme provided by the application can realize accurate judgment by multi-source data fusion and dynamic grading, rely on blockchain smart contract and game theory model to optimize cross-department resource preemption and cooperative scheduling, ensure the real-time of rescue decision, the efficiency of resource allocation and the traceability of process in complex disaster scene.

Description

Technical Field

[0001] This invention relates to the technical field of UAV collaborative scheduling, and in particular to a dynamic permission allocation method and system for UAV collaborative scheduling based on blockchain and game theory. Background Technology

[0002] With the increasing frequency of extreme weather events due to global climate change, the demand for drone technology in flood emergency rescue is becoming increasingly prominent. However, existing drone dispatch systems have revealed significant limitations in complex rescue environments. Permission allocation mechanisms tend to be static, often employing predefined or dynamic authorization models. This prevents rescue departments from flexibly adjusting permissions based on real-time situational changes when missions suddenly shift, and lacks efficient cross-departmental collaboration and resource sharing mechanisms, easily leading to duplicate deployments of drone resources or dispatch conflicts. Traditional dispatch architectures often rely on a single central node for data aggregation and command distribution, resulting in severely insufficient system robustness. In harsh low-altitude environments such as floods, if the central command node fails due to environmental damage or communication link disruptions, the entire dispatch network faces the risk of paralysis, severely restricting the continuity of emergency response and rescue efficiency.

[0003] Furthermore, existing technologies still have significant shortcomings in data security and accountability. Critical operational records such as flight trajectories, mission instructions, and execution feedback during drone search and rescue missions often lack robust anti-tampering mechanisms, making operational data susceptible to loss or malicious alteration. This poses significant challenges to subsequent rescue effectiveness assessments and accident liability determination. While blockchain technology's decentralized architecture, consensus mechanism, and immutability offer theoretical possibilities for solving these problems, in practical applications, there is a lack of mature technologies for real-time feature extraction from multi-source heterogeneous sensor data, for achieving dynamic resource allocation and optimization across departments through complex game theory models while ensuring privacy, and for achieving autonomous disaster recovery scheduling and incremental data synchronization of drone swarms under extreme conditions of wide-area communication disruptions. Therefore, properly addressing these issues has become an urgent task for the industry. Summary of the Invention

[0004] This invention addresses the aforementioned problems in existing technologies by providing a blockchain-game theory-based method and system for dynamic permission allocation in drone collaborative scheduling. It achieves accurate assessment of disaster sites through multi-source data fusion and dynamic scoring, and solves the problem of resource contention and collaborative scheduling of drones across departments by relying on smart contracts and game theory models. While ensuring optimal allocation of global resources, blockchain-based notarization guarantees the transparency and reliability of the process, improving the response efficiency and collaborative decision-making level of emergency rescue in complex environments.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: A blockchain-game theory-based method for dynamic permission allocation in drone collaborative scheduling is provided, including: S1, collect real-time field observation data obtained by multi-source heterogeneous sensors mounted on the UAV platform, and convert the collected real-time field observation data into a disaster environment and personnel feature vector containing a unified timestamp and normalized scaling. S2, input the disaster environment and personnel feature vectors into a preset dynamic priority calculation model based on a multi-level evaluation system for quantification, and output a comprehensive priority score for each current parallel rescue task; S3, when the comprehensive priority score is detected to be greater than the preset critical assessment threshold, the emergency access smart contract deployed at the bottom layer of the distributed blockchain network is triggered to generate a dynamic preemption instruction. S4. The dynamic preemption instruction is loaded as an initialization input variable into the game theory collaborative scheduling model in the blockchain network architecture. Multi-stage game solving calculation based on dynamic decision tree is performed among multiple department drone nodes. The corresponding global resource allocation results and local optimal path planning results are output and encrypted and stored on the blockchain.

[0006] Furthermore, in step S4, the multi-stage game-theoretic calculation based on dynamic decision trees is performed among multiple departmental drone nodes, outputting the corresponding global resource allocation results and local optimal path planning results, and then encrypted and stored on the blockchain, including: S41, the command center node, which acts as the leader of the game, receives the dynamic preemption instruction and constructs a global resource allocation objective function that maximizes the benefits of system resource allocation and minimizes the cost of inter-departmental task conflicts based on the preset departmental function weight index and task urgency variable. S42, the global resource allocation objective function is broadcast to the drone nodes of each department as game followers, driving the drone nodes of each department to generate preliminary local obstacle avoidance paths and corresponding estimated resource consumption data in combination with their own battery remaining energy consumption physical constraints. S43, the collaborative scheduling model is called in real time through the application programming interface of the emergency access smart contract to obtain the estimated resource consumption data and real-time environmental parameters. Under the convergence condition of satisfying the preset conflict penalty coefficient, the Stackelberg game equilibrium solution of the global resource allocation objective function is solved to generate the local optimal path planning result. The local optimal path planning result includes the avoidance and detour strategy of three-dimensional spatial coordinates and the airspace release instruction.

[0007] Furthermore, in step S2, the disaster environment and personnel feature vectors are input into a preset dynamic priority calculation model based on a multi-level evaluation system for quantification, and a comprehensive priority score for each current parallel rescue task is output, including: S21, analyze and extract the body temperature index features, respiratory rate features and absolute flood diffusion speed features contained in the disaster environment and personnel feature vector, perform weighted summation calculation based on deviation degree and physical critical threshold comparison operation respectively, and calculate the quantitative vital signs comprehensive score and environmental risk assessment value. S22, obtain the real-time matching degree variable of the disaster environment and personnel feature vector with respect to the number of drones owned by the current department, the reserve capacity of special materials and the weight of the task type, and calculate the functional correlation score of each participating rescue department; S23, substitute the comprehensive vital signs score, the environmental risk assessment value, and the functional correlation score of each participating rescue department into the preset priority quantification scoring function matrix, and call the time series analysis of the evaluation weight vector distribution in the function matrix with the introduction of dynamic forgetting factor to output the comprehensive priority score.

[0008] Furthermore, in step S4, after outputting the corresponding global resource allocation result and local optimal path planning result, and encrypting and storing them on the blockchain, the process further includes: S5. Monitor the connectivity status of the wide area communication link between the drone nodes of each department and the command center node in the area in real time. When the connectivity status parameter is detected to be lower than the preset network connectivity judgment threshold and the wide area communication network is determined to be disconnected and paralyzed, switch the underlying communication hardware module of the corresponding drone node to the ultra-wideband technology communication channel and broadcast the recent airspace scheduling records and equipment physical status information cached locally to the outside. S6, receive the recent airspace scheduling records and equipment physical status information broadcast by neighboring UAV nodes through the ultra-wideband technology communication channel, wherein the neighboring UAV nodes are in the same spatial physical range, and autonomously initiate a master node election procedure among multiple interconnected UAV node groups based on an improved distributed fault-tolerant consensus protocol, and generate a temporary master control scheduling node with temporary decision-making authority. S7, the temporary master control scheduling node elected performs offline disaster recovery decision calculation based on majority voting mechanism according to the aggregated physical status information of the equipment, reorders and generates the task priority order table in the current offline local area network environment and issues corresponding flight attitude physical control commands to guide the drone group in the fault area to perform low-risk life search and rescue cruise missions along the emergency flight corridor of the pre-set safety boundary.

[0009] Furthermore, in step S7, after the drone swarm guiding the fault area performs a low-risk life-saving search and rescue patrol mission along an emergency flight corridor with a pre-defined safety boundary, the process also includes: S8, continuously execute the monitoring program for the connectivity status of the wide area communication link, and when it is confirmed that the connectivity status parameters of the communication link have recovered to the normal working threshold range, package and encapsulate the recent airspace scheduling records and offline task operation execution logs generated in the local memory of the UAV during the wide area communication network disconnection into an incremental synchronization data packet; S9, control the airborne edge computing processing unit installed inside the drone nodes of each department to process the incremental synchronization data packet, extract the key coordinate hash feature value containing task attributes, perform lossless data compression encoding processing, and then send the compressed and encoded incremental synchronization data packet to the data receiving node of the distributed blockchain network through the wide area low power IoT communication protocol; S10, trigger the conflict verification smart contract deployed in the underlying architecture of the distributed blockchain network, compare and verify the spatial physical deviation between the key coordinate hash feature value and the dataset reported by the neighboring system nodes, and when it is determined that there is a cross-departmental collaborative competition conflict in the same three-dimensional physical space, the valid operation is determined by the dual standard of the priority principle of the timestamp sequence of the network global clock and the urgency level of the vital signs of the trapped person, and the conflict decision log containing timestamp information is solidified and written into the block body of the block storage layer.

