Kubernetes-based unmanned aerial vehicle task scheduling and fault handling system and method

Abstract

The application discloses a UAV task scheduling and fault handling system and method based on Kubernetes, which integrates Kubernetes technology with a UAV system. The system framework of the application is composed of a distributed cloud control center, task UAVs and backup UAVs. The distributed cloud control center is the core of the whole system, mainly responsible for task scheduling, resource management and fault detection. The task UAVs are mainly responsible for task execution, communication and fault reporting. The backup UAVs are always on standby and mainly responsible for task replacement. Through the task scheduling and distribution algorithm of the resource management module, the application can automatically select appropriate UAVs to execute tasks according to task types, priorities and resource requirements and other indicators. This optimized resource utilization method can improve task execution efficiency, reduce idle time of UAVs, and reduce energy consumption and cost.

Description

Technical Field

[0001] This invention belongs to the field of communication technology and relates to a Kubernetes-based drone mission scheduling and fault handling system. Background Technology

[0002] With the development of cloud computing, 5G mobile networks, edge computing, and artificial intelligence technologies, unmanned technology has matured significantly, and drone systems are widely used in various fields, such as border patrol, pesticide application, and terrain surveying. However, traditional drone systems have some technical challenges, such as:

[0003] (1) Real-time communication and data processing schemes between UAVs and cloud systems may suffer from data transmission and processing delays. Furthermore, there is a lack of specific research on fault response and safety scheduling in large-scale environmental monitoring.

[0004] (2) The research on unmanned aerial vehicle (UAV) systems used for urban emergency response has not included specific studies on fault response and safe dispatch, and there is a lack of in-depth discussion on system robustness and security. In addition, research on the collaborative work and task allocation of multiple UAVs in large-scale emergency response is also insufficient.

[0005] (3) The collaborative work and task allocation of large-scale multi-UAV systems have not been fully considered. In addition, insufficient research on real-time communication and data transmission schemes may affect the real-time performance and efficiency of the system.

[0006] (4) The tasks are divided and assigned to different drones in advance, and there is no dynamic scheduling capability. Once a fault occurs or the task changes, manual intervention is required for adjustment. Summary of the Invention

[0007] The purpose of this invention is to provide a Kubernetes-based drone task scheduling and fault handling system, which solves the problems of low task scheduling efficiency and lack of fault handling methods in existing drone systems.

[0008] The first technical solution adopted in this invention is: a Kubernetes-based drone task scheduling and fault handling system, including: a distributed cloud control center, a task drone, and a backup drone; wherein the distributed cloud control center is the central hub of the entire system and is mainly responsible for task scheduling, resource management, and fault detection; the task drone is mainly responsible for task execution, communication, and fault reporting; and the backup drone is always in standby mode and is mainly responsible for task takeover.

[0009] The invention is further characterized by:

[0010] The distributed cloud control center includes:

[0011] Security control module: Utilizes the multi-layered security mechanisms provided by Kubernetes, including authentication, authorization, and network isolation, to ensure that only authenticated users can access the system;

[0012] The communication monitoring module includes two functions: real-time monitoring and resource utilization analysis. The real-time monitoring function monitors the status of each UAV in the task UAV cluster; the resource utilization analysis function obtains the resource consumption of each UAV.

[0013] Resource management module: It utilizes the Kubernetes scheduler to schedule and allocate drone tasks; including dynamic adjustment and task scheduling. Dynamic adjustment achieves load balancing and resource optimization based on the real-time status of the drones. The task scheduling function automatically selects the appropriate drone to execute the task based on the task's attributes and requirements, as well as the drone's available resources.

[0014] The fault handling module includes fault repair and task switching functions. The fault repair function automatically attempts to repair a faulty drone when it is detected. If the fault repair is successful, the system can continue to work normally. The task switching function selects the best backup drone to take over the task when the faulty drone cannot be repaired. It also notifies the resource management module to send the unfinished tasks of the faulty drone to the best backup drone, ensuring seamless task switching and guaranteeing task continuity and completion.

[0015] The organizational structure of the mission drone includes a communication module, a health self-check module, a mission processing module, and a data processing module. The communication module is responsible for communicating with the cloud control center, receiving mission instructions, and sending mission completion signals, ensuring a normal communication connection between the mission drone and the cloud control center. The health self-check module is responsible for periodically checking the status and health of the mission drone, including battery level, sensor status, and communication connectivity, and determining if any faults or anomalies have occurred. The mission processing module is responsible for mission execution and control, controlling the drone to execute mission segments sequentially according to mission requirements and monitoring mission completion. The data processing module is responsible for processing and analyzing the data collected by the mission drone, performing image processing, target recognition, and data compression to extract useful information and optimize data transmission.

