A radar self-checking and self-recovery control method, device and equipment for spacer intelligent installation unmanned aerial vehicle

Abstract

The application discloses a radar self-checking and self-recovery control method, device and equipment for intelligent installation of an unmanned aerial vehicle, relates to the technical field of reliability control of unmanned aerial vehicle systems and intelligent construction of power transmission lines, and comprises the following steps: monitoring a radar working state in real time and grading abnormality levels, logically isolating abnormal radars to prohibit abnormal data from participating in decision-making, executing differentiated self-recovery and recovery access interlocking based on abnormality levels, removing the isolation when conditions are met, identifying task stage risk characteristics and executing differentiated safety degradation control during the isolation period. The application improves the reliability and continuity of the unmanned aerial vehicle operation in a complex electromagnetic environment.

Description

Technical Field

[0001] This invention belongs to the technical field of unmanned aerial vehicle (UAV) system reliability control and intelligent power transmission line construction, specifically relating to a radar self-testing and self-healing control method, device and equipment for UAVs used for intelligent installation of spacer bars. Background Technology

[0002] In unmanned installation of spacers on power transmission lines, drones typically need to perform high-precision hovering and docking operations in the near-field area of ​​the conductors. This scenario is characterized by a complex electromagnetic environment, frequent changes in flight attitude, and significant power supply fluctuations. Radar, as a key sensor for drones to perceive obstacles and assist in positioning, directly determines the safety of the operation. Current technologies for handling radar anomalies often employ a single threshold-triggered restart mechanism or a global degradation strategy, lacking a refined classification and judgment of the severity of the anomaly. This leads to minor faults potentially causing unnecessary power outages and restarts, while severe faults may result in the system operating with defects due to insufficient recovery verification. Furthermore, existing safety degradation strategies often do not differentiate between the current mission phase of the drone, applying the same avoidance logic to high-risk near-field operation phases and low-risk cruise phases. This can easily lead to secondary safety risks such as line collisions during critical operational windows due to a one-size-fits-all degradation action, or cause operational continuity interruptions due to excessive conservatism. Therefore, there is an urgent need for a radar self-checking and self-healing control method capable of performing differentiated self-healing and degradation control based on the anomaly level and mission phase risk characteristics. Summary of the Invention

[0003] To address the shortcomings of existing technologies, the present invention aims to provide a radar self-inspection and self-healing control method, device, and equipment for spacer bar intelligent installation drones, thereby achieving refined hierarchical control of radar status and scene-adaptive safety fault tolerance.

[0004] This invention provides a radar self-testing and self-healing control method for a spacer bar intelligent installation UAV, comprising:

[0005] Real-time monitoring of radar equipment's operational status;

[0006] Based on the working status information, it is determined whether the radar device is in an abnormal state, and the abnormal state is classified into multiple abnormal levels, including at least mild abnormality and severe abnormality.

[0007] In response to determining that the radar device is in the abnormal state, the radar device is logically isolated to prevent abnormal data from participating in perception fusion and flight control decisions;

[0008] Differentiated self-healing control is executed based on the anomaly level, and a recovery access interlock is executed after the self-healing control is executed. The logical isolation is released when the preset recovery conditions are met. During the isolation of the radar equipment, the risk characteristics of the current mission stage are identified, and differentiated security degradation control is executed based on the risk characteristics.

[0009] The step of performing differentiated self-healing control based on the anomaly level, and performing a recovery access interlock after the self-healing control is executed, includes: setting self-healing strategy parameters, which include a maximum number of self-healing attempts, a single self-healing timeout, and a self-healing action priority corresponding to different anomaly levels; sending a self-healing control command to the power management unit to execute the corresponding self-healing action based on the anomaly level and the self-healing action priority; and triggering a mission abort or return-to-home alarm strategy when the maximum number of self-healing attempts is reached and the radar equipment has not yet recovered.

[0010] The process of identifying the risk characteristics of the current task phase and performing differentiated safety degradation control based on the risk characteristics includes: identifying the risk level of the current task phase of the UAV while the radar equipment is isolated; if it is in a high-risk operation phase, performing stop or rollback safety actions; if it is in a low-risk operation phase, switching to an obstacle avoidance strategy that relies on backup sensors and limiting motion parameters.

[0011] The step of sending a self-healing control command to the power management unit to execute a corresponding self-healing action based on the anomaly level and the self-healing action priority includes: when the anomaly level is a minor anomaly, prioritizing the execution of a soft reset operation, the soft reset operation including sending a reset command through the control register of the radar device; when the anomaly level is a severe anomaly, performing a power-off restart operation, the power-off restart operation including controlling the power management unit to cut off the power supply to the radar device and re-energize it after a preset time.

[0012] The step of releasing the logical isolation when the preset recovery condition is met includes: after the self-healing control is executed, monitoring the working status information of the radar device again; when the data frame rate, packet loss rate, communication link status, power supply voltage, and working mode configuration consistency flag all recover to the preset health threshold and remain stable within multiple consecutive sampling periods, it is determined that the preset recovery condition is met and the logical isolation is released; if the preset recovery condition is not met, the logical isolation is maintained, and the self-healing control is repeated or a fallback strategy is entered.

[0013] If the operation is in a high-risk phase, perform safety actions such as stopping or retreating. If the operation is in a low-risk phase, switch to an obstacle avoidance strategy that relies on backup sensors and limit motion parameters, including: if the current task phase is a near-field operation phase with the guide wire, perform safety actions such as hovering, retreating to a safe distance, or exiting the guide wire corridor; if the current task phase is a return or cruise phase, switch to a visual obstacle avoidance priority mode and limit the maximum flight speed and maximum acceleration.

[0014] The logical isolation of the radar equipment to prevent abnormal data from participating in perception fusion and flight control decisions includes: shielding radar observation data at the perception fusion layer, or reducing the radar observation weight to zero and outputting a radar unavailable status flag.

[0015] This invention also provides a radar self-testing and self-healing control device for a spacer-mounted intelligent UAV, comprising: a status monitoring module for real-time monitoring of the radar device's operating status information; a health determination module for determining whether the radar device is in an abnormal state based on the operating status information, and classifying the abnormal state into multiple abnormality levels, including at least mild and severe abnormalities; an abnormality isolation module for logically isolating the radar device in response to determining that the radar device is in the abnormal state, thereby prohibiting abnormal data from participating in perception fusion and flight control decisions; a self-healing control module for performing differentiated self-healing control based on the abnormality level, and performing a recovery access interlock after the self-healing control is executed, and releasing the logical isolation when a preset recovery condition is met; and a degradation control module for identifying risk characteristics of the current task stage during the isolation of the radar device, and performing differentiated safety degradation control based on the risk characteristics.