[0010] Furthermore, in step S3, triggering the emergency access smart contract deployed at the bottom layer of the distributed blockchain network to generate a dynamic preemption instruction includes: S31, when different rescue departments initiate a cross-domain restricted airspace permission application request within the comprehensive priority scoring triggering system, the underlying drone edge computing node that made the application request is controlled to perform cryptographic operations to generate a zero-knowledge concise non-interactive knowledge demonstration protocol validity logic certificate that conceals the geographical coordinates of the real trapped person, and the validity logic certificate is submitted and uploaded to the distributed blockchain network. S32, control the distributed consensus node cluster in the distributed blockchain network, use the matching cryptographic verification algorithm function to review and verify the validity logical certificate, after the review and verification is passed and the node's digital identity is confirmed, grant the corresponding level of high-priority restricted airspace occupation permission to the underlying drone edge computing node that made the application, and generate the dynamic preemption instruction based on the high-priority restricted airspace occupation permission. S33. When the system enters the closed-loop optimization phase after the scheduling cycle ends, it calls the federated learning model aggregation framework to update the underlying scheduling model algorithm parameters of the cross-department local drone nodes. It outputs the initial gradient parameter calculation matrix to each department endpoint through the smart contract, and superimposes Laplace distribution statistical noise into the matrix to generate system scheduling model iteration parameters with the ability to resist model reverse feature back-inference attacks, thereby completing the global differential privacy protection processing.

[0011] According to a second aspect of the present invention, a blockchain-game theory-based dynamic permission allocation system for collaborative scheduling of unmanned aerial vehicles (UAVs) is provided, comprising: The acquisition module is used to acquire real-time field observation data obtained by multi-source heterogeneous sensors mounted on the UAV platform, and convert the acquired real-time field observation data into disaster environment and personnel feature vectors containing a unified timestamp and normalized scaling. The output module is used to input the disaster environment and personnel feature vectors into a preset dynamic priority calculation model based on a multi-level evaluation system for quantitative processing, and output a comprehensive priority score for the current parallel rescue tasks. The generation module is used to trigger an emergency access smart contract deployed at the bottom layer of the distributed blockchain network to generate a dynamic preemption instruction when the comprehensive priority score is detected to be greater than the preset critical assessment threshold. The calculation module is used to load the dynamic preemption instruction as an initialization input variable into the game theory collaborative scheduling model in the blockchain network architecture, perform multi-stage game solving calculation based on dynamic decision tree among multiple department drone nodes, output the corresponding global resource allocation results and local optimal path planning results, and perform encrypted on-chain storage.

[0012] The present invention also provides a blockchain-game theory-based UAV collaborative scheduling dynamic permission allocation system, comprising interconnected processors and computer-readable storage media, wherein the computer-readable storage media stores a computer program, which is executed by the processor to implement the steps of the above-described blockchain-game theory-based UAV collaborative scheduling dynamic permission allocation method.

[0013] The present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the above-described blockchain-game theory-based UAV collaborative scheduling dynamic permission allocation method.

[0014] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the above-described blockchain-game theory-based UAV collaborative scheduling dynamic permission allocation method.

[0015] Compared with the prior art, the advantages of the present invention are as follows: This invention achieves precise perception and intelligent classification of rescue missions by standardizing and dynamically prioritizing multi-source sensor data, thereby improving the accuracy of emergency decision-making. Relying on the smart contract mechanism of distributed blockchain and the game theory-based collaborative scheduling model, it solves the resource competition problem of cross-departmental drone swarms in dynamic environments, ensuring optimal allocation of global resources and planning of local paths through multi-stage game theory. Simultaneously, the encrypted on-chain evidence storage mechanism guarantees the transparency and traceability of the scheduling process, enhancing the response speed, collaborative efficiency, and safety and reliability of rescue operations in complex disaster scenarios. Attached Figure Description

[0016] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0017] Figure 1 A flowchart of a dynamic permission allocation method for UAV collaborative scheduling based on blockchain-game theory is provided for embodiments of the present invention; Figure 2 A flowchart of another method for dynamic permission allocation in collaborative scheduling of unmanned aerial vehicles based on blockchain-game theory, provided for embodiments of the present invention; Figure 3 A flowchart illustrating yet another method for dynamic permission allocation in collaborative scheduling of unmanned aerial vehicles based on blockchain-game theory, provided as an embodiment of the present invention; Figure 4 A flowchart illustrating yet another method for dynamic permission allocation in collaborative scheduling of unmanned aerial vehicles based on blockchain-game theory, provided as an embodiment of the present invention; Figure 5 A flowchart illustrating yet another method for dynamic permission allocation in collaborative scheduling of unmanned aerial vehicles based on blockchain-game theory, provided as an embodiment of the present invention; Figure 6 A flowchart illustrating yet another method for dynamic permission allocation in collaborative scheduling of unmanned aerial vehicles based on blockchain-game theory, provided as an embodiment of the present invention; Figure 7 A structural diagram of a blockchain-game theory-based UAV collaborative scheduling dynamic permission allocation system is provided as an embodiment of the present invention; Figure 8 A schematic diagram of the data acquisition layer provided for an embodiment of the present invention; Figure 9 A schematic diagram of the decision core layer provided for embodiments of the present invention; Figure 10 A schematic diagram of the execution scheduling layer provided for embodiments of the present invention; Figure 11 A schematic diagram of an emergency protection layer provided for an embodiment of the present invention; Figure 12 A schematic diagram of a security protection layer provided for an embodiment of the present invention; Figure 13 A schematic diagram of the continuous optimization layer provided for an embodiment of the present invention. Detailed Implementation

[0018] The present invention will be further described below with reference to the accompanying drawings and specific preferred embodiments, but this does not limit the scope of protection of the present invention.

[0019] The technical solution adopted in this embodiment is as follows: Figure 1 As shown, it includes the following steps: S1, collect real-time field observation data obtained by multi-source heterogeneous sensors mounted on the UAV platform, and convert the collected real-time field observation data into a disaster environment and personnel feature vector containing a unified timestamp and normalized scaling. S2 inputs the disaster environment and personnel feature vectors into a preset dynamic priority calculation model based on a multi-level evaluation system for quantification, and outputs a comprehensive priority score for each current parallel rescue task; S3, when the overall priority score is detected to be greater than the preset critical assessment threshold, the emergency access smart contract deployed at the bottom layer of the distributed blockchain network is triggered to generate a dynamic preemption instruction; S4 loads the dynamic preemption command as an initialization input variable into the game theory collaborative scheduling model in the blockchain network architecture. It then performs multi-stage game solving calculations based on dynamic decision trees among multiple departmental drone nodes, outputting the corresponding global resource allocation results and local optimal path planning results, and encrypting and storing them on the blockchain.

[0020] In this embodiment, the blockchain-game theory-based UAV collaborative scheduling dynamic permission allocation method is mainly applied to flood emergency rescue scenarios in the context of low-altitude economy.

[0021] Step S1 involves using multiple departmental drone nodes deployed within the operational area to collect real-time field observation data from their onboard multi-source heterogeneous sensors. These multi-source heterogeneous sensors include, but are not limited to, visible light cameras, infrared thermal imagers, lidar, and airborne life detection radar. Due to significant differences in sampling frequency, data format, and dimensions among the various sensors, the system preprocesses the collected real-time field observation data. Through resampling and feature mapping, the data is converted into a disaster environment and personnel feature vector containing a unified timestamp and a normalized scaling. This disaster environment and personnel feature vector constructs a multi-dimensional space, transforming complex on-site physical information into a computer-recognizable digital representation. The unified timestamp ensures time synchronization for multi-drone collaboration, while the normalized scaling eliminates the dimensional influence between different physical quantities.

[0022] Step S2 involves inputting the constructed disaster environment and personnel feature vectors into a pre-defined dynamic priority calculation model based on a multi-level assessment system for quantification. This multi-level assessment system encompasses a vital signs layer, an environmental risk layer, and a resource adaptation layer. The system extracts key indicators from the feature vectors, such as body temperature, respiratory rate, and absolute flood diffusion rate, to calculate a quantified comprehensive vital signs score and an environmental risk assessment value. Simultaneously, it combines the current number of drone nodes in each department, material reserves, and task weights to derive a functional relevance score. The dynamic priority calculation model calls a pre-defined priority quantification scoring function matrix and incorporates a time-series analysis method with a dynamic forgetting factor to dynamically weigh various parallel rescue tasks in real time, outputting a comprehensive priority score for each current parallel rescue task. This comprehensive priority score reflects the urgency of the disaster site and the necessity of task handling in real time.

[0023] After obtaining the score, step S3 is executed to monitor the comprehensive priority score in real time. When the system detects that the comprehensive priority score is greater than the preset critical assessment threshold, such as when a major drowning risk or signs of dam collapse are detected, the system will automatically trigger an emergency access smart contract deployed at the bottom layer of the distributed blockchain network. The distributed blockchain network is jointly maintained by the command center, edge servers of various rescue departments, and drone nodes. After the smart contract is triggered, the system will automatically generate a dynamic preemption instruction. This dynamic preemption instruction not only includes a temporary access permission request for the target's restricted airspace, but also carries a validity logic certificate generated based on a zero-knowledge concise non-interactive knowledge demonstration protocol. This embodiment can ensure that, without disclosing the precise privacy coordinates of the trapped person, high-priority permissions are granted to the requesting drone node through cryptographic verification by the blockchain consensus node cluster. This achieves the legalization and dynamic preemption of cross-departmental resources at both institutional and technical levels, ensuring the absolute priority execution right of critical tasks.