[0016] Backup drones are used to take over the work and complete the remaining tasks when the mission drone fails or malfunctions.

[0017] The second technical solution adopted in this invention is: a Kubernetes-based method for drone task scheduling and fault handling, comprising:

[0018] Step 1: System Initialization: Start the distributed cloud control center, mission drone, and backup drone. Each component performs initialization and preparation tasks;

[0019] Step 2: Task Analysis and Division: After receiving a task request, the distributed cloud control center will analyze the task based on its characteristics, priority, and resource requirements, and then divide the task into task segments.

[0020] Step 3: Task scheduling: The distributed cloud control center will monitor the status of the task drones in real time and assign the task segments to the most suitable drones for execution based on the monitoring results;

[0021] Step 4: Fault Detection: The mission drone will periodically perform a health self-check. When a fault or abnormality is detected, the drone will first perform automatic repair. If the repair fails, it will send a fault signal to the cloud.

[0022] Step 5: Fault Handling: Upon receiving a fault signal, the cloud will select the optimal backup drone to take over the task from the faulty drone.

[0023] The invention is further characterized by:

[0024] Step 2 is as follows:

[0025] Step 2.1: Calculate the total estimated runtime T of the task;

[0026] Step 2.2: Based on the number and performance of available drones, determine the working capacity of each drone, i.e., the number of tasks it can perform per unit of time;

[0027] Step 2.3: Calculate the working time t for each drone, i.e. the time window for task execution, t = T / N, where N is the number of drones;

[0028] Step 2.4: Divide the task: First, initialize an empty task segment; for each task segment, the start time t s Iterate from 0 to T at intervals of t; for each task segment, determine the end time t of that segment. e , i.e. t e =t s +t;

[0029] Step 2.5: Determine the task data required for the segment and add the segment to the task segment list;

[0030] Step 2.6: Return a list of task segments, where each task segment contains a start time, an end time, and related task data.

[0031] Step 3 is as follows:

[0032] Step 3.1: Task Requirements Analysis: When a new task request is received, the resource management module of the distributed cloud control center will conduct a comprehensive analysis of the task based on its nature, priority, resource requirements, and location.

[0033] Step 3.2: Calculate the task suitability score: The distributed cloud control center maintains an information database of all drones, including the performance parameters and sensor type information of each drone; the cloud will calculate the suitability score based on the degree of matching between the capability information of each drone and the task requirements.

[0034] Step 3.3: Calculate the availability score: The drone information database maintained by the distributed cloud control center includes the battery status, communication status, and sensor status of each drone. The cloud will combine this information to calculate the availability score of each drone.

[0035] Step 3.4: Calculate the distance score: The distributed cloud control center calculates the Euclidean distance between the task segment location and the UAV location, and then multiplies it by the corresponding weight to obtain the distance score;

[0036] Step 3.5: Add the fit score, availability score and distance score to get the evaluation score, sort the evaluation scores, and select the drone with the highest evaluation score as the optimal drone to execute the current task segment.

[0037] Step 4 is as follows:

[0038] Step 4.1: Formulate fault detection rules: Based on the design and operation characteristics of the UAV, formulate fault detection rules, specifically: define a sharp drop or interruption in the signal strength of the communication channel as a communication fault, a continuous drop or rise in battery voltage below or above the safe range or a sharp drop in battery capacity as a battery fault, and a sharp drop in the flight speed or altitude of the UAV as a power system fault.

[0039] Step 4.2: Obtain real-time status data of the UAV: ​​The distributed cloud control center will detect and collect various real-time status data of the UAV in real time, including: battery level, sensor data, communication status, flight speed and altitude;

[0040] Step 4.3: Fault Determination: Compare and analyze the real-time status data obtained in Step 4.2 with the fault detection rules set in Step 4.1. If the status data of a certain UAV meets the abnormal conditions defined in the fault detection rules, it is diagnosed as having a fault.

[0041] Step 4.4: Fault Diagnosis: When Step 4.3 determines that a certain UAV has malfunctioned, the system will trigger the fault diagnosis process, including further data analysis, fault mode identification, and error code matching, in order to determine the specific type and cause of the fault.