[0016] The aforementioned device, through the coordinated operation of its various functional modules, solidifies the logical flow of the aforementioned method into executable software or hardware units, realizing the modular encapsulation of radar health management functions. This facilitates integration into different types of UAV flight control systems, thereby improving the system's scalability and ease of maintenance.

[0017] The present invention also provides a spacer bar intelligent installation drone, including a drone body, radar equipment, power management unit and flight control computer, wherein the flight control computer is configured to execute the radar self-test and self-healing control method for spacer bar intelligent installation drone as described above.

[0018] The aforementioned UAV products deeply embed radar self-testing and self-healing control methods into the flight control computer, enabling the UAV to autonomously cope with perception system failures in complex electromagnetic environments. This significantly improves the automation level and safety margin of spacer bar installation operations and reduces reliance on external manual intervention.

[0019] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the radar self-testing and self-healing control method for a spacer-mounted unmanned aerial vehicle as described above.

[0020] The aforementioned storage medium serves as the carrier of the software product, supporting offline distribution, online upgrades, and cloud deployment of the control method, and providing a standardized software delivery foundation for the batch operation and maintenance and algorithm iteration of UAV swarms.

[0021] This invention discloses a radar self-testing and self-healing control method, device, and equipment for spacer bar intelligent installation UAVs. By establishing a differentiated self-healing and recovery access interlock mechanism based on anomaly level, it changes the crude mode of the single restart strategy in the prior art. It accurately matches soft reset or power failure restart resources according to the severity of the fault, and combined with continuous multi-cycle stability verification, it eliminates the risk of system oscillation and secondary faults caused by false recovery or over-recovery in principle, and significantly improves the self-recovery reliability and data credibility of the radar subsystem under complex operating conditions.

[0022] The method of this invention breaks through the scenario limitations of traditional global degradation strategies by introducing differentiated safety degradation control based on the risk characteristics of the task stage. In high-risk stages such as the near field of the conductor, it actively performs spatial avoidance actions to build a physical safety barrier. In low-risk stages such as cruise, it maintains task continuity through multi-source fusion and motion constraints. This scenario-adaptive fault-tolerant mechanism effectively mitigates secondary safety risks during perception failure and ensures the safety and efficiency of intelligent spacer installation operations.

[0023] In addition, the end-to-end logical isolation and status recording feedback mechanism enables the immediate blocking of abnormal data and the complete tracing of fault information. This not only prevents the contamination of flight decisions by erroneous perception, but also provides detailed data support for subsequent operation and maintenance analysis and algorithm optimization, thereby enhancing the engineering implementation capability and full life cycle management level of the entire UAV system. Attached Figure Description

[0024] Figure 1 This is a flowchart of a radar self-testing and self-healing control method for a spacer bar intelligent installation UAV according to an embodiment of this application. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0026] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application.

[0027] like Figure 1 As shown, this embodiment provides a radar self-testing and self-healing control method for spacer bar intelligent installation UAVs. This method solves the problems of poor system stability and high operational safety risks caused by the coarse radar anomaly handling strategies in existing technologies by constructing a closed-loop control link consisting of state monitoring, anomaly classification, logic isolation, differentiated self-healing and access interlocking, and differentiated safety degradation. It should be understood that although... Figure 1 The process sequence is shown as linear, but in actual engineering implementation, some steps, such as status monitoring and safety degradation control, can be executed in parallel or triggered asynchronously, as long as the overall closed-loop control logic is satisfied.

[0028] Specifically, the method includes the following steps:

[0029] Step S101: Monitor the working status information of the radar equipment in real time.

[0030] In this embodiment, the monitoring of operational status information is not limited to whether a single data output is normal, but encompasses multi-dimensional feature vectors reflecting the health of the radar equipment. This multi-dimensional monitoring mechanism aims to provide sufficient data support for subsequent accurate classification, avoiding misjudgments of system status due to instantaneous fluctuations in a single indicator. For example, operational status information may include, but is not limited to, the completeness and timeliness of data output, the stability of communication links, and the matching degree between internal electrical parameters and environmental parameters. By collecting this low-level status information in real time, the system can establish a comprehensive perception of the radar equipment's operational status, which is the foundation for achieving refined fault-tolerant control.

[0031] Step S102: Determine whether the radar equipment is in an abnormal state based on the working status information, and classify the abnormal state into multiple abnormal levels, including at least mild abnormality and severe abnormality.

[0032] The core of this step lies in mapping continuous, multi-dimensional state monitoring data into discrete, semantically clear anomaly level identifiers. Unlike traditional binary judgments (normal / fault), the hierarchical judgment mechanism adopted in this embodiment can distinguish the nature and severity of anomalies. Specifically, mild anomalies typically indicate a decline in radar performance but still possessing some reliability, or the presence of transient interference that can be automatically recovered; while severe anomalies indicate loss of radar function, completely unreliable data, or the risk of hardware-level failure. The technical value of this hierarchical mechanism lies in providing a decision-making basis for subsequent resource scheduling, enabling the system to match the most appropriate handling strategy according to the severity and urgency of the fault. This avoids treating all anomalies as fatal faults and triggering a costly system-wide restart, and also prevents the accumulation of hidden dangers due to only taking low-intensity intervention for severe faults.

[0033] Step S103: In response to determining that the radar equipment is in an abnormal state, the radar equipment is logically isolated to prevent abnormal data from participating in perception fusion and flight control decisions.

[0034] It is important to emphasize that the logical isolation in this embodiment is essentially a soft isolation mechanism implemented at the perception fusion algorithm layer, rather than a physical hardware power-off or signal cutoff. Specifically, when a radar anomaly is detected, the system dynamically adjusts the weight of the radar observation data to zero in the multi-source sensor fusion algorithm, or directly blocks the data input of that channel, while simultaneously outputting a clear radar unavailability status flag to the downstream decision module. The reason for adopting this algorithm-level soft isolation instead of hardware hard cutoff is twofold: firstly, the response speed of soft isolation is much faster than the action time of hardware relays, which can block the propagation of erroneous data within milliseconds, preventing the flight control system from generating incorrect attitude correction commands due to receiving sudden changes in data; secondly, maintaining the physical power supply and communication link of the radar equipment preserves the necessary interaction channel for subsequent online diagnosis and rapid self-healing, avoiding the lengthy initialization process caused by physical power-off and power-back, which is crucial for ensuring the continuous response capability of the UAV in complex electromagnetic environments.

[0035] Step S104: Perform differentiated self-healing control based on the anomaly level, and perform recovery access interlock after the self-healing control is executed. Release the logic isolation when the preset recovery conditions are met.