[0024] Step S4 is executed, entering the core game-theoretic scheduling phase. The system uses the dynamic preemption command as an initialization input variable and loads it into the game-theoretic collaborative scheduling model within the blockchain network architecture. Multi-stage game-theoretic calculations based on dynamic decision trees are performed among drone nodes from multiple departments. Specifically, the system constructs a Stackelberg game model, where the command center, as the leader node, proposes a global resource allocation objective function to maximize overall system benefits and minimize departmental conflict costs. Each department's drone node, as a follower, makes an optimal response based on its own remaining battery energy consumption physical constraints. Through repeated iterations and convergence calculations of the multi-stage game, the system outputs the corresponding global resource allocation results and locally optimal path planning results. The locally optimal path planning results include precise three-dimensional space avoidance strategies and airspace release commands, ensuring that no spatial physical collisions occur when drone nodes execute preemption tasks. All generated scheduling decisions, allocation results, and path data are encrypted and permanently written into the blockchain ledger as evidence, ensuring the transparency and traceability of the entire scheduling process.

[0025] Furthermore, to address the extreme situation of potential wide-area communication link paralysis at flood disaster sites, the system monitors the network connectivity status between drone nodes in various departments and the command center node in real time. Once the connectivity status is determined to be below a threshold, the underlying communication hardware module of the corresponding drone node will be forcibly switched to an ultra-wideband (UWB) communication channel. In this state, neighboring drone node groups will autonomously initiate a master node election based on a distributed fault-tolerant consensus protocol, generating a temporary master control scheduling node with temporary decision-making authority. This temporary master control scheduling node summarizes the physical status information of each device in the local area, performs offline disaster recovery decision calculations, regenerates the task priority order table for the current offline environment, and guides the drone group to perform tasks along the emergency flight corridor with a preset safety boundary. After the wide-area communication link is restored, the system uploads the incremental synchronization data packets generated during the offline period to the distributed blockchain network through lossless compression encoding, and performs spatiotemporal consistency comparison of historical data through a conflict verification smart contract, ensuring seamless connection and data integrity between offline and online scheduling logic.

[0026] After the scheduling cycle ends, the system enters a closed-loop optimization phase, where it invokes a federated learning model aggregation framework to update the underlying scheduling model algorithm parameters distributed across various departmental nodes. To protect inter-departmental trade secrets and sensitive rescue information, the system superimposes Laplace distribution statistical noise when outputting the initial gradient parameter matrix and performs global differential privacy protection processing. The system can achieve iterative evolution of the game theory collaborative scheduling model and the dynamic priority calculation model without acquiring the original private data from each department, improving the system's robustness and intelligence in dealing with complex and unknown disaster environments. Through the coordinated efforts of the above steps, this invention constructs a UAV collaborative scheduling system that possesses both strong central command efficiency and distributed self-healing capabilities.

[0027] This system adopts a layered design, specifically divided into a data acquisition layer, a decision-making core layer, a scheduling and execution layer, an emergency support layer, a security protection layer, and a continuous optimization layer. A schematic diagram of the data acquisition layer is attached. Figure 8 As shown in the attached diagram, the core decision-making layer is... Figure 9 The schematic diagram of the scheduling execution layer is shown in the attached figure. Figure 10 The diagram of the emergency support layer is shown in the attached figure. Figure 11 The schematic diagram of the safety protection layer is shown in the attached figure. Figure 12 The diagrams shown below, along with the schematics of the continuous optimization layer, are attached. Figure 13 As shown, each layer interacts with data and coordinates control through a blockchain network and smart contracts, ensuring the efficient operation of the system.

[0028] 1. Real-time disaster data acquisition and preprocessing includes multi-source data access and dynamic classification and feature extraction.

[0029] 1.1 Multi-source data access includes sensor data acquisition and data standardization processing.

[0030] 1.1.1 Sensor data acquisition refers to the use of infrared / millimeter-wave radar by UAVs to acquire heat maps, location coordinates and vital signs of trapped individuals; hydrological sensors to collect data on water level, flow velocity and flood spread in real time; and meteorological modules to access forecast information such as rainfall, wind speed and lightning warnings for the next n hours.

[0031] 1.1.2 Data standardization processing refers to the normalization of vital sign data into an urgency level, ranging from 0 to 1, with a value above 0.8 indicating a critical level; hydrological / meteorological data are converted into influencing factors, such as marking water levels exceeding the warning line by 10% as high risk. All the above information generates a unified format timestamp and is associated with all sensor data.

[0032] 1.2 Dynamic grading and feature extraction include the grading of trapped individuals and the encoding of environmental features.

[0033] 1.2.1 The classification of trapped persons is based on vital signs, such as body temperature <35℃ or >40℃, and respiratory rate <10 breaths / minute or >30 breaths / minute. The above situations are marked as emergency.

[0034] 1.2.2 Environmental feature coding refers to converting flood diffusion speed (m / min), rainfall (mm / h), and UAV remaining battery power (%) into binary feature vectors.

[0035] 2. Dynamic priority calculation and decision-making includes a multi-level evaluation system and dynamic weight calculation.

[0036] 2.1 The multi-level evaluation system includes the target layer, the criteria layer, and the indicator layer.

[0037] 2.1.1 Objective Level. The core objective of the assessment system is to complete rescue missions in the shortest time and at the lowest cost during flood relief, maximize rescue efficiency, ensure the safety of those trapped, and reduce property damage.

[0038] 2.1.2 Criterion Layer. The three dimensions of the criterion layer are task urgency, departmental functional relevance, and resource consumption.

[0039] Task urgency: Assessment is based on the vital signs data of the trapped individuals, such as body temperature, respiratory rate, and blood oxygen saturation. The normalized urgency level of vital signs is a crucial metric: a value above 0.8 indicates critical urgency, corresponding to extremely high urgency; a value between 0.6 and 0.8 indicates high urgency; a value between 0.3 and 0.6 indicates moderate urgency; and a value below 0.3 indicates low urgency. The urgency assessment is also dynamically adjusted based on the environmental danger level of the trapped individuals, such as whether they are in the core flood-affected area or whether there are dangerous buildings nearby.

[0040] Departmental Functional Relevance: Different departments have different functions and varying degrees of importance in flood relief. The fire department plays a crucial role in rescuing trapped people and addressing fire hazards; the medical department is indispensable in treating the injured. The degree of relevance between each department's functions and the specific needs of the rescue mission should be determined. For example, in situations where a large number of injured people require urgent treatment, the relevance of the medical department's functions is significantly increased.

[0041] Resource Consumption: A comprehensive assessment of resource consumption during mission execution, including drone power consumption, material losses, and equipment wear and tear. Accurate calculation of the percentage of total drone power consumed during the current mission, and estimation of the required materials and equipment wear and tear to complete the mission. For missions with excessively high resource consumption and relatively low urgency, their priority is appropriately reduced; conversely, for urgent missions with manageable resource consumption, their priority is maintained or increased.

[0042] 2.1.3 Indicator Layer. The indicator layer includes vital sign indicators, environmental factor indicators, and departmental resource reserve indicators.

[0043] Vital signs indicators: In addition to basic vital signs, indicators such as heart rate variability and blood pressure are included to form a more comprehensive vital signs assessment system. Reasonable threshold ranges are set for each indicator. Based on the degree to which the vital signs of the trapped person deviate from the normal range, a weighted comprehensive vital signs score is calculated, serving as a key quantitative basis for measuring the urgency of the mission.

[0044] Let the set of vital signs indicators of the trapped person be... The weights include heart rate, blood oxygen, etc., and the weight of each indicator is w. i The degree of deviation from the normal range is The formula for the overall score is as follows:

[0045] Environmental factors: A thorough analysis of the impact of flood spread rate, water level trends, and meteorological conditions (such as rainfall, wind speed, and lightning intensity) on rescue operations is conducted. Historical and real-time monitoring data, combined with Geographic Information System (GIS) technology, are used to establish an environmental factor assessment model. For example, if the flood spread rate exceeds a certain threshold and the water level continues to rise, accompanied by heavy rainfall and lightning activity, the difficulty and risk of rescue operations will significantly increase. In such cases, the priorities of relevant tasks need to be reassessed and adjusted.

[0046]

[0047] in, This is the flood spread rate, measured in m / min; It is the critical diffusion rate, measured in m / min; This is the real-time water level, measured in meters (m). This is the warning water level, measured in meters (m). This is the lightning intensity index, which is dimensionless data; the weighting coefficients are respectively... , , ,and .

[0048] Departmental resource reserve indicators: This includes statistics on the number of drones owned by each department, their performance parameters (endurance, payload capacity, etc.), material reserves (medical supplies, rescue tools, etc.), and the availability of professional personnel. Based on the matching degree between departmental resource reserves and current mission requirements, corresponding quantitative evaluation standards will be developed. If a department's drones have strong endurance and sufficient material reserves, enabling it to better handle long-distance, high-difficulty rescue missions, then the department's mission priority can be appropriately increased when performing such missions.