[0042] Step 4.5: Fault Handling: Once the fault diagnosis results in Step 4.4 are confirmed, the system will take corresponding fault handling measures, including: sending alarms to relevant personnel, executing automated fault recovery procedures, and dispatching backup drones to replace the faulty drone.

[0043] Step 4.6: Fault Recording and Analysis: After each fault occurs, the system will record detailed information about the fault event, including the fault type, occurrence time, and diagnostic results. These records will be used for subsequent fault analysis, system design improvement, and maintenance process optimization.

[0044] Step 5 is as follows:

[0045] Step 5.1: Task Requirements and Priority Assessment: The distributed cloud control center first determines the requirements and priorities of the tasks that the faulty drone has not completed. Different tasks have different performance requirements for backup drones. High-priority tasks need to be assigned to more reliable and high-performance backup drones.

[0046] Step 5.2: Availability Assessment: The distributed cloud control center assesses whether all backup drones are available based on real-time information from the backup drones. After filtering out unavailable backup drones, it evaluates the availability score of the available backup drones based on battery status, communication status, and sensor status information.

[0047] Step 5.3: Task Adaptability Assessment: After filtering out unavailable backup drones in Step 5.2, the cloud will analyze and record the resource status of all available backup drones, including computing power, communication bandwidth, and battery level; and then calculate the task adaptability score of each backup drone based on the analysis results of Step 5.1.

[0048] Step 5.4: Distance Assessment: Calculate the Euclidean distance between the mission segment location and the backup drone location, and then multiply it by the corresponding weight to obtain the distance score;

[0049] Step 5.5: Matching Assessment and Decision: Add the availability score, task suitability score, and distance score to obtain the assessment score. Sort the assessment scores and select the backup drone with the highest assessment score to execute the task that the faulty drone did not complete.

[0050] Key innovations of this invention:

[0051] 1. The distributed cloud control center of the UAV system proposed in this invention utilizes Kubernetes technology to build a distributed architecture, which consists of a security control module, a resource management module, a fault handling module, and a communication monitoring module, ensuring the efficiency and reliability of task scheduling and fault handling.

[0052] 2. This invention proposes a task partitioning and moving window technology, which can divide tasks into task segments based on the expected running time and distance, and set task moving windows for UAVs to optimize task allocation and execution.

[0053] 3. This invention proposes a fault replacement and task handover mechanism for backup drones. The backup drone is in a follow-flying state and takes over the task from the faulty drone when the mission drone fails.

[0054] The beneficial effects of this invention are:

[0055] (1) More reasonable allocation of system resources: The cloud server, as the control center, is driven by specific tasks when planning end system tasks, determines the end node drones that need to be securely coordinated, and provides full-process security services for the drones; the end node drones do not need to store additional authentication materials in advance, nor do they need a complicated certificate management mechanism, which greatly reduces the various expenses of resource-constrained end systems such as drones, and maximizes the advantages of cloud servers in computing and storage.

[0056] (2) Lower requirements for control link stability: The high dynamic operation of UAVs often causes the control link between the cloud server and the UAV to be intermittent. This invention only requires the cloud server to establish a communication connection with either of the UAVs of the two parties involved in the key negotiation. Through the authentication materials issued by the cloud server, the two UAVs can achieve key negotiation and establish an end-to-end secure channel within a specific range.

[0057] (3) Enhanced Session Key Security: When calculating the session key, this invention uses a long-term public-private key pair combined with a digital signature algorithm to authenticate messages, preventing man-in-the-middle attacks. Using long-term, short-term, and temporary public-private key pairs, along with verifiable timestamps in the request message, combined with a multi-round DH algorithm for key negotiation, it can resist replay attacks and ensure forward security. Each key negotiation session additionally uses a temporary public-private key pair; that is, each request message uses a different temporary public key, effectively preventing harm caused by the leakage of long and short-term secrets.

[0058] (4) Session key validity period is linked to the task cycle: Since the UAV updates the short-term public key stored on the cloud server after calculating the session key, the previous authentication materials and session key will become invalid once the current task ends, effectively linking the validity period of the session key to the task cycle. The negotiated session key is only valid during the execution period of this task, and even if the current session key is leaked, it will not cause harm to previous or subsequent tasks. Attached Figure Description

[0059] Figure 1 This is a schematic diagram of the system architecture principle of the present invention;

[0060] Figure 2 This is a schematic diagram illustrating the structural relationship between the present invention, the distributed cloud control center, the mission drone, and the backup drone. Detailed Implementation

[0061] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0062] Example 1

[0063] The present invention relates to a Kubernetes-based drone task scheduling and fault handling system, such as... Figure 1-2 As shown, it includes: a distributed cloud control center, a mission drone, and a backup drone; the distributed cloud control center is the central hub of the entire system, mainly responsible for mission scheduling, resource management, and fault detection; the mission drone is mainly responsible for mission execution, communication, and fault reporting; the backup drone is always in standby mode and is mainly responsible for mission takeover.