[0036] This step comprises two closely related sub-processes: differentiated self-healing and recovery access interlock. Differentialized self-healing refers to the system selecting a matching sequence of self-healing actions from a pre-defined policy library based on the anomaly level determined in step S102. For example, a minor anomaly might only trigger software-level configuration reload or cache cleanup, while a severe anomaly might trigger hardware-level power cycling or deep reset. This policy matching mechanism ensures accurate deployment of self-healing resources, improving recovery efficiency and reducing unnecessary disturbances to the overall system operation. Recovery access interlock refers to the process where, after the self-healing action is completed, the fusion weights of the radar data are not immediately restored. Instead, a strict verification window is initiated. Only when the radar's various status indicators not only recover to within the healthy threshold range but also exhibit sustained stability over time are the preset recovery conditions met and the logical isolation released. This interlocking mechanism is equivalent to adding a time-domain filter to the self-healing output, which fundamentally eliminates the false recovery phenomenon caused by the radar being in a critical recovery state or intermittent failure. It effectively prevents system oscillations caused by frequent jumps in sensing data between normal and abnormal states, and ensures that the radar data reconnected to the system is truly reliable and stable.

[0037] Step S106: During the isolation of the radar equipment, identify the risk characteristics of the current mission phase and perform differentiated security degradation control based on the risk characteristics.

[0038] After the radar is logically isolated, the UAV enters a degraded operation mode using single-sensor or multi-sensor fusion. At this point, the system does not execute a globally unified safety strategy, but first identifies the current mission phase and its corresponding risk characteristics. The risk characteristics of the mission phase reflect the UAV's dependence on radar perception capabilities in the current spatiotemporal environment and the severity of the consequences of a collision. Based on this, the system executes differentiated safety degradation control: if a high-risk operational phase is identified (such as close-range docking or narrow space traversal), a conservative stop, hover, or active retreat strategy is adopted, trading space for safety and prioritizing the elimination of dynamic collision hazards; if a low-risk operational phase is identified (such as open airspace cruising or returning to base), a backup sensor-dominated obstacle avoidance mode is switched, supplemented by kinematic parameter constraints, to maintain mission continuity as much as possible while ensuring basic safety. This scene-aware adaptive degradation strategy effectively resolves the contradiction of traditional one-size-fits-all degradation schemes, which may cause secondary risks due to overly aggressive actions in near-field operations or mission interruptions due to excessive conservatism in far-field operations, achieving a dynamic optimal balance between safety and operational efficiency.

[0039] Through the coordinated operation of steps S101 to S106 described above, this embodiment constructs an intelligent fault-tolerant closed loop with self-perception, self-diagnosis, self-repair, and scenario adaptation capabilities. This closed loop not only ensures the purity and stability of the data flow at the micro level through logical isolation and access interlocking, but also ensures the behavioral safety of the drone during perception failures at the macro level through differentiated self-healing and scenario-based degradation. This significantly improves the operational reliability and mission success rate of the spacer bar intelligent installation drone under complex working conditions.

[0040] Furthermore, the radar device operating status information in step S101 is described in detail. In this embodiment, the radar device operating status information includes data frame rate f (unit: Hz), packet loss rate p (unit: %), internal temperature t (unit: ℃), power supply voltage v (unit: V), operating current i (unit: mA), and operating mode configuration consistency flag m (1 indicates consistency, 0 indicates inconsistency).

[0041] Furthermore, the health status determination in step S102 is explained in detail. In this embodiment, the flight control computer compares the collected state vector with a preset threshold and executes a graded determination logic:

[0042] If the data frame rate f is less than 20Hz and lasts for more than 1 second, or the packet loss rate p is greater than 10% and lasts for more than 2 seconds, but the communication link is still intact, it is judged as a minor anomaly.

[0043] If the communication link is interrupted for more than 3 seconds, or the power supply voltage v is less than 4.0V or greater than 6.0V, or the internal temperature t of the radar is greater than 85℃, it is considered a serious anomaly.

[0044] Generate an anomaly level identifier (0-normal, 1-mild anomaly, 2-severe anomaly) and trigger an anomaly signal.

[0045] Furthermore, the differentiated self-healing control mechanism in step S104 and the recovery access interlock mechanism in step S105 are explained in detail. In this embodiment, differentiated self-healing control is executed based on the anomaly level, and a recovery access interlock is executed after the self-healing control is executed. Specifically, this includes setting self-healing strategy parameters, which include the maximum number of self-healing attempts, the timeout for a single self-healing attempt, and the priority of self-healing actions corresponding to different anomaly levels. In particular, the configuration of the self-healing strategy parameters constitutes a safety fence for the system's fault tolerance boundary. The maximum number of self-healing attempts is set to prevent the system from falling into an infinite restart loop due to irreversible physical damage to the radar hardware, thereby avoiding the unnecessary consumption of the UAV's precious onboard power and delaying mission decision-making. The timeout for each self-healing attempt is calibrated based on the hardware startup characteristics and initialization process of the radar equipment, ensuring that each self-healing operation has a clear time window constraint and preventing control thread blocking due to device response suspension. The self-healing action priority mapping table corresponding to different anomaly levels is the core logic for achieving precise resource matching. It ensures that the system can dynamically schedule the most appropriate recovery method according to the severity of the fault, avoiding unnecessary system disturbances caused by "over-treating minor problems" and preventing residual faults caused by "under-treating major problems".

[0046] At the execution level, based on the anomaly level and self-healing action priority, a self-healing control command is sent to the power management unit to execute the corresponding self-healing action. To balance recovery efficiency and system stability, this embodiment adopts a differentiated execution strategy with hardware and software layers. When the anomaly level is mild, a soft reset operation is prioritized. The soft reset operation includes sending a reset command through the radar device's control register. For example, the flight control computer can directly write a specific reset word to the radar's internal status control register via the I2C or SPI bus. This method only resets the radar's digital logic units and cache state without cutting off its main power supply, thus exhibiting millisecond-level fast response characteristics and avoiding surge impacts on the overall power supply network of the UAV. It is particularly suitable for resolving self-recoverable faults such as temporary data quality degradation or cache overflow caused by electromagnetic interference. Conversely, when the anomaly level is severe, a power-off restart operation is performed. The power-off restart operation includes controlling the power management unit to cut off the power supply to the radar device and then powering it back on after a preset time. Specifically, the flight control computer sends a power-off signal to the power management unit (PMU), which then drives the load switch to disconnect the radar power supply circuit. This process is maintained for a preset time sufficient to fully discharge the onboard capacitors before the circuit is reclosed. Although this process is time-consuming and involves instantaneous current fluctuations, it thoroughly clears hardware-level lock-up states, releases static electricity buildup, and forces the firmware to be reloaded from scratch. It is the only effective means of resolving deep-seated faults such as complete communication link interruptions, internal reference voltage drift, or program crashes. This layered execution mechanism ensures a high recovery success rate while minimizing the negative impact of the self-healing process on flight attitude control and multi-sensor synchronous sampling.