[0049] 2.2 Dynamic Weight Calculation. Dynamic weight calculation is based on time-series weight updates, feedback adjustment mechanisms, and machine learning algorithms to optimize weights.

[0050] 2.2.1 Time-Series-Based Weight Update: Time-series analysis is used to analyze historical task data and real-time collected data. The weights of factors at each criterion level are dynamically adjusted based on the changing trends of task urgency, departmental functional relevance, and resource consumption over time. For example, in the early stages of a flood, the focus of rescue efforts is on searching for and rescuing trapped individuals, so the weight of task urgency can be appropriately increased; as the rescue progresses, the importance of medical treatment and resource allocation increases, and the weight of departmental functional relevance is adjusted accordingly.

[0051] Let the criterion layer weight vector be at time t. W(t) = [W emerg (t),W func (t),W cost (t)]), Dynamic adjustment via exponential smoothing: W(t) = λ W(t-1)+(1-λ) D(t) Where D(t) is the urgency / relevance / cost assessment vector at the current moment, λ is the forgetting factor, and λ∈[0,1].

[0052] 2.2.2 Feedback and Adjustment Mechanism: The weights are dynamically adjusted based on the execution results of the rescue mission and real-time feedback information. If a mission is executed according to its current weighted priority, but resource consumption far exceeds expectations or the urgency of the mission changes (e.g., the vital signs of the trapped individuals deteriorate), the weights are adjusted promptly, and the mission priority is reassessed. Simultaneously, the mission execution results are fed back into the evaluation system as an important basis for subsequent weight adjustments, forming a closed-loop dynamic adjustment mechanism.

[0053] 2.2.3 Machine Learning Algorithm Optimization of Weights: Machine learning algorithms are introduced to learn and train on a large amount of historical and real-time data. The algorithms automatically learn the complex relationships and weight allocations between various factors, continuously optimizing the weight calculation model. For example, neural networks are used to learn from task data under different flood scenarios, automatically adjusting the weights of task urgency, departmental functional relevance, and resource consumption to improve the accuracy and adaptability of priority assessment.

[0054] The formula for calculating task priority is as follows:

[0055] in, As a weight for urgency, The weight of functional relevance, The weights are the resource consumption weights, and the sum of the three weights is 1. Assess the overall health of the organism. For the degree of relevance of departmental functions, This refers to the resource consumption of the task. This represents the maximum resource carrying capacity threshold for a single-unit drone.

[0056] 3. Blockchain-Game Theory Coordinated Scheduling and Execution, including automatic execution of smart contracts and collaborative optimization of game theory models.

[0057] 3.1 Smart contracts automatically trigger dynamic permission allocation by analyzing sensor data in real time, such as vital signs ≥0.8 or water levels exceeding the warning line by 10%. For example, when an emergency task is detected, the system immediately freezes the airspace occupied by low-priority drones, broadcasts a replanning instruction to relevant nodes via PBFT consensus, and simultaneously updates the global task priority table on the blockchain. After receiving the instruction, the drone uses its onboard edge computing module to verify the digital signature and execute detour path planning, while releasing airspace resources. If multiple departments compete for the same airspace, the system arbitrates based on millisecond-level timestamps, prioritizing emergency tasks and encrypting and uploading conflict logs to the blockchain for post-event traceability.

[0058] 3.2 Game theory models for collaborative optimization include extended game theory and Stackelberg game theory.

[0059] 3.2.1 The extended game model models the UAV path planning as a multi-stage dynamic decision tree. The nodes define the environmental state, which includes the flood spread rate and the remaining power. The edges map the actions. The subgame equilibrium is solved by inverse induction to ensure the shortest path selection under sudden obstacles.

[0060] 3.2.2 In the Stackelberg game, the command center, as the leader, optimizes the objective function of global resource allocation based on departmental weights and task urgency.

[0061]

[0062] in, The value of is between (0,1), and i can be any integer between 1 and n.

[0063] Indicates the benefits of resource allocation. This indicates the cost of task conflict.

[0064] when =1 indicates task assignment; When =0, it means no task is assigned.

[0065] Assigning weights to departmental functions, such as the medical department ( =0.2) The urgency level of the task, with a value range of (0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 1, 1, 1, 1, 2 ...2, 1, 2, 1, 2, 1 1) Resource conflict cost between tasks (i) and (j) Conflict penalty coefficient Followers (department drones) generate locally optimal paths based on the leader's strategy, combining the A* algorithm with energy constraints (such as avoiding flood whirlpools), and dynamically release redundant resources (such as airspace and batteries). Smart contracts call the game model in real time via API to update environmental parameters (wind speed, water level) and drone status (battery level, coordinates), and the blockchain consensus synchronizes the calculation results to ensure global consistency of the scheduling scheme.

[0066] Given a task allocation scheme (x), solve the local path planning problem:

[0067] Constraints:

[0068] Where R thresh The preset environmental risk threshold is η, where η is the energy consumption penalty coefficient.

[0069] 4. Disaster recovery and data synchronization protection, including emergency handling of communication interruptions and data synchronization and conflict arbitration.

[0070] 4.1 Emergency Handling of Communication Interruptions. In flood relief scenarios, communication networks often experience regional or even global outages due to base station damage, electromagnetic interference, or extreme weather. To address such extreme situations, the system employs a multi-layered disaster recovery mechanism. Each drone carries a lightweight blockchain node, locally caching the airspace scheduling records, task priority allocation tables, and departmental operation logs for the most recent 10 minutes, ensuring basic scheduling functions can be maintained based on local data even during network outages. When communication is completely interrupted, the drone broadcasts its status (such as coordinates, remaining battery power, and task progress) to neighboring drones via Ultra Wideband (UWB) technology, achieving local coordination based on an improved Paxos consensus protocol. For example, if five drones in a certain area simultaneously lose cloud connectivity, they will autonomously elect a temporary master node, deciding on detour routes or task priority adjustments through majority voting, avoiding group chaos caused by single-point decisions. Furthermore, the system pre-defines emergency flight corridors—low-risk paths along the flood safety line—allowing drones to automatically cruise according to preset rules during communication interruptions, prioritizing life-saving search and rescue missions to minimize rescue delays. When the signal is restored, the drone will incrementally synchronize the locally cached data to the blockchain network. The smart contract automatically verifies the timestamp and operation hash. Conflicting data (such as multiple drones reporting the same airspace occupation) is arbitrated according to the time priority principle to ensure the consistency of the global state.

[0071] 4.2 Data Synchronization and Conflict Arbitration. Under normal communication conditions, the drone uploads compressed data packets to the blockchain node every 30 seconds via the LoRaWAN network. These packets contain encrypted GPS coordinates, sensor readings (such as infrared image feature values), mission execution status (such as material delivery completion rate), and energy consumption statistics. Before uploading, the data packets are preprocessed by the onboard edge computing unit to extract key features (such as the hash value of the trapped drone's coordinates) and discard redundant information (such as background terrain data) to reduce transmission load. After receiving the data, the blockchain network performs multi-level verification via a smart contract: first, it compares the coordinate hashes reported by neighboring drones; if the deviation exceeds 50 meters (possibly due to positioning drift or malicious forgery), an anomaly alarm is triggered and the node's operating permissions are frozen; second, Merkle root verification ensures data integrity and prevents tampering during transmission. For data that needs to be shared across departments (such as disaster heat maps), the system uses a federated learning framework. After training, each local model only uploads encrypted gradient parameters, which are then decrypted by the aggregation server to update the global model, avoiding the risk of original data leakage. In scenarios involving communication fluctuations or partial node offline situations, data synchronization employs an eventual consistency strategy, allowing for temporary data discrepancies. Once the network is restored, these discrepancies are automatically merged through version number conflict detection (Version Vector), prioritizing the retention of operation records from higher-priority departments. For example, if both medical and firefighting drones modify the same airspace status during a network outage, the system automatically determines the final valid operation based on task urgency and marks the conflict log for future auditing.

[0072] 5. Security and privacy protection mechanisms, including zero-knowledge proofs and differential privacy protection.

[0073] 5.1 Zero-Knowledge Proof. In flood relief scenarios, the coordinated dispatch of drones involves sensitive data from multiple departments (such as the location of trapped individuals, rescue mission priorities, and departmental resource allocation), requiring assurance of data security and privacy. For example, when a Red Cross drone needs to verify its identity to obtain high-priority airspace access, traditional identity authentication requires submitting the precise coordinates of the trapped individual, posing a risk of privacy leakage. Therefore, we choose to adopt the zk-SNARKs (Zero-Knowledge Concise Non-Interactive Knowledge Proof) protocol. When a drone generates identity proof, it only needs to submit a "validity proof" (such as a departmental qualification hash) to the blockchain, without revealing the coordinates of the trapped individual. Verifying parties (such as other departmental nodes) can confirm the legitimacy of the identity but cannot deduce sensitive information.

[0074] 5.2 Differential Privacy Protection. When multiple departments share disaster heat maps, it is necessary to prevent the inference of task details for a particular department through data aggregation, such as inferring medical resource allocation from the number of trapped people in a certain area. Therefore, before aggregating local model parameters in federated learning, Laplace noise is added to the gradient data of each department to ensure that the impact of a single data point on the global model is controllable.