[0064] The distributed cloud control center includes:

[0065] The security control module utilizes Kubernetes' multi-layered security mechanisms, including authentication, authorization, and network isolation, to ensure that only authenticated users can access the system, and that communication and data transmission are secure. The authentication mechanism uses Kubernetes' built-in authentication functionality to verify user identity, ensuring that only legitimate users can access the cloud control center. The authorization mechanism leverages Kubernetes' authorization features to assign different permissions and roles to users, restricting their access to and operations on resources. Network isolation utilizes Kubernetes' network policy features to define network isolation rules, ensuring that communication between different users or services is controlled and secure. Through these principles, functions, and workflows, the security control module ensures the security and data protection of the distributed cloud control center. It uses Kubernetes' authentication, authorization, and network isolation mechanisms to verify user legitimacy and control permissions, preventing unauthorized access and operations. Furthermore, the security control module also ensures system communication security and data confidentiality through secure communication and data protection measures.

[0066] The communication monitoring module utilizes Kubernetes' resource management capabilities to monitor the status and resource usage of drones, implementing two main functions: real-time monitoring and resource utilization analysis. Real-time monitoring tracks the status of each drone in the task drone cluster, acquiring key information such as current status, available resources, and operational status. Through this function, the communication monitoring module can obtain information such as drone health status, task execution status, and communication quality. The resource utilization analysis function acquires the resource consumption of each drone, such as CPU utilization and memory usage, and calculates the overall resource utilization. By analyzing resource utilization, the communication monitoring module can assess the operational status and load of the drone cluster.

[0067] The resource management module is a key component of the distributed cloud control center. It utilizes the Kubernetes scheduler for scheduling and allocating drone tasks, achieving rational resource allocation and efficient task execution. This includes dynamic adjustment and task scheduling. Dynamic adjustment achieves load balancing and resource optimization based on the real-time status of the drones. The task scheduling function automatically selects the appropriate drone to execute the task based on its attributes and requirements, as well as the available resources of the drones. It fully considers indicators such as task priority, resource requirements, and availability, and uses scheduling algorithms to assign tasks to the most suitable drones.

[0068] Fault Handling Module: The fault handling module is also a crucial component of the distributed cloud control center. It leverages Kubernetes' autoscaling and fault recovery mechanisms to achieve seamless repair of faulty drones and seamless task switching. This includes fault repair and task switching functions. The fault repair function automatically attempts to repair a faulty drone upon detection; if repair is successful, the system can resume normal operation. The task switching function selects the optimal backup drone to take over the task when a faulty drone cannot be repaired. It also notifies the resource management module to send any unfinished tasks from the faulty drone to the optimal backup drone, ensuring seamless task switching and guaranteeing task continuity and completion.

[0069] The organizational structure of the mission drone includes a communication module, a health self-check module, a mission processing module, and a data processing module. The communication module is responsible for communicating with the cloud control center, receiving mission instructions, and sending mission completion signals, ensuring a normal communication connection between the mission drone and the cloud control center. The health self-check module is responsible for periodically checking the status and health of the mission drone, including battery level, sensor status, and communication connectivity, and determining if any faults or anomalies have occurred. The mission processing module is responsible for mission execution and control, controlling the drone to execute mission segments sequentially according to mission requirements and monitoring mission completion. The data processing module is responsible for processing and analyzing the data collected by the mission drone, performing image processing, target recognition, and data compression to extract useful information and optimize data transmission.

[0070] Backup drones are used to take over the work and complete the remaining tasks when the mission drone malfunctions or fails. The mission drone's organizational structure includes a communication module, a health self-check module, a task processing module, and a data processing module. The communication module is responsible for communicating with the cloud control center, receiving mission instructions, and sending mission completion signals, ensuring a normal communication connection between the mission drone and the cloud control center. The health self-check module is responsible for periodically checking the status and health of the mission drone, detecting battery level, sensor status, communication connectivity, etc., and determining whether any faults or anomalies have occurred. The task processing module is responsible for mission execution and control, controlling the drone to execute mission segments sequentially according to mission requirements and monitoring mission completion. The data processing module is responsible for processing and analyzing the data collected by the mission drone, performing functions such as image processing, target recognition, and data compression to extract useful information and optimize data transmission.