[0047] After the self-healing action is completed, the system does not immediately release the logical isolation, but instead enters a strict recovery access interlock phase. Logical isolation is released when preset recovery conditions are met. This process requires monitoring the radar equipment's operating status information again after the self-healing control is executed. When the data frame rate, packet loss rate, communication link status, power supply voltage, and operating mode configuration consistency flag all recover to preset health thresholds and remain stable over multiple consecutive sampling periods, the preset recovery conditions are deemed met, and logical isolation is released. This mechanism is designed to create a creative defense against system oscillations caused by "false recovery." In actual engineering scenarios, radar equipment often experiences a transient instability period after power-on initialization or reset, characterized by drastic fluctuations in output data and parameters not yet converging. If recovery is judged successful based solely on a single successful sampling, this transient spike can easily be misinterpreted as normal operation, leading to the perception fusion layer re-accessing dirty data, which in turn causes frequent switching of flight control decisions and UAV attitude jitter. Therefore, this embodiment introduces stability verification in the time domain, requiring that the five key indicators mentioned above must remain within the healthy range for N consecutive sampling periods (e.g., 3 to 5 consecutive periods). This is equivalent to connecting a digital low-pass filter in series in the recovery access phase, which can effectively filter out signal noise and parameter jitter in the early recovery stage, ensuring that the radar is only allowed to re-participate in flight control decisions after it has truly entered the steady-state operating range, thereby guaranteeing the absolute reliability of the sensing data and the smoothness of system operation.

[0048] Furthermore, if the preset recovery conditions are not met, logical isolation is maintained, and self-healing control is repeatedly executed or a rollback strategy is entered. This means that if the radar still fails to pass the continuous stability verification of the above five-dimensional indicators within the specified timeout period or after reaching the maximum number of self-healing attempts, the system will determine that the device no longer has reliable working capabilities in the current environment. At this time, the system will not blindly continue to attempt self-healing, but will decisively maintain the isolation state to prevent erroneous data contamination, and trigger a mission abort command or automatic return-to-home alarm according to the preset rollback strategy, guiding the UAV to a safe area. This closed-loop fallback mechanism ensures that even in extreme working conditions or in the event of permanent equipment failure, the UAV can still maintain a minimum level of safe behavior, avoiding the risk of overall loss of control due to a single sensor failure.

[0049] Furthermore, the differentiated safety degradation control logic in step S106 is explained in detail. During the isolation of the radar equipment, the risk level of the current mission phase of the UAV is identified. Specifically, the identification of the mission phase risk level does not rely solely on a single timestamp or geographical location, but rather integrates multi-source information such as the real-time relative distance between the UAV and the target guideline, the current flight control system's operating mode flag, and environmental complexity assessment. For example, when the system detects that the UAV is in automatic docking mode and its relative distance to the guideline is less than a preset near-field threshold, it is determined to be a high-risk operation phase; while when the UAV is in waypoint flight mode and far from the target area, it is determined to be a low-risk operation phase. This dynamic risk identification mechanism based on multi-dimensional state vectors ensures that the degradation strategy accurately matches the current actual safety requirements, avoiding strategy mismatch caused by static partitioning.

[0050] If the operation is in a high-risk phase, safety actions such as stopping or retreating should be performed. For the specific scenario of spacer installation, the high-risk phase typically corresponds to the near-field operation phase with the conductor. Safety actions performed at this stage include hovering, retreating to a safe distance, or exiting the conductor corridor. During this phase, because the drone is extremely close to the high-voltage conductor, any blind spot could directly lead to a catastrophic conductor collision. Therefore, this embodiment employs a progressive spatial avoidance strategy: first, a hovering command is immediately executed to eliminate the drone's inertial displacement and prevent collisions with obstacles due to continued movement at the moment of perception failure; then, the drone is controlled to retreat along the original path to a preset safe distance, proactively creating a physical safety buffer zone to buy time and space for subsequent decisions; if the radar has not recovered during the retreat or environmental conditions do not allow continued operation, the exit from the conductor corridor strategy is further implemented, i.e., guiding the drone to ascend vertically or move laterally to an open airspace away from the conductor. This tiered approach, from static to dynamic and from local adjustments to global evacuation, essentially sacrifices temporary operational progress for absolute physical safety, completely avoiding the secondary risks of complex maneuvers in high-risk areas under conditions of perceived untrustworthiness.

[0051] In low-risk operational phases, the system switches to an obstacle avoidance strategy relying on backup sensors and restricts motion parameters. Specifically, when the UAV is in the return or cruise phase, the system automatically switches to a visual obstacle avoidance priority mode and limits the maximum flight speed and maximum acceleration. In open airspace far from guide lines or on predetermined routes, even if radar fails, the visual sensors can still provide sufficient environmental perception to deal with large obstacles at a distance. At this time, the system no longer adopts conservative hovering or retreating strategies, but instead activates the visual obstacle avoidance algorithm to maintain basic navigation and obstacle avoidance functions, ensuring mission continuity. At the same time, to compensate for the shortcomings of visual sensors in ranging accuracy and response speed compared to radar, the system simultaneously applies strict kinematic constraints, limiting the UAV's maximum flight speed and maximum acceleration within preset safety thresholds. The physical significance of reducing speed and acceleration is to extend the control system's reaction time window to sudden obstacles and reduce the kinetic energy impact during potential collisions, thereby maintaining an acceptable safety margin even with degraded perception performance. This combined strategy of "backup sensing + motion constraints" effectively solves the problem of frequent task interruptions or ineffective power consumption caused by excessive conservatism in traditional global degradation schemes during low-risk phases, achieving a dynamic optimal balance between safety and operational efficiency.

[0052] It should be understood that the above division between the near-field operation phase and the return or cruise phase is merely an illustrative example. In practical applications, the risk level of each mission phase can be dynamically adjusted more precisely based on factors such as the voltage level of the transmission line, weather conditions, and the payload status of the UAV. Similarly, specific safety action sequences and motion parameter limits should be adaptively calibrated based on the UAV's dynamic characteristics and operating environment. As long as they adhere to the differentiated control principle of "prioritizing position safety in high-risk situations and considering mission continuity in low-risk situations," they fall within the scope of protection of this application.

