An unmanned aerial vehicle autonomous mission flight control system and unmanned aerial vehicle

Abstract

The present application relates to the technical field of unmanned aerial vehicle flight control, and particularly relates to an autonomous task flight control system of unmanned aerial vehicle and unmanned aerial vehicle; a task planning module is used for presetting autonomous flight global route with task target space-time anchor point, dividing independent task section with accounting window, and establishing scene-adaptive task execution effectiveness dynamic quantitative evaluation system; a task effectiveness calculation module is used for collecting space-time-aligned task load operation data and flight pose data, and outputting linkage comparison result with control threshold; a flight path iterative correction module is used for taking task execution effectiveness comparison result as core control input; a flight closed-loop execution module is used for controlling unmanned aerial vehicle to execute preset global route or iterative correction flight path; through taking task effectiveness as core input, through section-level planning, real-time calculation, iterative correction and closed-loop verification, autonomous closed-loop guarantee and whole-process traceability of unmanned aerial vehicle task quality are realized.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) flight control technology, and in particular to an autonomous mission flight control system for UAVs and the UAV itself. Background Technology

[0002] Currently, drones have been widely used in fields such as power line inspection, oil and gas pipeline inspection, agricultural plant protection, and aerial surveying and mapping. Their autonomous mission flight control system usually adopts the control paradigm of "preset route + track tracking". In the existing technology, the drone flight control system takes track tracking accuracy as the core control target, and uses GPS / RTK positioning, inertial navigation, visual odometry and other means to make the actual flight trajectory of the drone as close as possible to the preset route.

[0003] However, the aforementioned existing technologies generally suffer from the following problems: First, the existing control logic takes "accurate flight path" as its core objective. Even in schemes used for mission quality correction, it does not fundamentally establish a closed-loop control system with the actual mission completion quality as the core evaluation standard. Simultaneously, even when the flight path perfectly follows the preset route, dynamic factors such as target obstruction by the environment, deviations between the actual and preset target positions, changes in lighting conditions, and differences in crop growth status can still lead to mission failures—for example, missing key equipment during power line inspections or missed or over-sprayed areas during plant protection operations. Existing systems cannot autonomously detect and correct such mission quality issues. Furthermore, existing technologies lack a technical solution that uses "mission execution effectiveness" as the core control input and forms a closed-loop logic of "compliance judgment → iterative correction → re-judgment," making it difficult to guarantee the mission completion quality of UAVs in complex dynamic environments. Summary of the Invention

[0004] The purpose of this invention is to provide an autonomous mission flight control system for unmanned aerial vehicles (UAVs) and the UAV itself. This invention aims to solve the technical problem that existing control logics in the prior art, which take "flight accuracy" as the core objective and are used for mission quality correction, do not fundamentally establish a closed-loop control system with the actual mission completion quality as the core evaluation standard.

[0005] To achieve the above objectives, this invention employs an autonomous mission flight control system for unmanned aerial vehicles (UAVs), comprising a mission planning module, a mission effectiveness calculation module, a trajectory iterative correction module, and a flight closed-loop execution module; wherein: The task planning module is used to preset the global autonomous flight route with task target spatiotemporal anchor points, divide the independent task segments with accounting windows, establish a dynamic quantitative evaluation system for the effectiveness of task execution adapted to the scenario, and set multi-dimensional linkage control thresholds. The task effectiveness calculation module is used to collect spatiotemporally aligned task payload operation data and flight attitude data. Relying on the dynamic quantitative evaluation system, the airborne terminal calculates the task execution effectiveness of the current task segment in real time and outputs the linkage comparison results with the control threshold. The trajectory iterative correction module is used to establish an inverse solution mapping model between the effectiveness difference and trajectory parameters, with the mission execution effectiveness comparison result as the core control input, to generate an iteratively corrected trajectory that satisfies multiple constraints, and to trigger closed-loop flight action commands. The flight closed-loop execution module is used to control the UAV to execute a preset global route or iteratively correct the trajectory. After completing the closed-loop verification of the effectiveness of the entire mission segment, it generates and transmits back the mission execution key feature data with spatiotemporal tags and the trajectory correction link record.