[0071] Example 2

[0072] The specific workflow of the distributed cloud control center in the Kubernetes-based UAV mission scheduling and fault handling system of this invention is as follows:

[0073] S1: Security Authentication: In this invention, drones need to communicate and interact with the distributed cloud control center; first, they must pass the identity authentication of the security protection module to ensure that only legitimate devices can access the system, preventing unauthorized access and attacks;

[0074] S2: Task Division: The resource management module divides the received tasks into task segments. At the same time, in order to further optimize task execution, the resource management module sets a task segment movement window, that is, a time window for task segments, which specifies the time requirement for the task drone to complete the task segment within the specified time period. The task drone needs to complete the task segment within the specified time to ensure the continuity and smoothness of the task.

[0075] S3: Surveillance Drones: The monitoring module will monitor the status of all drones in real time. Once a fault signal is detected from a task drone, the fault handling module will select the best backup drone to take over the task to ensure the continuity and successful completion of the task.

[0076] S4: Task Scheduling: The resource management module selects appropriate drones to execute different task segments based on the drone's power consumption, distance, and other relevant factors to optimize task execution efficiency and resource utilization. The next task segment will only be sent after the task completion signal is received from the task drone. This ensures that tasks are executed continuously in a predetermined order, preventing task conflicts and data loss.

[0077] S5: Data Synchronization: The cloud control center will synchronize data with the drones in the system in real time to ensure that the drones' status and mission progress can be monitored.

[0078] S6: Task Completed: When the communication monitoring module detects that there are still drones that have not completed their tasks, it will continue to monitor them. Only when it receives the task completion signal from all drones will the cloud control center end the monitoring work and exit safely.

[0079] Example 3

[0080] The Kubernetes-based drone task scheduling and fault handling methods include:

[0081] Step 1: System Initialization: Start the distributed cloud control center, mission drone, and backup drone. Each component performs initialization and preparation tasks;

[0082] Step 2: Task Analysis and Division: After receiving a task request, the distributed cloud control center will analyze the task based on its characteristics, priority, and resource requirements, and then divide the task into task segments.

[0083] Step 2.1: Calculate the total estimated runtime T of the task;

[0084] Step 2.2: Based on the number and performance of available drones, determine the working capacity of each drone, i.e., the number of tasks it can perform per unit of time;

[0085] Step 2.3: Calculate the working time t for each drone, i.e. the time window for task execution, t = T / N, where N is the number of drones;

[0086] Step 2.4: Divide the task: First, initialize an empty task segment; for each task segment, the start time t s Iterate from 0 to T at intervals of t; for each task segment, determine the end time t of that segment. e , i.e. t e =t s +t;

[0087] Step 2.5: Determine the task data required for the segment and add the segment to the task segment list;

[0088] Step 2.6: Return a list of task segments, where each task segment contains a start time, an end time, and related task data;

[0089] Step 3: Task scheduling: The distributed cloud control center will monitor the status of the task drones in real time and assign the task segments to the most suitable drones for execution based on the monitoring results;

[0090] Step 3.1: Task Requirements Analysis: When a new task request is received, the resource management module of the distributed cloud control center will conduct a comprehensive analysis of the task based on its nature, priority, resource requirements, and location.

[0091] Step 3.2: Calculate the task suitability score: The distributed cloud control center maintains an information database of all drones, including the performance parameters and sensor type information of each drone; the cloud will calculate the suitability score based on the degree of matching between the capability information of each drone and the task requirements.

[0092] Step 3.3: Calculate the availability score: The drone information database maintained by the distributed cloud control center includes the battery status, communication status, and sensor status of each drone. The cloud will combine this information to calculate the availability score of each drone.

[0093] Step 3.4: Calculate the distance score: The distributed cloud control center calculates the Euclidean distance between the task segment location and the UAV location, and then multiplies it by the corresponding weight to obtain the distance score;

[0094] Step 3.5: Add the fit score, availability score and distance score to get the evaluation score, sort the evaluation scores, and select the drone with the highest evaluation score as the optimal drone to execute the current task segment;

[0095] Step 4: Fault Detection: The mission drone will periodically perform a health self-check. When a fault or abnormality is detected, the drone will first perform automatic repair. If the repair fails, it will send a fault signal to the cloud.