[0075] 6. System closed-loop and continuous optimization, including blockchain evidence storage analysis and dynamic model updates.

[0076] 6.1 Blockchain Evidence Storage and Analysis. All drone lifecycle data (trajectory coordinate hashes, mission status codes, energy consumption logs) are encrypted and stored on the blockchain, forming an immutable evidence storage chain. After a rescue operation, the system calculates core indicators including average response time (≤3 minutes), mission completion rate (≥95%), and conflict airspace hotspots (e.g., 5 preemption attempts triggered in a certain area within 10 minutes), generating heatmaps to pinpoint bottlenecks and proposing optimization solutions.

[0077] 6.2 Dynamic Model Update. Federated learning is used to aggregate local model parameters from multiple departments (such as game weights and path safety thresholds). Laplace noise (ε=0.1, sensitivity Δf=0.05) is added before gradient aggregation to achieve differential privacy and prevent back-calculation of departmental task details. The system has a built-in disaster scenario library (such as flood spread rate grading: 1-5m / min is low risk, >5m / min is high risk), and reinforcement learning (Q-learning reward function: +10 for task completion, -20 for power depletion) is used to train the model to dynamically adapt to extreme environments. If the failure rate of a task in a certain scenario exceeds 15% for three consecutive scheduling cycles, the model is triggered to roll back to the historical best version and incremental retraining is started (iterations ≥1000). The optimized parameters are pushed to the decision layer via smart contracts, and the priority calculation logic is updated in real time, forming a closed-loop process of data collection, strategy execution, effect evaluation, and parameter iteration.

[0078] Federated learning gradient aggregation (differential privacy protection): Let g be the local model gradient of the k-th department. k Global aggregation after adding noise:

[0079] Among them, g k For the local model gradient of the k-th department, It is gradient sensitivity. It's a privacy budget. It is the Laplace noise generation function, and K is the total number of departments participating in federated learning.

[0080] Reinforcement learning reward function (Q-learning)

[0081] in, For learning rate, As a discount factor, and These represent the current state and the next state, respectively. and These are the current action and the candidate action, respectively.

[0082] The technical solution in this embodiment achieves precise perception and intelligent classification of rescue missions through standardized processing and dynamic priority evaluation of multi-source sensor data, thereby improving the accuracy of emergency decision-making. Relying on the smart contract mechanism of distributed blockchain and the game theory collaborative scheduling model, it solves the resource competition problem of cross-departmental drone clusters in dynamic environments, and ensures the optimal allocation of global resources and the planning of local paths through multi-stage game theory. At the same time, the encrypted on-chain evidence storage mechanism ensures the transparency and traceability of the scheduling process, and enhances the response speed, collaborative efficiency and safety reliability of rescue operations in complex disaster scenarios.

[0083] Preferred, such as Figure 2 As shown, in step S4, the multi-stage game-theoretic calculation based on dynamic decision trees is performed among multiple departmental drone nodes, outputting the corresponding global resource allocation results and local optimal path planning results, and then encrypted and stored on the blockchain, including: S41, the command center node, which acts as the leader of the game, receives the dynamic preemption instruction and constructs a global resource allocation objective function that maximizes the benefits of system resource allocation and minimizes the cost of inter-departmental task conflicts based on the preset departmental function weight index and task urgency variable. S42, the global resource allocation objective function is broadcast to the drone nodes of each department as game followers, driving the drone nodes of each department to generate preliminary local obstacle avoidance paths and corresponding estimated resource consumption data in combination with their own battery remaining energy consumption physical constraints. S43, the collaborative scheduling model is called in real time through the application programming interface of the emergency access smart contract to obtain the estimated resource consumption data and real-time environmental parameters. Under the convergence condition of satisfying the preset conflict penalty coefficient, the Stackelberg game equilibrium solution of the global resource allocation objective function is solved to generate the local optimal path planning result. The local optimal path planning result includes the avoidance and detour strategy of three-dimensional spatial coordinates and the airspace release instruction.

[0084] The core of step S4, which involves performing multi-stage game-theoretic calculations based on dynamic decision trees among drone nodes in multiple departments, lies in establishing a hierarchical and iterative decision-making framework to solve the resource competition and spatiotemporal conflict problems of drones across departments when performing high-priority preemptive tasks.

[0085] In step S41, the command center node, acting as the game leader, receives the dynamic preemption command generated in the preceding steps. As the aggregator of global information, the command center node can retrieve pre-set departmental function weight indicators and task urgency variables in real time. These departmental function weight indicators include the functional priority coefficients of different departments such as fire fighting, medical services, and security in specific disaster scenarios. Through mathematical modeling, the command center node constructs a global resource allocation objective function that maximizes the system's resource allocation benefits while minimizing the costs of inter-departmental task conflicts. This global resource allocation objective function not only considers the execution efficiency of the current preemption task but also takes into account the overall scheduling stability of the system and the continuity of other ongoing parallel rescue tasks, thus establishing the optimal direction for resource scheduling and the initial boundary of the game at a macro level.

[0086] In step S42, the system distributes the constructed global resource allocation objective function in real time to the drone nodes of each department, which act as followers in the game, through the broadcast mechanism of the distributed blockchain network. Upon receiving the global guidance signal, each drone node makes a secondary decision based on its real-time status detected by its onboard edge computing unit. Specifically, each drone node must strictly adhere to its own physical constraints regarding remaining battery power consumption and calculate the optimal flight range supported by its current power and load conditions. Based on this constraint, each node uses a dynamic decision tree algorithm to optimize its local path, generating a preliminary local obstacle avoidance path, and simultaneously calculates the estimated resource consumption data required to complete the path, such as expected power consumption and airspace occupation time, thereby achieving an organic combination of centralized guidance and decentralized response.

[0087] The system executes step S43, invoking the collaborative scheduling model in real time via the application programming interface of the emergency access smart contract. The system uses the collected estimated resource consumption data of each node and real-time environmental parameters as input variables, performing deep computation within a game theory framework. These real-time environmental parameters include gust wind speed and dynamic coordinates of obstacles. To ensure rapid convergence to the optimal solution, the system introduces a conflict penalty coefficient, imposing a high penalty on any decision path that might lead to physical collisions in the three-dimensional airspace or resource over-allocation. Under preset convergence conditions, the model iterates multiple times to find the Stackelberg game equilibrium solution of the global resource allocation objective function. This Stackelberg game equilibrium solution represents the optimal cooperative state of the drone nodes in each department under the current complex constraints, generating a locally optimal path planning result. This locally optimal path planning result includes avoidance and detour strategies in three-dimensional spatial coordinates and airspace release commands, guiding the drones to perform refined maneuvers in the three-dimensional airspace. The system encrypts the above game-solving process, global allocation decision and path planning results, and solidifies them as key evidence in the blockchain ledger, thus completing encrypted on-chain evidence storage, thereby ensuring the legality, immutability and full traceability of the entire dynamic permission allocation and collaborative scheduling process.

[0088] Preferred, such as Figure 3 As shown, in step S2, the disaster environment and personnel feature vectors are input into a preset dynamic priority calculation model based on a multi-level evaluation system for quantification, and a comprehensive priority score for each current parallel rescue task is output, including: S21, analyze and extract the body temperature index features, respiratory rate features and absolute flood diffusion speed features contained in the disaster environment and personnel feature vector, perform weighted summation calculation based on deviation degree and physical critical threshold comparison operation respectively, and calculate the quantitative vital signs comprehensive score and environmental risk assessment value. S22, obtain the real-time matching degree variable of the disaster environment and personnel feature vector with respect to the number of drones owned by the current department, the reserve capacity of special materials and the weight of the task type, and calculate the functional correlation score of each participating rescue department; S23, substitute the comprehensive vital signs score, the environmental risk assessment value, and the functional correlation score of each participating rescue department into the preset priority quantification scoring function matrix, and call the time series analysis of the evaluation weight vector distribution in the function matrix with the introduction of dynamic forgetting factor to output the comprehensive priority score.

[0089] In this embodiment, regarding step S2 of the blockchain-game theory-based UAV collaborative scheduling dynamic permission allocation method, the system achieves quantitative and accurate analysis of complex rescue situations through a dynamic priority calculation model based on a multi-level evaluation system.

[0090] In step S21, the system extracts physiological and environmental parameters from the disaster environment and personnel feature vector generated in step S1 using a preset feature analysis module. These parameters include body temperature characteristics, respiratory rate characteristics, and absolute flood diffusion velocity characteristics. Weighted summation calculations based on deviation are then performed on these physiological and environmental parameters. By measuring the degree of deviation between real-time data and safety benchmark values, the system reflects the risk coefficient of the trapped individual and the environmental state. Simultaneously, the system performs a physical critical threshold comparison calculation, comparing the real-time parameters with pre-stored physiological limits or disaster evolution thresholds. Through the comprehensive processing of the above algorithms, a quantified comprehensive vital sign score and environmental risk assessment value are calculated.