[0053] This embodiment provides a radar self-testing and self-healing control device for spacer-mounted unmanned aerial vehicles (UAVs). This device adopts a modular architecture design, solidifying the method flow of the aforementioned embodiments into executable logical functional units. It should be understood that the device described in this embodiment is a functional architecture based on computer programs or firmware logic. Each module can run as an independent software process, be integrated into the same control program, or be partially or entirely implemented by dedicated hardware circuits, as long as the corresponding functional logic can be achieved.

[0054] Specifically, the device includes a status monitoring module, a health determination module, an anomaly isolation module, a self-healing control module, and a degradation control module.

[0055] The status monitoring module is used to monitor the operational status information of the radar equipment in real time. As the data sensing entry point for the entire device, this module periodically collects multi-dimensional status vectors of the radar equipment through the underlying driver interface. In terms of data flow, the status monitoring module transmits the collected raw status information to the health assessment module in real time, providing basic data support for subsequent anomaly identification. Its monitoring content covers multiple dimensions, including data output quality, communication link stability, and internal electrical parameters, ensuring the comprehensiveness and timeliness of status perception.

[0056] The health assessment module determines whether the radar equipment is in an abnormal state based on its operational status information, classifying abnormal states into multiple levels, including at least mild and severe anomalies. This module receives status information from the status monitoring module and maps it to discrete anomaly level identifiers through its built-in classification logic. In control flow interaction, the health assessment module is the system's decision-making center. Its output anomaly level identifiers not only trigger the actions of the anomaly isolation module but also serve as the basis for strategy selection by the self-healing and degradation control modules. This one-to-many signal distribution mechanism ensures that the isolation, self-healing, and degradation subsystems can respond synchronously based on a unified fault semantic, avoiding control conflicts caused by inconsistent judgment criteria.

[0057] The anomaly isolation module is used to logically isolate the radar equipment in response to an anomaly, preventing abnormal data from participating in perception fusion and flight control decisions. Upon receiving an anomaly trigger signal from the health assessment module, this module immediately performs data masking or weight zeroing operations at the perception fusion algorithm layer and generates a radar unavailable status flag, broadcasting it downstream. Functionally, the anomaly isolation module achieves soft decoupling between the faulty sensor and the flight decision system. Its response speed directly determines the timeliness of blocking erroneous data, serving as the first line of defense to prevent malfunctions in the event of a fault.

[0058] The self-healing control module executes differentiated self-healing control based on the anomaly level and performs a recovery access interlock after the self-healing control is executed. Logical isolation is released when preset recovery conditions are met. This module matches the corresponding self-healing action sequence from a preset strategy library based on the anomaly level identifier provided by the health assessment module and sends control commands to the power management unit or radar register. After the self-healing action is completed, this module is also responsible for performing strict recovery access verification. Only when all indicators continuously and stably meet the standards is a release isolation command sent to the anomaly isolation module. This closed-loop control logic ensures that the radar only reconnects to the system after truly recovering its health, effectively preventing system oscillations caused by false recovery.

[0059] The degradation control module identifies risk characteristics of the current mission phase during radar equipment isolation and executes differentiated safety degradation control based on these characteristics. This module works in parallel with the anomaly isolation module, proactively acquiring the UAV's current mission context information after sensing a radar unavailability status indicator. Based on the identified risk level, the degradation control module outputs corresponding safety maintenance commands to the flight control system, such as triggering hover reversal in high-risk phases or switching to visual obstacle avoidance and limiting motion parameters in low-risk phases. Through collaboration with the self-healing control module, the degradation control module fills the safety perception gap during radar failure, ensuring the UAV's behavioral safety under fault conditions.

[0060] In practical engineering deployments, the aforementioned modules can be flexibly configured within the UAV's flight control computer, utilizing the computer's existing computing resources for integrated operation. Alternatively, they can be deployed in a separate onboard monitoring unit, interacting with the flight control system via a bus to achieve physical isolation between health monitoring and flight control functions, further enhancing system reliability and anti-interference capabilities. Regardless of the deployment method, the data flow and control flow relationships between the modules remain consistent, ensuring complete correspondence between the overall device functionality and the method described above.

[0061] This embodiment provides a spacer bar intelligent installation drone. This drone serves as the physical carrier of the aforementioned radar self-testing and self-healing control method, achieving autonomous fault tolerance in complex power transmission line environments through hardware and software co-design. Specifically, the spacer bar intelligent installation drone includes a drone body, radar equipment, a power management unit, and a flight control computer. The flight control computer is configured to execute the radar self-testing and self-healing control method for the spacer bar intelligent installation drone as described in the foregoing embodiment.

[0062] At the hardware architecture level, the UAV airframe provides the platform and power foundation, and its structural design must meet the wind resistance, image stabilization, and insulation safety requirements for near-field operations on power transmission lines. The radar equipment, mounted on the UAV airframe, serves as the core sensor for obstacle perception and auxiliary positioning. Its selection can encompass one or more combinations of millimeter-wave radar, lidar, or ultrasonic sensors, as long as it can output observation data such as distance, speed, and reflection intensity required by the aforementioned methods. The power management unit (PMU) is independent of the flight control computer and is specifically responsible for fine-grained power supply control of the radar equipment. It connects to the flight control computer via a high-speed communication bus (such as CAN bus, Ethernet, or SPI / I2C interface) to receive self-healing control commands from the flight control computer and provide feedback on the execution status. This independent configuration of the PMU allows the flight control computer to perform power-off restarts or soft resets on the radar equipment independently without cutting off power to itself or other critical subsystems, providing the necessary physical actuator for the implementation of the differentiated self-healing strategy in the aforementioned embodiments.

[0063] At the software configuration and control logic level, the flight control computer, as the decision-making center of the entire system, has a computer program embedded in its internal memory that implements the aforementioned radar self-testing and self-healing control methods. When the processor executes this program, it constitutes the functional units such as the status monitoring module, health judgment module, anomaly isolation module, self-healing control module, and degradation control module in the aforementioned embodiments. It should be understood that the execution of the aforementioned methods by the flight control computer is not limited to a single processing chip or single-board computer. In actual engineering, a distributed architecture with separate main flight control and mission computer can also be adopted, or a heterogeneous computing platform using FPGA and ARM for collaborative processing can be used. As long as the overall system has the ability to collect radar status in real time, run hierarchical judgment algorithms, issue self-healing commands, and switch degradation strategies, it falls within the protection scope of this application.