[0006] The task planning module includes an anchor point route preset unit, a task segment spatiotemporal division unit, a dynamic effectiveness system construction unit, and a multi-dimensional threshold linkage unit; wherein: The anchor point preset unit is used to plan a global flight path with target spatiotemporal anchor points according to the geographical coordinates, physical size and observation requirements of the mission target, and to fix the waypoint three-dimensional coordinates, reference flight altitude, cruise speed, payload field of view and exposure trigger logic. The task segment spatiotemporal partitioning unit is used to divide the global flight path into multiple continuous task segments with independent validity calculation windows based on the spatial clustering results and time series of the task objectives. Each task segment is bound to a unique set of target spatiotemporal anchor points. The dynamic effectiveness system construction unit is used to establish a dynamic quantitative evaluation system for task execution effectiveness that is linked to environmental complexity and target importance for two core operation scenarios: inspection and plant protection, and to solidify the core evaluation indicators and dynamic weight calculation logic. The multi-dimensional threshold linkage unit is used to bind the dynamic quantitative evaluation system and link the effective achievement threshold, the maximum number of iterations for single-task segment trajectory correction, the minimum power threshold for safe return, and the environmental complexity trigger threshold.

[0007] The task segment spatiotemporal partitioning unit includes a target spatiotemporal clustering subunit, a calculation window adaptation subunit, and an anchor point binding subunit; wherein: The target spatiotemporal clustering subunit is used to use an adaptive clustering method based on spatial density to cluster all task targets covered by the global flight path according to their geographic coordinates, and at the same time match the time series of task execution to output multiple sets of target sets with spatiotemporal boundaries. The calculation window adaptation subunit is used to adapt the effective calculation sliding window of the task segment length and corresponding duration of 100m-500m according to the spatial span, number of targets and observation requirements of each target set, and output the adapted segmentation rules. The anchor point binding subunit is used to bind the spatiotemporal anchor points of each target set to a single independent task segment, and output segmented routes with target spatiotemporal anchor points and calculation windows.

[0008] The dynamic validity system construction unit includes a scenario indicator solidification subunit, a dynamic weight calculation subunit, and a calculation logic solidification subunit; wherein: The scenario index solidification subunit is used to solidify the target shooting coverage, target feature clarity, unobstructed image ratio, and target feature integrity as core evaluation indicators for inspection-type operation scenarios; and to solidify the target area coverage integrity, droplet adhesion uniformity, missed spray rate, and re-spray rate as core evaluation indicators for plant protection operation scenarios. The dynamic weight calculation subunit is used to pre-set weight benchmarks based on target importance, environmental complexity, and flight conditions using the analytic hierarchy process, and to adjust the dynamic weight ratio of each core evaluation indicator in real time in conjunction with environmental complexity. The weight adjustment range is 10%-60%. The calculation logic solidification subunit is used to calculate the frame-level effectiveness within a single calculation window through a weighted summation algorithm, and then output the cumulative task execution effectiveness comprehensive score of the current task segment through a moving average algorithm.

[0009] The task effectiveness calculation module includes a spatiotemporal aligned data acquisition unit, an airborne lightweight feature extraction unit, a real-time effectiveness calculation unit, and a threshold-linked comparison unit; wherein: The spatiotemporal alignment data acquisition unit is used to synchronously acquire the operational data of the airborne mission payload and the attitude data of the flight control system based on the spatiotemporal anchor point of the mission segment, and uniformly mark all data with millisecond-level timestamps and centimeter-level position stamps; among which, the operational data includes visible light, multispectral data, and plant protection nozzle flow and pressure data, and the attitude data includes three-dimensional coordinates, flight altitude, attitude angle, and velocity vector. The airborne lightweight feature extraction unit is used to extract target features and perform semantic segmentation on the collected operation data through a lightweight semantic segmentation model based on deep learning, identify target regions, non-target regions, and occluded regions, and output a standardized feature dataset. The real-time validity calculation unit is used to perform frame-level validity calculation on the standardized feature dataset based on the dynamic quantitative evaluation system, and then calculate the cumulative validity comprehensive score of the current task segment in real time through the sliding window algorithm. The threshold linkage comparison unit is used to compare the real-time calculated comprehensive score of effectiveness with the threshold set in the linkage frame by frame, and output two-level judgment results of compliance and non-compliance, and synchronously output the effectiveness difference data.

[0010] The task effectiveness calculation module also includes a precise positioning unit for areas that do not meet the standards. The sub-unit for precise location of non-compliant areas is used to locate the three-dimensional geographic boundaries, area, non-compliance type and cause of non-compliant areas for frame-level data within the non-compliant calculation window through pixel-level semantic segmentation and inverse geographic coordinate calculation, and output a feature dataset of non-compliant areas with spatiotemporal labels.