[0091] To ensure a high degree of alignment between rescue resources and mission requirements, this embodiment executes step S22. The system retrieves dynamic information regarding resource distribution and mission attributes from the disaster environment and personnel feature vectors in real time, obtaining real-time matching variables for the number of drones currently in use by the relevant department, the capacity of special material reserves, and the weight of mission types. The capacity of special material reserves includes, but is not limited to, items such as lifebuoys, first-aid kits, and communication relay modules. The system establishes a multi-dimensional resource efficiency evaluation matrix. By calculating the matching coefficient between the existing strength of each rescue department and the current mission requirements, it derives a functional correlation score for each participating rescue department. This ensures that during the allocation of authority, departments with stronger specialized rescue capabilities and more sufficient equipment reserves receive higher decision-making reference weights, thereby optimizing cross-departmental resource allocation efficiency at the macro-level scheduling level.

[0092] The system executes step S23, taking the comprehensive vital signs score, the environmental risk assessment value, and the functional correlation score corresponding to each participating rescue department calculated in the previous steps as core input parameters, and substituting them into a preset priority quantification scoring function matrix. To address the rapidly changing timeliness requirements at disaster sites, the system introduces a time-series analysis method with a dynamic forgetting factor during the calculation process. By weighted attenuation of historical dispatch data, the system automatically adjusts the distribution of evaluation weight vectors in the function matrix, ensuring the model focuses on reflecting the latest changes in the on-site situation and effectively filtering out outdated noise information. Through high-dimensional nonlinear mapping operations of this matrix, the system outputs a comprehensive priority score for each current parallel rescue task.

[0093] Preferred, such as Figure 4As shown, in step S4, after outputting the corresponding global resource allocation result and local optimal path planning result, and encrypting and storing them on the blockchain, the process further includes: S5. Monitor the connectivity status of the wide area communication link between the drone nodes of each department and the command center node in the area in real time. When the connectivity status parameter is detected to be lower than the preset network connectivity judgment threshold and the wide area communication network is determined to be disconnected and paralyzed, switch the underlying communication hardware module of the corresponding drone node to the ultra-wideband technology communication channel and broadcast the recent airspace scheduling records and equipment physical status information cached locally to the outside. S6, receive the recent airspace scheduling records and equipment physical status information broadcast by neighboring UAV nodes through the ultra-wideband technology communication channel, wherein the neighboring UAV nodes are in the same spatial physical range, and autonomously initiate a master node election procedure among multiple interconnected UAV node groups based on an improved distributed fault-tolerant consensus protocol, and generate a temporary master control scheduling node with temporary decision-making authority. S7, the temporary master control scheduling node elected performs offline disaster recovery decision calculation based on majority voting mechanism according to the aggregated physical status information of the equipment, reorders and generates the task priority order table in the current offline local area network environment and issues corresponding flight attitude physical control commands to guide the drone group in the fault area to perform low-risk life search and rescue cruise missions along the emergency flight corridor of the pre-set safety boundary.

[0094] This invention provides a blockchain-game theory-based dynamic permission allocation method for UAV collaborative scheduling. After completing step S4, which outputs the corresponding global resource allocation results and local optimal path planning results and performs encrypted on-chain storage, a highly robust disaster recovery mechanism is further constructed to cope with the potential wide-area communication network paralysis at flood sites. Specifically, in step S5, the system uses an airborne edge computing unit to monitor the connectivity status of the wide-area communication links between the UAV nodes of each department and the command center nodes in their respective areas in real time. The wide-area communication links typically use 4G / 5G or satellite communication protocols for data transmission. The system obtains real-time connectivity parameters by calculating key indicators such as link packet loss rate, latency, and signal strength. Once the connectivity parameters are detected to be lower than the preset network connectivity judgment threshold, and it is determined that the wide-area communication network has been disconnected and paralyzed, the system will immediately activate the underlying emergency plan and switch the underlying communication hardware module of the corresponding UAV node to an ultra-wideband (UWB) communication channel. In this UWB communication channel mode, the UAV node will broadcast locally cached recent airspace scheduling records and equipment physical status information. By forcibly switching the underlying hardware, it is ensured that even in extreme scenarios where centralized command signals are completely lost, the drone swarm in a local area still has basic data exchange capabilities.

[0095] The system executes step S6 to establish a local autonomous negotiation mechanism. Neighboring UAV nodes within the same physical spatial range receive recent airspace scheduling records and equipment physical status information broadcast by other nodes via the ultra-wideband communication channel. Based on this, the system autonomously initiates a master node election process among multiple interconnected UAV node groups using an improved distributed fault-tolerant consensus protocol. Since the recent airspace scheduling records contain the task permissions and priority status of each node before disconnection, and the equipment physical status information reflects the current remaining power and load, the election process can generate a temporary master scheduling node with temporary decision-making authority based on node performance and task importance weights. This temporary master scheduling node acts as a local temporary command tower, taking over the coordination functions originally undertaken by the command center node, thereby achieving rapid restructuring of the organizational structure in a decentralized environment.

[0096] In step S7, the elected temporary master control scheduling node aggregates the physical status information of the devices within the current local area network through data fusion and performs offline disaster recovery decision calculations based on a majority voting mechanism. During this decision-making process, the temporary master control scheduling node reorders and generates a task priority order table for the current offline local area network environment based on the location of the stranded drones and environmental risks, and issues corresponding flight attitude physical control commands based on this table. To ensure offline flight safety, the system pre-defines an emergency flight corridor with safety boundaries. The temporary master control scheduling node guides the drone swarm in the fault area to strictly follow the emergency flight corridor to perform low-risk life search and rescue patrol missions. This disaster recovery strategy based on "majority voting" effectively filters out potential single-point sensor failure interference, ensuring that the drone swarm can still maintain basic search and rescue functions in a coordinated manner under disconnection conditions, minimizing the risk of secondary accidents.

[0097] Preferred, such as Figure 5 As shown, in step S7, after the drone swarm guiding the fault area performs a low-risk life search and rescue patrol mission along the emergency flight corridor with a pre-defined safety boundary, the process further includes: S8, continuously execute the monitoring program for the connectivity status of the wide area communication link, and when it is confirmed that the connectivity status parameters of the communication link have recovered to the normal working threshold range, package and encapsulate the recent airspace scheduling records and offline task operation execution logs generated in the local memory of the UAV during the wide area communication network disconnection into an incremental synchronization data packet; S9, control the airborne edge computing processing unit installed inside the drone nodes of each department to process the incremental synchronization data packet, extract the key coordinate hash feature value containing task attributes, perform lossless data compression encoding processing, and then send the compressed and encoded incremental synchronization data packet to the data receiving node of the distributed blockchain network through the wide area low power IoT communication protocol; S10, trigger the conflict verification smart contract deployed in the underlying architecture of the distributed blockchain network, compare and verify the spatial physical deviation between the key coordinate hash feature value and the dataset reported by the neighboring system nodes, and when it is determined that there is a cross-departmental collaborative competition conflict in the same three-dimensional physical space, the valid operation is determined by the dual standard of the priority principle of the timestamp sequence of the network global clock and the urgency level of the vital signs of the trapped person, and the conflict decision log containing timestamp information is solidified and written into the block body of the block storage layer.

[0098] After completing the low-risk life search and rescue patrol mission along the emergency flight corridor with a pre-set safety boundary as described in step S7, the system enters the data backhaul and conflict resolution stage in order to achieve efficient alignment and consistency verification between offline operation data and the global scheduling system.

[0099] In step S8, the system continuously executes the monitoring program for the connectivity status of the wide-area communication link through the airborne communication module. As the flood-affected environment gradually stabilizes and communication base stations or relay equipment resume operation, the system will capture key parameters such as signal strength, bandwidth, and packet loss rate of the wide-area communication link in real time. Once it is confirmed that the connectivity parameters of the communication link have recovered to the normal operating threshold range, the system determines that the physical conditions for large-scale data interaction with the distributed blockchain network are now met. At this time, the system retrieves and integrates the recent airspace scheduling records and offline task operation execution logs generated in the UAV's local memory during the wide-area communication network disconnection period, and packages them into an incremental synchronization data packet. This incremental synchronization data packet fully records the flight trajectory, permission changes, and task execution details of each UAV node during the offline period.

[0100] The system executes step S9, entering the data feature extraction and transmission optimization stage. In this embodiment, the onboard edge computing processing unit, mounted inside the drone nodes of each department, performs deep analysis of the incremental synchronization data packets. To ensure verification efficiency and security under limited bandwidth conditions, the onboard edge computing processing unit does not directly transmit the original logs. Instead, it extracts key coordinate hash feature values ​​containing task attributes using a hash algorithm. These key coordinate hash feature values ​​can characterize the drone's occupancy status in the three-dimensional airspace with a minimal amount of data. Simultaneously, the system performs lossless data compression encoding to further reduce the proportion of redundant information. The system utilizes wide-area low-power IoT communication protocols, such as NB-IoT or LoRaWAN, which have high penetration capabilities, to send the compressed incremental synchronization data packets to the data receiving nodes of the distributed blockchain network. Through edge-side preprocessing and protocol optimization, the reliable transmission problem under network load fluctuations during the initial stage of disconnection recovery is solved, ensuring the timely on-chain uploading of key scheduling evidence.