[0096] Step 4.1: Formulate fault detection rules: Based on the design and operation characteristics of the UAV, formulate fault detection rules, specifically: define a sharp drop or interruption in the signal strength of the communication channel as a communication fault, a continuous drop or rise in battery voltage below or above the safe range or a sharp drop in battery capacity as a battery fault, and a sharp drop in the flight speed or altitude of the UAV as a power system fault.

[0097] Step 4.2: Obtain real-time status data of the UAV: ​​The distributed cloud control center will detect and collect various real-time status data of the UAV in real time, including: battery level, sensor data, communication status, flight speed and altitude;

[0098] Step 4.3: Fault Determination: Compare and analyze the real-time status data obtained in Step 4.2 with the fault detection rules set in Step 4.1. If the status data of a certain UAV meets the abnormal conditions defined in the fault detection rules, it is diagnosed as having a fault.

[0099] Step 4.4: Fault Diagnosis: When Step 4.3 determines that a certain UAV has malfunctioned, the system will trigger the fault diagnosis process, including further data analysis, fault mode identification, and error code matching, in order to determine the specific type and cause of the fault.

[0100] Step 4.5: Fault Handling: Once the fault diagnosis results in Step 4.4 are confirmed, the system will take corresponding fault handling measures, including: sending alarms to relevant personnel, executing automated fault recovery procedures, and dispatching backup drones to replace the faulty drone.

[0101] Step 4.6: Fault Recording and Analysis: After each fault occurs, the system will record detailed information about the fault event, including the fault type, occurrence time, and diagnostic results. These records will be used for subsequent fault analysis, system design improvement, and maintenance process optimization.

[0102] Step 5: Fault Handling: Upon receiving a fault signal, the cloud will select the optimal backup drone to take over the task from the faulty drone.

[0103] Step 5.1: Task Requirements and Priority Assessment: The distributed cloud control center first determines the requirements and priorities of the tasks that the faulty drone has not completed. Different tasks have different performance requirements for backup drones. High-priority tasks need to be assigned to more reliable and high-performance backup drones.

[0104] Step 5.2: Availability Assessment: The distributed cloud control center assesses whether all backup drones are available based on real-time information from the backup drones. After filtering out unavailable backup drones, it evaluates the availability score of the available backup drones based on battery status, communication status, and sensor status information.

[0105] Step 5.3: Task Adaptability Assessment: After filtering out unavailable backup drones in Step 5.2, the cloud will analyze and record the resource status of all available backup drones, including computing power, communication bandwidth, and battery level; and then calculate the task adaptability score of each backup drone based on the analysis results of Step 5.1.

[0106] Step 5.4: Distance Assessment: Calculate the Euclidean distance between the mission segment location and the backup drone location, and then multiply it by the corresponding weight to obtain the distance score;

[0107] Step 5.5: Matching Assessment and Decision: Add the availability score, task suitability score, and distance score to obtain the assessment score. Sort the assessment scores and select the backup drone with the highest assessment score to execute the task that the faulty drone did not complete.

[0108] The workflow of the mission-specific drone is as follows:

[0109] (1) Mission preparation: The mission UAV performs mission preparation work, including system settings, sensor calibration, route planning, etc.

[0110] (2) Health Monitoring: The mission drone periodically monitors its own status and health condition through a health self-check module. By monitoring key parameters such as battery level, sensor status, and communication connectivity, the module ensures the drone remains operational. If a malfunction occurs, the health self-check module will attempt to repair it.

[0111] If the repair is successful, the drone will continue its mission; if the repair fails, the communication module will send a fault signal to the cloud.

[0112] (3) Mission Fragment Reception: After the mission drone confirms that its own status is normal, it will receive mission fragments from the cloud control center.

[0113] (4) Mission Segment Execution: The mission processing module controls the mission UAV to execute mission segments sequentially according to the mission requirements and instructions. Each mission segment covers tasks such as flight, data acquisition, and image capture in a specific area.

[0114] (5) Sending a completion signal: Whenever the mission drone completes a mission segment, the communication module will send a mission completion signal to the cloud control center to notify the cloud of the mission's progress.

[0115] (6) Data synchronization: The communication module synchronizes the data, images and other information collected by the data processing module with the cloud control center to ensure that the cloud can monitor the progress of the task and the data collection status in real time.