[0064] To address the unique challenges of strong electromagnetic interference and high reliability requirements in spacer installation operations, this embodiment incorporates tailored adaptive designs for the entire UAV. For example, the power supply cable between the radar equipment and the power management unit uses shielded twisted-pair cable with a magnetic ring filter to suppress the interference of corona discharge noise from high-voltage conductors on the radar power quality. The communication link between the flight control computer and the radar equipment incorporates CRC checksum and retransmission mechanisms to prevent false alarms or command loss due to electromagnetic pulses. These hardware-level anti-interference measures, combined with the aforementioned software-level self-checking and self-healing methods, form a complementary dual-protection system: the hardware design reduces the probability of anomalies, while the software approach provides rapid recovery and a safety net when anomalies are unavoidable. Through this integrated hardware and software system-level design, this intelligent spacer installation UAV can operate stably for extended periods in the near-field region of transmission lines with voltage levels of 500kV and above. This effectively solves the engineering problem of mission interruptions or safety accidents caused by frequent radar interference in traditional UAVs, significantly improving the equipment level and operational efficiency of intelligent power grid construction.

[0065] This embodiment provides a computer-readable storage medium storing a computer program. When executed by a processor, the computer program implements the radar self-testing and self-healing control method for a spacer-mounted intelligent UAV as described in any of the preceding embodiments one to three. It should be understood that the computer-readable storage medium in this embodiment is a software carrier of the aforementioned method flow, which solidifies the abstract control logic into a distributable, storable, and executable data entity, thereby achieving a degree of decoupling between the technical solution and the specific hardware platform.

[0066] Specifically, the computer-readable storage medium can take many forms, including both non-volatile and volatile memory. For example, non-volatile memory can be read-only memory, programmable read-only memory, electrically erasable programmable read-only memory, flash memory, disk, or optical disk; volatile memory can be random access memory, used as external cache or runtime memory. In practical engineering deployment scenarios, the storage medium can be an embedded flash memory chip integrated within the UAV flight control computer, used to embed the factory-version self-testing and self-healing algorithm; it can also be a portable storage device such as an independent SD card or USB flash drive, used for firmware flashing and log export during on-site maintenance; or it can be a digital image file on a ground station server or cloud platform, supporting remote OTA upgrades of the UAV swarm via wireless network. Regardless of its physical form, as long as it can carry and output the instruction sequence for implementing the aforementioned methods, it falls within the protection scope of this application.

[0067] To meet the stringent requirements of data security and system reliability in power transmission line operations, the computer program stored on this medium can further include security verification and version management modules. For example, program code can be digitally signed or encrypted when written to the medium. Before loading the program, the flight control computer must verify the legality and integrity of the signature to prevent control logic failure due to malicious tampering or transmission errors. Simultaneously, the medium can store multiple versions of the algorithm program and corresponding configuration files, supporting dynamic loading of adapted versions based on batch differences in radar hardware or changes in the operating environment, enhancing system compatibility and maintainability. By encapsulating the radar self-testing and self-healing control method into a standardized software product, it not only facilitates rapid portability and reuse between different types of spacer-mounted UAVs but also provides a standardized data foundation for subsequent algorithm iteration, fault backtracking, and full lifecycle maintenance, significantly improving the commercial application value and engineering implementation efficiency of the technical solution.

[0068] This embodiment provides a complete operational example in a 500kV transmission line spacer installation scenario, aiming to verify the effectiveness of the radar self-testing and self-healing control methods in the aforementioned embodiments through specific parameters and timing logic. It should be understood that all specific values, equipment models, and action parameters mentioned in this embodiment are preferred examples, used only to help those skilled in the art understand the engineering implementation of the technical solution, and are not necessary limitations on the scope of protection of this application. In practical applications, relevant thresholds and strategies can be adaptively adjusted according to the characteristics of the UAV platform, radar sensor specifications, and operational environment requirements.

[0069] In this application example, the operating platform uses a certain type of industrial-grade hexacopter UAV, equipped with a 77GHz millimeter-wave radar as the core obstacle avoidance and positioning perception unit. The flight control computer has a main frequency of 1.2GHz, and the power management unit has independent channel control capabilities. The operating environment is a 500kV ultra-high voltage transmission line corridor, with a complex electromagnetic environment and interference factors such as strong winds and hardware reflections. The following is combined with... Figure 1 The flowcharts shown illustrate the entire system response process under two typical abnormal operating conditions.

[0070] The first operating scenario is the self-healing process for minor anomalies during the cruise phase. When the UAV is in the cruise phase heading to the work site, the flight control computer executes the status monitoring in step S101 at a frequency of 50Hz. At a certain moment, it is detected that the radar data frame rate drops to 22Hz and the packet loss rate rises to 8%, but the communication link remains online and the power supply voltage remains at the normal level of 5.0V. Based on preset logic, the health judgment module determines that this state is a minor anomaly due to degraded data quality, but it is still partially usable. Subsequently, the anomaly isolation module resets the radar observation weight to zero at the perception fusion layer and outputs a radar unavailable flag. The entire process takes less than 20 milliseconds, effectively preventing low-quality data from contaminating the visual-GPS fusion positioning algorithm.

[0071] Following this, the self-healing control module prioritized a soft reset strategy based on the mild anomaly level. The flight control computer sent a reset command to the radar control register via the I2C bus. This operation did not cut off the radar power supply, but only reset the internal digital logic and cache, taking approximately 150 milliseconds. After the reset was completed, the system entered the recovery access interlock verification phase. In the subsequent three consecutive sampling cycles (i.e., a 60-millisecond window), the radar frame rate was observed to steadily rise to 28Hz, the packet loss rate dropped to below 1%, the link status was normal, the voltage was stable, and the configuration consistency flag was set to 1. Since all five indicators were within the preset health thresholds and met the continuous stability requirements, the system determined that the recovery was successful, immediately released the logic isolation, and restored the radar weights to normal values. From the anomaly trigger to complete recovery, the entire process took less than 1 second. The UAV did not change its cruise attitude or speed throughout the process, achieving imperceptible fault self-healing. This fully demonstrates the efficiency of the differentiated self-healing strategy in handling transient interference and its ability to ensure mission continuity.

[0072] The second operating scenario involves a severe anomaly and safety degradation during the near-field operation phase of the conductor. When the UAV enters the docking phase approximately 1.5 meters from the conductor, it experiences strong corona discharge interference, causing a sudden interruption of the radar communication link for over 3 seconds, and the power supply voltage drops instantly to 3.8V. The health assessment module immediately classifies this as a severe anomaly. At this point, the system triggers two control branches in parallel: in the self-healing branch, the self-healing control module determines that a soft reset is ineffective and directly instructs the power management unit to perform a power-off restart operation, cutting off the radar power supply for 5 seconds before resuming power to completely clear the hardware lock-up state; in the safety degradation branch, the degradation control module identifies the current high-risk near-field operation phase of the conductor and immediately takes over flight control, first issuing a hover command to eliminate inertial displacement, and then controlling the UAV to retreat along the original path at a speed of 0.5m / s to a preset safe distance of 2 meters from the conductor.