[0011] The trajectory iterative correction module includes a trajectory parameter inverse calculation unit, a multi-constraint supplementary flight trajectory generation unit, and a closed-loop execution scheduling unit; wherein: The inverse calculation unit for track parameters is used to establish an inverse calculation mapping model between the effective difference and track parameters, with the effective difference and the non-compliant area feature dataset as the core inputs. It matches the non-compliant type and outputs the optimal track correction parameters, which include flight altitude, flight speed, lateral offset of the flight path, load field of view, and flight path overlap rate. The multi-constraint flight path generation unit is used to generate an iterative flight path that satisfies global optimum based on the optimal path correction parameters, combined with UAV flight dynamics constraints, remaining battery power constraints, field of view coverage constraints, and safety obstacle avoidance constraints, through a heuristic path search algorithm. The closed-loop execution scheduling unit is used to receive the effectiveness comparison and judgment results and generate closed-loop flight action commands: for task segments that meet the effectiveness criteria, it generates continuous flight commands that are extended to the next task segment; for task segments that do not meet the effectiveness criteria, it generates control commands to pause the global flight path and execute iterative supplementary flight paths.

[0012] The inverse calculation unit for trajectory parameters includes a single-factor precise correction subunit, a multi-factor coupled correction subunit, and an iteration over-limit termination subunit; wherein: The single-factor precision correction subunit is used to precisely correct the corresponding flight path parameters for a single type of non-compliance problem: when the target clarity is not up to standard, the reference flight altitude is lowered by 10%-30% and the flight speed is reduced by 40%; when the target coverage is not up to standard, the lateral overlap rate of the flight path is increased by 15%-30% and the payload field of view is adjusted; when the image occlusion is not up to standard, the lateral offset of the flight path is adjusted and the flight altitude is optimized. The multi-factor coupled correction subunit is used to solve multi-parameter coupled correction schemes for multiple types of composite non-compliance problems through multi-objective optimization algorithms, and generate optimal trajectory parameters that take into account both correction effect and energy consumption. The iteration over-limit termination subunit is used to terminate the current mission segment's supplementary flight process when the mission segment's effectiveness still fails to meet the target after the number of supplementary flight iterations reaches the preset maximum correction number. It pushes the full feature data of the non-targeted area to the ground control station, triggers an anomaly alarm, and generates an instruction to restore the execution of the global flight path.

[0013] The flight closed-loop execution module also includes a power emergency dispatch unit; The power emergency dispatch unit is used to compare the remaining airborne power with the preset minimum power threshold for safe return in real time, and establish a multi-objective optimization model for the completion of tasks with remaining power. When the remaining power is lower than the threshold, the execution order of task segments is dynamically adjusted according to the target weight priority in the effectiveness evaluation system. The task segments corresponding to high-weight targets are completed first to maximize the global task execution effectiveness. After completing the high-priority tasks, an automatic return command is generated, and the effectiveness calculation record and trajectory correction link data are retained.

[0014] The present invention also provides an unmanned aerial vehicle (UAV) including the aforementioned autonomous mission flight control system for the UAV.

[0015] This invention discloses an autonomous mission flight control system and an unmanned aerial vehicle (UAV). The mission planning module is used to preset a global autonomous flight path with mission target spatiotemporal anchor points, divide independent mission segments with calculation windows, establish a scenario-adaptive dynamic quantitative evaluation system for mission execution effectiveness, and set multi-dimensional linkage control thresholds. The mission effectiveness calculation module is used to collect spatiotemporally aligned mission payload operation data and flight attitude data, and, based on the dynamic quantitative evaluation system, calculate the mission execution effectiveness of the current mission segment in real time on the airborne end, outputting a linkage comparison result with the control thresholds. The trajectory iteration correction module is used to establish an inverse calculation mapping model between effectiveness difference and trajectory parameters, using the mission execution effectiveness comparison result as the core control input, generating an iteratively corrected trajectory that satisfies multiple constraints, and triggering closed-loop flight action commands. The flight closed-loop execution module is used to control the UAV to execute the preset global path or iteratively corrected trajectory, and after completing the closed-loop verification of the effectiveness of the entire mission segment, generate and transmit back key mission execution feature data with spatiotemporal tags and trajectory correction link records. Through a five-stage closed-loop control process of "segment-level planning - real-time calculation - compliance judgment - iterative correction - closed-loop verification", with the effectiveness of task execution as the core control input, the UAV has achieved the technical effect of autonomously ensuring the quality of task completion in dynamic and complex environments. It completely replaces the existing flight control logic that focuses on the accuracy of trajectory tracking, and significantly improves the reliability and quality of operations in scenarios such as power line inspection and agricultural plant protection. Attached Figure Description

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

[0017] Figure 1 This is a schematic diagram of the autonomous mission flight control system of the UAV of the present invention.

[0018] Figure 2 This is a schematic diagram of the task planning module of the present invention.

[0019] Figure 3 This is a schematic diagram of the task segment spatiotemporal partitioning unit of the present invention.