[0116] (7) Mission completion: After the mission drone completes all mission segments, the communication module sends a signal to the cloud that the mission is complete and reports the results to the cloud control center.

[0117] Through the modules and workflow described above, the mission drone can receive mission commands and execute specific mission operations. The modules work together to ensure the mission drone can successfully perform its mission, achieving functions such as flight, control, and data acquisition. It also interacts with the cloud control center via the communication module to exchange information and receive mission status feedback.

[0118] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for scheduling and handling faults of unmanned aerial vehicles based on Kubernetes, characterized in that, A Kubernetes-based drone mission scheduling and fault handling system is adopted, including: Step 1: System Initialization: Start the distributed cloud control center, mission drone and backup drone, and perform initialization and preparation work for each component; Step 2: Task Analysis and Division: After receiving a task request, the distributed cloud control center will analyze the task based on its characteristics, priority, and resource requirements, and then divide the task into task segments; Step 2 is detailed below: Step 2.1: Calculate the total estimated runtime T of the task; Step 2.2: Based on the number and performance of available drones, determine the working capacity of each drone, i.e., the number of tasks it can perform per unit of time; Step 2.3: Calculate the working time t for each drone, i.e., the time window for task execution, t=T / N, where N is the number of drones; Step 2.4: Divide the task: First, initialize an empty task segment; for each task segment, the start time... Iterate from 0 to T at intervals of t; for each task segment, determine the end time of that segment. ,Right now ; Step 2.5: Determine the task data required for the segment and add the segment to the task segment list; Step 2.6: Return a list of task segments, where each task segment contains a start time, an end time, and related task data; Step 3: Task scheduling: The distributed cloud control center will monitor the status of the task drones in real time and assign the task segments to the most suitable drones for execution based on the monitoring results; Step 4: Fault Detection: The mission drone will periodically perform a health self-check. When a fault or abnormality is detected, the drone will first perform automatic repair. If the repair fails, it will send a fault signal to the cloud. Step 5: Fault Handling: Upon receiving a fault signal, the cloud will select the optimal backup drone to take over the task from the faulty drone.

2. The Kubernetes-based drone task scheduling and fault handling method as described in claim 1, characterized in that, The Kubernetes-based drone mission scheduling and fault handling system includes: a distributed cloud control center, mission drones, and backup drones; the distributed cloud control center is the central hub of the entire system, responsible for mission scheduling, resource management, and fault detection; the mission drones are responsible for mission execution, communication, and fault reporting; and the backup drones are always on standby, responsible for mission takeover.

3. The Kubernetes-based UAV task scheduling and fault handling method according to claim 2, characterized in that, The distributed cloud control center includes: Security control module: Utilizes the multi-layered security mechanisms provided by Kubernetes, including authentication, authorization, and network isolation, to ensure that only authenticated users can access the system; The communication monitoring module includes two functions: real-time monitoring and resource utilization analysis. The real-time monitoring function monitors the status of each UAV in the task UAV cluster; the resource utilization analysis function obtains the resource consumption of each UAV. Resource management module: It utilizes the Kubernetes scheduler to schedule and allocate drone tasks; including dynamic adjustment and task scheduling. The dynamic adjustment achieves load balancing and resource optimization based on the real-time status of the task drones; the task scheduling function automatically selects the appropriate drone to execute the task based on the task's attributes and requirements, as well as the drone's available resources. The fault handling module includes fault repair and task switching functions. The fault repair function automatically attempts to repair a faulty drone when it is detected. If the repair is successful, the system can continue to operate normally. The task switching function selects the optimal backup drone to take over the task when the faulty drone cannot be repaired. It also notifies the resource management module to send the unfinished tasks of the faulty drone to the optimal backup drone, ensuring seamless task switching and guaranteeing task continuity and completion.

4. The Kubernetes-based UAV task scheduling and fault handling method according to claim 2, characterized in that, The organizational structure of the mission drone includes a communication module, a health self-check module, a task processing module, and a data processing module. The communication module is responsible for communicating with the cloud control center, receiving task instructions, and sending task completion signals, ensuring a normal communication connection between the mission drone and the cloud control center. The health self-check module is responsible for periodically checking the status and health of the mission drone; it can detect the drone's battery level, sensor status, and communication connection, and determine whether any faults or abnormalities have occurred. The task processing module is responsible for task execution and control; it can control the drone to execute task segments sequentially according to task requirements and monitor the task completion status. The data processing module is responsible for processing and analyzing the data collected by the mission drone, enabling image processing, target recognition, and data compression to extract useful information and optimize data transmission.