[0073] During the 5-second window after the radar power failure and restart, the UAV maintained hovering stability entirely through its visual sensors and IMU, and actively moved away to create a physical safety buffer zone, completely avoiding the risk of contact with the line due to blind spots. However, during the access verification after power-on, although the radar link was restored, the power supply voltage fluctuated between 4.2V and 4.6V for five consecutive sampling cycles, failing to meet the continuous and stable access conditions. Based on this, the system determined that self-healing had failed and maintained a logical isolation state. Given that the preset maximum number of self-healing attempts (3) had been reached, the self-healing control module triggered a fallback strategy, and the flight control computer automatically switched to return-to-home mode, guiding the UAV safely away from the work area.

[0074] The two complete application examples above demonstrate that the technical solution provided in this application exhibits refined adaptive control capabilities when facing operational scenarios with varying degrees of faults and risks. Under mild anomalies, through soft reset and rigorous continuous stability verification, it achieves millisecond-level rapid recovery while avoiding system oscillations caused by false recovery. In severe anomalies coupled with high-risk scenarios, through the synergy of power-off restart and proactive retreat avoidance, it prioritizes absolute safety while attempting deep repair, and orderly executes return-to-home as a fallback in case of repair failure. This closed-loop control mechanism, which deeply integrates anomaly classification, self-healing layering, and scenario-specific degradation, effectively solves the technical challenges of poor reliability and low safety of traditional single restart or global degradation strategies in complex electromagnetic environments, significantly enhancing the engineering practical value of spacer bar intelligent installation drones.

[0075] To more clearly illustrate the necessity and advancement of the access restoration interlock mechanism in this application's technical solution, this comparative example provides a radar self-testing and self-healing control scheme lacking continuous stability verification, demonstrating the technical defects caused by solely relying on instantaneous threshold judgment. Unlike the aforementioned embodiments, which require stability over multiple consecutive sampling periods before releasing logical isolation, the access restoration condition in this comparative example is determined based solely on a single sampling result. Specifically, after the self-healing control is executed, the system monitors the radar equipment's operating status information in real time. Once the data frame rate, packet loss rate, communication link status, power supply voltage, and operating mode configuration consistency flag all reach preset health thresholds at a certain sampling moment, the system immediately determines that the preset recovery condition is met and immediately releases the logical isolation, restoring the fusion weight of the radar observation data to its normal value.

[0076] This strategy of triggering recovery based solely on a single successful sampling is highly susceptible to false recovery phenomena in engineering practice due to the transient characteristics of radar equipment, leading to serious system oscillation problems. Specifically, during the initialization phase after a power-off restart or soft reset, the radar equipment often requires a certain convergence time for internal clock synchronization, signal processing unit calibration, and firmware loading. During this period, the radar's output data stream is usually unstable, potentially exhibiting intermittent data spikes, drastic frame rate fluctuations, or brief parameter overshoots. If recovery admission verification relies solely on a single isolated point in time, the system may mistakenly assume the radar has fully recovered if the sampling time falls within the brief normal window between two abnormal spikes.

[0077] The direct consequence of this is a drastic jump in the data input of the perception fusion layer. When the system releases the isolation based on an erroneous recovery judgment, the radar weights are reloaded, but the radar may output abnormal or invalid data again in the next sampling cycle. This rapid switching from "complete shielding" to "full participation" and then to "forced re-isolation" in a very short time can cause a precipitous change in the output of the multi-source sensor fusion algorithm. For example, in obstacle avoidance scenarios, the obstacle distance calculated by the fusion positioning module may suddenly change from a safe value to zero or infinity within milliseconds. After receiving such high-frequency and large-amplitude perception errors, the flight control computer will interpret them as a rapid change in the environment, thus frequently issuing contradictory flight control commands, such as alternating between emergency braking and acceleration evasion maneuvers.

[0078] This control command oscillation induced by false recovery manifests physically as abnormal shaking of the UAV body, attitude instability, and even overload protection of the power system. Especially in the near-field region of the conductor during spacer bar installation, any slight loss of attitude control can directly lead to a line contact accident. In contrast, the stability verification mechanism using multiple consecutive sampling periods employed in the aforementioned embodiments of this application essentially connects a time-domain low-pass filter in series during the recovery access phase. This mechanism forces the radar state to exhibit continuous determinism in the time dimension, effectively filtering out transient noise and intermittent faults in the initial recovery phase, ensuring that the radar is only allowed to re-enter the decision loop once it has truly entered the steady-state operating region. As demonstrated by the counter-example in this comparison, introducing continuous stability verification is not simply parameter optimization, but a key means to solve the specific technical problem of false recovery and system oscillation during radar self-healing, playing an irreplaceable role in ensuring the operational safety of UAVs in complex electromagnetic environments.

[0079] To further illustrate the necessity and advancement of the differentiated safety degradation control based on the risk characteristics of the mission phase in this application, this comparative example also provides a radar self-checking and self-healing control scheme employing a globally unified degradation strategy. Unlike the aforementioned embodiments that dynamically adjust the control strategy according to the risk level of the mission phase, the safety degradation logic in this comparative example does not distinguish between the specific operating environment or mission node currently occupied by the UAV. Specifically, once an anomaly is detected in the radar equipment and it is logically isolated, regardless of whether the system determines that it is currently in a high-risk wire-guided near-field operation phase or a low-risk return-to-home or cruise phase, it unconditionally executes the same preset degradation action, that is, immediately reducing the flight speed to zero and maintaining a hovering state until the radar returns to normal or the timeout return-to-home mechanism is triggered.

[0080] This one-size-fits-all degradation strategy, lacking scene awareness, exposes serious safety flaws and efficiency bottlenecks in the specific application scenario of spacer bar installation. First, when the drone is in the near-field operation phase with the high-voltage power line (such as during docking or position fine-tuning), its physical distance to the high-voltage power line is extremely close, and the surrounding electromagnetic environment is extremely complex. If only a hovering strategy is implemented at this time, the drone will be continuously exposed to the strong electric field interference area. Not only will the onboard battery power be rapidly consumed to maintain attitude stability, increasing the risk of forced landing due to battery depletion, but prolonged close-range hovering also significantly increases the probability of accumulated electrostatic discharge damaging electronic components. More importantly, at the moment of perception failure, the drone often still retains some forward inertia or drifts due to wind conditions. Simple deceleration hovering commands cannot actively eliminate this spatial approaching tendency. Without radar data correction, relying solely on the IMU for short-term blind hovering is highly susceptible to position estimation divergence due to integration errors, causing the drone to unknowingly cross the safety boundary and ultimately cause a power line collision. In contrast, this application proactively retreats to a safe distance or exits the conductor corridor when a high-risk phase is identified. Essentially, it constructs a physical safety barrier by actively moving away from the conductor, fundamentally avoiding the multiple overlapping risks caused by waiting in place.