[0101] The system executes step S10, using blockchain technology to automate compliance auditing and conflict resolution for offline decisions. When the data receiving node receives the incremental synchronization data packet, the system automatically triggers a conflict verification smart contract deployed in the underlying architecture of the distributed blockchain network. This smart contract performs reverse verification on the hash feature value of the key coordinates and compares it with the dataset reported by neighboring system nodes to verify spatial physical deviations. This identifies whether uncoordinated conflicts or track crossings occurred between drones from different departments in the same three-dimensional physical airspace during the wide-area communication network outage. When determining if a cross-departmental collaborative conflict exists within the same three-dimensional physical airspace, to maintain the rigor of scheduling, the smart contract enforces a dual standard of prioritizing the timestamp sequence of the network's global clock and the urgency level of the trapped individuals' vital signs. Specifically, the system prioritizes operations with earlier timestamps or involving trapped individuals with higher urgency, using this to determine the validity of the scheduling. The system solidifies the generated conflict resolution log, containing timestamp information, into the block body of the block storage layer, completing the evidence preservation of the rescue process. Through the coordinated operation of S8 to S10, it can be ensured that the drone swarm can still maintain the closed-loop consistency and ownership clarity of the global scheduling logic after experiencing a complete cycle of network interruption and recovery.

[0102] Preferred, such as Figure 6 As shown, in step S3, triggering the emergency access smart contract deployed at the bottom layer of the distributed blockchain network to generate a dynamic preemption instruction includes: S31, when different rescue departments initiate a cross-domain restricted airspace permission application request within the comprehensive priority scoring triggering system, the underlying drone edge computing node that made the application request is controlled to perform cryptographic operations to generate a zero-knowledge concise non-interactive knowledge demonstration protocol validity logic certificate that conceals the geographical coordinates of the real trapped person, and the validity logic certificate is submitted and uploaded to the distributed blockchain network. S32, control the distributed consensus node cluster in the distributed blockchain network, use the matching cryptographic verification algorithm function to review and verify the validity logical certificate, after the review and verification is passed and the node's digital identity is confirmed, grant the corresponding level of high-priority restricted airspace occupation permission to the underlying drone edge computing node that made the application, and generate the dynamic preemption instruction based on the high-priority restricted airspace occupation permission. S33. When the system enters the closed-loop optimization phase after the scheduling cycle ends, it calls the federated learning model aggregation framework to update the underlying scheduling model algorithm parameters of the cross-department local drone nodes. It outputs the initial gradient parameter calculation matrix to each department endpoint through the smart contract, and superimposes Laplace distribution statistical noise into the matrix to generate system scheduling model iteration parameters with the ability to resist model reverse feature back-inference attacks, thereby completing the global differential privacy protection processing.

[0103] Based on the comprehensive priority score calculated in the aforementioned steps, this embodiment implements step S3 to achieve dynamic allocation of permissions and long-term system optimization in emergency situations through the trusted computing mechanism of the blockchain underlying layer.

[0104] In step S31, when the comprehensive priority score triggers a cross-domain restricted airspace permission application request between different rescue departments within the system—for example, a medical support department needs to enter the airway under the jurisdiction of the fire department for the transfer of the wounded—the system controls the underlying drone edge computing node that made the request to perform specific cryptographic operations. In this step, the system introduces a zero-knowledge concise non-interactive knowledge demonstration protocol to generate a valid logical credential that conceals the actual geographical coordinates of the trapped individual. The essence of this valid logical credential design lies in its ability to allow the requesting node to prove, through mathematical logic, that its application meets preset crisis assessment standards without disclosing sensitive privacy coordinates to the outside world or other nodes in the blockchain. The system then submits and uploads the generated valid logical credential to the distributed blockchain network as the legal basis for granting permissions.

[0105] In step S32, the distributed consensus node cluster in the distributed blockchain network invokes the configured cryptographic verification algorithm function to automatically verify the legitimacy of the validity logical credential. After successful verification and identification of the digital identity of the underlying drone edge computing node that submitted the application, the system automatically grants the node high-priority restricted airspace occupancy rights at the corresponding level. Based on these high-priority restricted airspace occupancy rights, the system generates a dynamic preemption command. This dynamic preemption command possesses a mandatory airspace exclusivity attribute, guiding the node to legally enter the restricted area, thereby ensuring the immediacy, compliance, and exclusive execution of extremely urgent rescue operations.

[0106] To ensure the continuous evolution of the underlying scheduling model algorithm stored in each department's nodes, while also considering data security in cross-departmental collaboration, this embodiment executes step S33. When the system enters the closed-loop optimization phase after the scheduling cycle ends, it invokes the federated learning model aggregation framework to update the underlying scheduling model algorithm parameters of the cross-departmental local drone nodes. In this optimization process, the system outputs the initial gradient parameter calculation matrix to the endpoints of each participating department through a smart contract. To prevent malicious attackers from reverse-engineering each department's private rescue strategies or the distribution characteristics of trapped individuals by analyzing gradient change information, the system superimposes Laplace distribution statistical noise into the initial gradient parameter calculation matrix, thereby generating system scheduling model iteration parameters capable of resisting model reverse feature back-inference attacks. This enables global differential privacy protection processing, allowing the system to achieve a synergistic improvement in the performance of the global scheduling model while ensuring that the original data of each participant never leaves its local location.

[0107] In one embodiment, Figure 7 This is a block diagram illustrating a blockchain-game theory-based dynamic permission allocation system for collaborative scheduling of unmanned aerial vehicles (UAVs). For example... Figure 7 The system for collaborative scheduling of unmanned aerial vehicles (UAVs) includes a data acquisition module 71, an output module 72, a generation module 73, and a calculation module 74.

[0108] The acquisition module 71 is used to acquire real-time field observation data obtained by multi-source heterogeneous sensors mounted on the UAV platform, and convert the acquired real-time field observation data into a disaster environment and personnel feature vector containing a unified timestamp and normalized scaling. The output module 72 is used to input the disaster environment and personnel feature vectors into a preset dynamic priority calculation model based on a multi-level evaluation system for quantitative processing, and output a comprehensive priority score for the current parallel rescue tasks. The generation module 73 is used to trigger an emergency access smart contract deployed at the bottom layer of the distributed blockchain network to generate a dynamic preemption instruction when the comprehensive priority score is detected to be greater than the preset critical assessment threshold. The calculation module 74 is used to load the dynamic preemption instruction as an initialization input variable into the game theory collaborative scheduling model in the blockchain network architecture, perform multi-stage game solving calculation based on dynamic decision tree among multiple department drone nodes, output the corresponding global resource allocation results and local optimal path planning results, and perform encrypted on-chain storage.

[0109] The data acquisition module 71, the output module 72, the generation module 73, and the calculation module 74 included in the block diagram of the blockchain-game theory-based UAV collaborative scheduling dynamic permission allocation system are controlled to execute the blockchain-game theory-based UAV collaborative scheduling dynamic permission allocation method described in any of the above embodiments.

[0110] This embodiment also provides a blockchain-game theory-based UAV collaborative scheduling dynamic permission allocation system, including interconnected processors and computer-readable storage media. The computer-readable storage media stores a computer program, which is executed by the processor to implement the steps of the above-described blockchain-game theory-based UAV collaborative scheduling dynamic permission allocation method.

[0111] This embodiment also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the above-described blockchain-game theory-based UAV collaborative scheduling dynamic permission allocation method.

[0112] This embodiment also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the above-described blockchain-game theory-based UAV collaborative scheduling dynamic permission allocation method.

[0113] Compared with the prior art, the advantages of this embodiment are: By standardizing and dynamically prioritizing multi-source sensor data, precise perception and intelligent classification of rescue missions were achieved, improving the accuracy of emergency decision-making. Relying on the smart contract mechanism of distributed blockchain and the game theory collaborative scheduling model, the resource competition problem of cross-departmental drone swarms in dynamic environments was solved. Through multi-stage game theory, the optimal allocation of global resources and the planning of local paths were ensured. At the same time, the encrypted on-chain evidence storage mechanism ensured the transparency and traceability of the scheduling process, enhancing the response speed, collaborative efficiency, and safety and reliability of rescue operations in complex disaster scenarios.

[0114] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-readable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process. Figure 1 One or more processes and / or boxes Figure 1 The functions specified in one or more boxes. These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable apparatus for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0115] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A dynamic permission allocation method for collaborative scheduling of unmanned aerial vehicles (UAVs) based on blockchain and game theory, characterized in that, include: S1, collect real-time field observation data obtained by multi-source heterogeneous sensors mounted on the UAV platform, and convert the collected real-time field observation data into a disaster environment and personnel feature vector containing a unified timestamp and normalized scaling. S2, input the disaster environment and personnel feature vectors into a preset dynamic priority calculation model based on a multi-level evaluation system for quantification, and output a comprehensive priority score for each current parallel rescue task; S3, when the comprehensive priority score is detected to be greater than the preset critical assessment threshold, the emergency access smart contract deployed at the bottom layer of the distributed blockchain network is triggered to generate a dynamic preemption instruction. S4. The dynamic preemption instruction is loaded as an initialization input variable into the game theory collaborative scheduling model in the blockchain network architecture. Multi-stage game solving calculation based on dynamic decision tree is performed among multiple department drone nodes. The corresponding global resource allocation results and local optimal path planning results are output and encrypted and stored on the blockchain.