5. The Kubernetes-based UAV task scheduling and fault handling method according to claim 2, characterized in that, The backup drone is used to take over the work and complete the remaining tasks when the mission drone fails or malfunctions.

6. The Kubernetes-based drone task scheduling and fault handling method as described in claim 1, characterized in that, Step 3 is as follows: Step 3.1: Task Requirements Analysis: When a new task request is received, the resource management module of the distributed cloud control center will conduct a comprehensive analysis of the task based on its nature, priority, resource requirements, and location. Step 3.2: Calculate the task suitability score: The distributed cloud control center maintains an information database of all drones, including the performance parameters and sensor type information of each drone; the cloud will calculate the suitability score based on the degree of matching between the capability information of each drone and the task requirements. Step 3.3: Calculate the availability score: The drone information database maintained by the distributed cloud control center includes the battery status, communication status, and sensor status of each drone. The cloud will combine this information to calculate the availability score of each drone. Step 3.4: Calculate the distance score: The distributed cloud control center calculates the Euclidean distance between the task segment location and the UAV location, and then multiplies it by the corresponding weight to obtain the distance score; Step 3.5: Add the fit score, availability score and distance score to get the evaluation score, sort the evaluation scores, and select the drone with the highest evaluation score as the optimal drone to execute the current task segment.

7. The Kubernetes-based drone task scheduling and fault handling method as described in claim 1, characterized in that, Step 4 is as follows: Step 4.1: Formulate fault detection rules: Based on the design and operation characteristics of the UAV, formulate fault detection rules, specifically: define a sharp drop or interruption in the signal strength of the communication channel as a communication fault, a continuous drop or rise in battery voltage below or above the safe range or a sharp drop in battery capacity as a battery fault, and a sharp drop in the flight speed or altitude of the UAV as a power system fault. Step 4.2: Obtain real-time status data of the UAV: ​​The distributed cloud control center will detect and collect various real-time status data of the UAV in real time, including: battery level, sensor data, communication status, flight speed and altitude; Step 4.3: Fault Determination: Compare and analyze the real-time status data obtained in Step 4.2 with the fault detection rules set in Step 4.

1. If the status data of a certain UAV meets the abnormal conditions defined in the fault detection rules, it is diagnosed as having a fault. Step 4.4: Fault Diagnosis: When Step 4.3 determines that a certain UAV has malfunctioned, the system will trigger the fault diagnosis process, including further data analysis, fault mode identification, and error code matching, in order to determine the specific type and cause of the fault. Step 4.5: Fault Handling: Once the fault diagnosis results in Step 4.4 are confirmed, the system will take corresponding fault handling measures, including: sending alarms to relevant personnel, executing automated fault recovery procedures, and dispatching backup drones to replace the faulty drone. Step 4.6: Fault Recording and Analysis: After each fault occurs, the system will record detailed information about the fault event, including the fault type, occurrence time, and diagnostic results. These records will be used for subsequent fault analysis, system design improvement, and maintenance process optimization.

8. The Kubernetes-based drone task scheduling and fault handling method as described in claim 1, characterized in that, Step 5 is as follows: Step 5.1: Task Requirements and Priority Assessment: The distributed cloud control center first determines the requirements and priorities of the tasks that the faulty drone has not completed. Different tasks have different performance requirements for backup drones. High-priority tasks need to be assigned to more reliable and high-performance backup drones. Step 5.2: Availability Assessment: The distributed cloud control center assesses whether all backup drones are available based on real-time information from the backup drones. After filtering out unavailable backup drones, it evaluates the availability score of the available backup drones based on battery status, communication status, and sensor status information. Step 5.3: Task Adaptability Assessment: After filtering out unavailable backup drones in Step 5.2, the cloud will analyze and record the resource status of all available backup drones, including computing power, communication bandwidth, and battery level. Then, based on the analysis results of step 5.1, calculate the task suitability score for each backup drone; Step 5.4: Distance Assessment: Calculate the Euclidean distance between the mission segment location and the backup drone location, and then multiply it by the corresponding weight to obtain the distance score; Step 5.5: Matching Assessment and Decision: Add the availability score, task suitability score, and distance score to obtain the assessment score. Sort the assessment scores and select the backup drone with the highest assessment score to execute the task that the faulty drone did not complete.

Images