[0081] Secondly, even during the cruise or return phases far from the guide wire, a globally uniform hovering strategy appears overly conservative and unnecessary. In these open airspaces, although the radar fails, the visual sensors typically still possess sufficient environmental awareness redundancy. The comparative example forces the UAV to hover urgently during cruise, directly leading to unexpected mission interruptions and a significant decrease in overall efficiency. In contrast, this application switches to a visual obstacle avoidance priority mode during low-risk phases and limits maximum flight speed and acceleration. This utilizes the capabilities of remaining sensors to maintain basic navigation continuity while ensuring a safety margin through kinematic constraints, achieving an optimal balance between safety and mission efficiency in fault conditions.

[0082] In summary, this comparative example fully reveals the limitations of a globally unified degradation strategy in dealing with complex and ever-changing operational scenarios: in the near-field high-risk zone, insufficient action force fails to guarantee absolute safety, while in the far-field safe zone, overly conservative actions sacrifice operational continuity. This conversely proves that the risk identification and differentiated degradation control introduced in this application are not simply a functional addition, but rather a targeted technical improvement made for the unique working conditions of spacer bar intelligent installation UAVs. Only through refined scenario adaptation can the rigid safety requirements of the anti-collision line and the flexible efficiency requirements of mission assurance be simultaneously met within the radar failure window. This is the key difference between this application and existing technologies, and the key to its substantial technological advancement.

[0083] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A radar self-testing and self-healing control method for a spacer bar intelligent installation UAV, characterized in that, include: Real-time monitoring of radar equipment's operational status; Based on the working status information, it is determined whether the radar device is in an abnormal state, and the abnormal state is classified into multiple abnormal levels, including at least mild abnormality and severe abnormality. In response to determining that the radar device is in the abnormal state, the radar device is logically isolated to prevent abnormal data from participating in perception fusion and flight control decisions; Based on the anomaly level, differentiated self-healing control is executed, and after the self-healing control is executed, a recovery access interlock is executed. When the preset recovery conditions are met, the logical isolation is released. During the isolation of the radar equipment, risk characteristics of the current mission phase are identified, and differentiated security degradation control is performed based on the risk characteristics.

2. The method according to claim 1, characterized in that, The execution of differentiated self-healing control based on the anomaly level, and the execution of restoration access interlock after the self-healing control is executed, includes: Set self-healing strategy parameters, including the maximum number of self-healing attempts, the timeout for a single self-healing attempt, and the priority of self-healing actions corresponding to different anomaly levels. Based on the anomaly level and the self-healing action priority, a self-healing control command is sent to the power management unit to execute the corresponding self-healing action. When the maximum number of self-healing attempts is reached and the radar equipment still fails to recover, a mission abort or return-to-base alarm strategy is triggered.

3. The method according to claim 1, characterized in that, The process of identifying risk characteristics at the current task stage and performing differentiated security degradation control based on those risk characteristics includes: While the radar equipment is isolated, the risk level of the current mission phase of the UAV is identified. If the operation is in a high-risk phase, execute safety actions such as stopping or reversing. If the operation is in a low-risk phase, switch to an obstacle avoidance strategy that relies on backup sensors and limit motion parameters.

4. The method according to claim 2, characterized in that, The step of sending a self-healing control command to the power management unit to execute the corresponding self-healing action based on the anomaly level and the self-healing action priority includes: When the anomaly level is a minor anomaly, a soft reset operation is performed first. The soft reset operation includes sending a reset command through the control register of the radar device. When the anomaly level is severe, a power-off restart operation is performed. The power-off restart operation includes controlling the power management unit to cut off the power supply to the radar equipment and then powering it back on after a preset time.

5. The method according to claim 2, characterized in that, The step of releasing the logical isolation when the preset recovery condition is met includes: After the self-healing control is executed, the working status information of the radar device is monitored again. When the data frame rate, packet loss rate, communication link status, power supply voltage and working mode configuration consistency flag are all restored to the preset health threshold and remain stable in multiple consecutive sampling periods, it is determined that the preset recovery condition is met and the logical isolation is released. If the preset recovery conditions are not met, the logical isolation is maintained, and the self-healing control is repeated or a rollback strategy is entered.

6. The method according to claim 3, characterized in that, If the operation is in the high-risk phase, a stop or rollback safety action is performed; if the operation is in the low-risk phase, an obstacle avoidance strategy relying on backup sensors is switched, and motion parameters are limited, including: If the current task stage is the conductor near-field operation stage, perform safety actions such as hovering, retreating to a safe distance, or exiting the conductor corridor; If the current mission phase is the return or cruise phase, switch to visual obstacle avoidance priority mode and limit the maximum flight speed and maximum acceleration.

7. The method according to claim 1, characterized in that, The logical isolation of the radar equipment to prevent abnormal data from participating in perception fusion and flight control decisions includes: The radar observation data is shielded at the perception fusion layer, or the radar observation weight is reduced to zero, and a radar unavailable status flag is output.

8. A radar self-testing and self-healing control device for a spacer-mounted intelligent UAV, characterized in that, include: The status monitoring module is used to monitor the working status information of the radar equipment in real time. The health determination module is used to determine whether the radar device is in an abnormal state based on the working status information, and to classify the abnormal state into multiple abnormal levels, including at least mild abnormality and severe abnormality. An anomaly isolation module is used to logically isolate the radar device in response to determining that the radar device is in the abnormal state, so as to prevent abnormal data from participating in perception fusion and flight control decision-making. The self-healing control module is used to perform differentiated self-healing control based on the anomaly level, and to perform a recovery access interlock after the self-healing control is executed. When the preset recovery conditions are met, the logical isolation is released. The degradation control module is used to identify risk characteristics of the current mission phase during the isolation of the radar equipment, and to perform differentiated security degradation control based on the risk characteristics.

9. A spacer bar intelligent installation drone, characterized in that, It includes an unmanned aerial vehicle (UAV) body, radar equipment, a power management unit, and a flight control computer, wherein the flight control computer is configured to execute the radar self-testing and self-healing control method for a spacer-mounted UAV as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the radar self-testing and self-healing control method for spacer bar intelligent installation UAV as described in any one of claims 1 to 7.

Images