2. The method for dynamic permission allocation of UAV collaborative scheduling based on blockchain-game theory as described in claim 1, characterized in that, In step S4, the multi-stage game-theoretic calculation based on dynamic decision trees is performed among multiple departmental drone nodes, outputting the corresponding global resource allocation results and local optimal path planning results, and then encrypted and stored on the blockchain, including: S41, the command center node, which acts as the leader of the game, receives the dynamic preemption instruction and constructs a global resource allocation objective function that maximizes the benefits of system resource allocation and minimizes the cost of inter-departmental task conflicts based on the preset departmental function weight index and task urgency variable. S42, the global resource allocation objective function is broadcast to the drone nodes of each department as game followers, driving the drone nodes of each department to generate preliminary local obstacle avoidance paths and corresponding estimated resource consumption data in combination with their own battery remaining energy consumption physical constraints. S43, the collaborative scheduling model is called in real time through the application programming interface of the emergency access smart contract to obtain the estimated resource consumption data and real-time environmental parameters. Under the convergence condition of satisfying the preset conflict penalty coefficient, the Stackelberg game equilibrium solution of the global resource allocation objective function is solved to generate the local optimal path planning result. The local optimal path planning result includes the avoidance and detour strategy of three-dimensional spatial coordinates and the airspace release instruction.

3. The method for dynamic permission allocation of UAV collaborative scheduling based on blockchain-game theory as described in claim 1, characterized in that, In step S2, the disaster environment and personnel feature vectors are input into a preset dynamic priority calculation model based on a multi-level evaluation system for quantification, and a comprehensive priority score for each current parallel rescue task is output, including: S21, analyze and extract the body temperature index features, respiratory rate features and absolute flood diffusion speed features contained in the disaster environment and personnel feature vector, perform weighted summation calculation based on deviation degree and physical critical threshold comparison operation respectively, and calculate the quantitative vital signs comprehensive score and environmental risk assessment value. S22, obtain the real-time matching degree variable of the disaster environment and personnel feature vector with respect to the number of drones owned by the current department, the reserve capacity of special materials and the weight of the task type, and calculate the functional correlation score of each participating rescue department; S23, substitute the comprehensive vital signs score, the environmental risk assessment value, and the functional correlation score of each participating rescue department into the preset priority quantification scoring function matrix, and call the time series analysis of the evaluation weight vector distribution in the function matrix with the introduction of dynamic forgetting factor to output the comprehensive priority score.

4. The method for dynamic permission allocation of UAV collaborative scheduling based on blockchain-game theory as described in claim 1, characterized in that, In step S4, after outputting the corresponding global resource allocation result and local optimal path planning result, and encrypting and storing them on the blockchain, the process further includes: S5. Monitor the connectivity status of the wide area communication link between the drone nodes of each department and the command center node in the area in real time. When the connectivity status parameter is detected to be lower than the preset network connectivity judgment threshold and the wide area communication network is determined to be disconnected and paralyzed, switch the underlying communication hardware module of the corresponding drone node to the ultra-wideband technology communication channel and broadcast the recent airspace scheduling records and equipment physical status information cached locally to the outside. S6, receive the recent airspace scheduling records and equipment physical status information broadcast by neighboring UAV nodes through the ultra-wideband technology communication channel, wherein the neighboring UAV nodes are in the same spatial physical range, and autonomously initiate a master node election procedure among multiple interconnected UAV node groups based on an improved distributed fault-tolerant consensus protocol, and generate a temporary master control scheduling node with temporary decision-making authority. S7, the temporary master control scheduling node elected performs offline disaster recovery decision calculation based on majority voting mechanism according to the aggregated device physical status information, reorders and generates the task priority order table in the current offline local area network environment and issues corresponding flight attitude physical control commands to guide the drone group in the fault area to perform low-risk life search and rescue cruise missions along the emergency flight corridor of the pre-set safety boundary.

5. The method for dynamic permission allocation of UAV collaborative scheduling based on blockchain-game theory as described in claim 4, characterized in that, In step S7, after the drone swarm guiding the fault area performs a low-risk life search and rescue patrol mission along an emergency flight corridor with a pre-defined safety boundary, the process further includes: S8, continuously execute the monitoring program for the connectivity status of the wide area communication link, and when it is confirmed that the connectivity status parameters of the communication link have recovered to the normal working threshold range, package and encapsulate the recent airspace scheduling records and offline task operation execution logs generated in the local memory of the UAV during the wide area communication network disconnection into an incremental synchronization data packet; S9, control the airborne edge computing processing unit installed inside the drone nodes of each department to process the incremental synchronization data packet, extract the key coordinate hash feature value containing task attributes, perform lossless data compression encoding processing, and then send the compressed and encoded incremental synchronization data packet to the data receiving node of the distributed blockchain network through the wide area low power IoT communication protocol; S10, trigger the conflict verification smart contract deployed in the underlying architecture of the distributed blockchain network, compare and verify the spatial physical deviation between the key coordinate hash feature value and the dataset reported by the neighboring system nodes, and when it is determined that there is a cross-departmental collaborative competition conflict in the same three-dimensional physical space, the valid operation is determined by the dual standard of the priority principle of the timestamp sequence of the network global clock and the urgency level of the vital signs of the trapped person, and the conflict decision log containing timestamp information is solidified and written into the block body of the block storage layer.

6. The method for dynamic permission allocation of UAV collaborative scheduling based on blockchain-game theory as described in claim 1, characterized in that, In step S3, triggering the emergency access smart contract deployed at the bottom layer of the distributed blockchain network to generate a dynamic preemption instruction includes: S31, when different rescue departments initiate a cross-domain restricted airspace permission application request within the comprehensive priority scoring triggering system, the underlying drone edge computing node that made the application request is controlled to perform cryptographic operations to generate a zero-knowledge concise non-interactive knowledge demonstration protocol validity logic certificate that conceals the geographical coordinates of the real trapped person, and the validity logic certificate is submitted and uploaded to the distributed blockchain network. S32, control the distributed consensus node cluster in the distributed blockchain network, use the matching cryptographic verification algorithm function to review and verify the validity logical certificate, after the review and verification is passed and the node's digital identity is confirmed, grant the corresponding level of high-priority restricted airspace occupation permission to the underlying drone edge computing node that made the application, and generate the dynamic preemption instruction based on the high-priority restricted airspace occupation permission. S33. When the system enters the closed-loop optimization phase after the scheduling cycle ends, it calls the federated learning model aggregation framework to update the underlying scheduling model algorithm parameters of the cross-department local drone nodes. It outputs the initial gradient parameter calculation matrix to each department endpoint through the smart contract, and superimposes Laplace distribution statistical noise into the matrix to generate system scheduling model iteration parameters with the ability to resist model reverse feature back-inference attacks, thereby completing the global differential privacy protection processing.

7. A blockchain-game theory-based dynamic permission allocation system for collaborative scheduling of unmanned aerial vehicles (UAVs), characterized in that, include: The acquisition module is used to acquire real-time field observation data obtained by multi-source heterogeneous sensors mounted on the UAV platform, and convert the acquired real-time field observation data into disaster environment and personnel feature vectors containing a unified timestamp and normalized scaling. The output module is used to input the disaster environment and personnel feature vectors into a preset dynamic priority calculation model based on a multi-level evaluation system for quantitative processing, and output a comprehensive priority score for the current parallel rescue tasks. The generation module is used to trigger an emergency access smart contract deployed at the bottom layer of the distributed blockchain network to generate a dynamic preemption instruction when the comprehensive priority score is detected to be greater than the preset critical assessment threshold. The calculation module is used to load the dynamic preemption instruction as an initialization input variable into the game theory collaborative scheduling model in the blockchain network architecture, perform multi-stage game solving calculation based on dynamic decision tree among multiple department drone nodes, output the corresponding global resource allocation results and local optimal path planning results, and perform encrypted on-chain storage.

8. The UAV collaborative scheduling dynamic permission allocation system based on blockchain-game theory as described in claim 7, characterized in that: The acquisition module, the output module, the generation module, and the calculation module are controlled to execute the UAV collaborative scheduling dynamic permission allocation method based on blockchain-game theory as described in any one of claims 2 to 6.

9. An electronic device, characterized in that, include: Communication interface, processor, memory; The memory is used to store program instructions, which, when executed by the processor connected to the memory via the communication interface, enable the electronic device to implement the blockchain-game theory-based UAV collaborative scheduling dynamic permission allocation method as described in any one of claims 1 to 6.

10. A computer-readable storage medium having program instructions stored thereon, characterized in that, When the program instructions are executed by the computer, the computer implements the dynamic permission allocation method for UAV collaborative scheduling based on blockchain-game theory as described in any one of claims 1 to 6.

Images