[0020] Figure 4 This is a schematic diagram of the dynamic effectiveness system construction unit of the present invention.

[0021] Figure 5 This is a schematic diagram of the task validity calculation module of the present invention.

[0022] Figure 6 This is a schematic diagram of the trajectory iterative correction module of the present invention.

[0023] Figure 7 This is a schematic diagram of the inverse calculation unit for trajectory parameters of the present invention.

[0024] 1-Task Planning Module, 101-Anchor Point Preset Route Unit, 102-Task Segment Spatiotemporal Division Unit, 1021-Target Spatiotemporal Clustering Subunit, 1022-Accounting Window Adaptation Subunit, 1023-Anchor Point Binding Subunit, 103-Dynamic Validity System Construction Unit, 1031-Scenario Indicator Solidification Subunit, 1032-Dynamic Weight Calculation Subunit, 1033-Accounting Logic Solidification Subunit, 104-Multi-Dimensional Threshold Linkage Unit, 2-Task Validity Calculation Module, 201-Spatiotemporal Alignment Data Acquisition Unit, 2 02-Airborne lightweight feature extraction unit, 203-Real-time validity calculation unit, 204-Threshold linkage comparison unit, 205-Precise positioning unit for non-compliant areas, 3-Trajectory iteration correction module, 301-Trajectory parameter inverse calculation unit, 3011-Single-factor precise correction subunit, 3012-Multi-factor coupling correction subunit, 3013-Iteration over-limit termination subunit, 302-Multi-constraint supplementary flight trajectory generation unit, 303-Closed-loop execution scheduling unit, 4-Flight closed-loop execution module, 401-Emergency power scheduling unit. Detailed Implementation

[0025] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application.

[0026] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used in this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0027] It should be understood that although the terms first, second, third, etc., may be used in this application to describe various information, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to determination."

[0028] Please see Figures 1-7 This invention provides an autonomous mission flight control system for an unmanned aerial vehicle (UAV), comprising a mission planning module 1, a mission effectiveness calculation module 2, a trajectory iterative correction module 3, and a flight closed-loop execution module 4; wherein: The task planning module 1 is used to preset the global autonomous flight route with task target spatiotemporal anchor points, divide the independent task segments with accounting windows, establish a dynamic quantitative evaluation system for the effectiveness of task execution adapted to the scenario, and set multi-dimensional linkage control thresholds. The task effectiveness calculation module 2 is used to collect spatiotemporally aligned task payload operation data and flight attitude data. Relying on the dynamic quantitative evaluation system, the airborne terminal calculates the task execution effectiveness of the current task segment in real time and outputs the linkage comparison result with the control threshold. The trajectory iterative correction module 3 is used to establish an inverse solution mapping model between the effectiveness difference and trajectory parameters, with the mission execution effectiveness comparison result as the core control input, to generate an iterative correction trajectory that satisfies multiple constraints, and to trigger closed-loop flight action commands. The flight closed-loop execution module 4 is used to control the UAV to execute a preset global route or iteratively correct the trajectory. After completing the closed-loop verification of the effectiveness of the entire mission segment, it generates and transmits the mission execution key feature data with spatiotemporal tags and the trajectory correction link record.

[0029] In this embodiment, the autonomous mission flight control system uses mission execution effectiveness as the core control input, replacing the traditional control logic centered on trajectory tracking accuracy. The mission planning and spatiotemporal configuration module pre-divides the global flight path into multiple mission segments with independent calculation windows, each segment bound to a unique target spatiotemporal anchor point. The real-time mission effectiveness calculation module calculates the comprehensive effectiveness score of the current mission segment frame by frame during flight. The trajectory-effectiveness coupled iterative correction module triggers trajectory correction only when the effectiveness is below standard, and the corrected re-flight trajectory still needs to undergo effectiveness re-verification. The flight closed-loop execution and data feedback module generates a full record only after the effectiveness of all mission segments meets the standard. Through this architecture, a closed-loop control of "measurement → comparison → correction → re-measurement" is formed, ensuring that the mission execution quality of each mission segment meets the preset standards.

[0030] Furthermore, the task planning module 1 includes an anchor point route preset unit 101, a task segment spatiotemporal division unit 102, a dynamic validity system construction unit 103, and a multi-dimensional threshold linkage unit 104; wherein: The anchor point route preset unit 101 is used to plan a global flight route with target spatiotemporal anchor points according to the geographical coordinates, physical size and observation requirements of the mission target, and to solidify the waypoint three-dimensional coordinates, reference flight altitude, cruise speed, payload field of view and exposure trigger logic. The task segment spatiotemporal division unit 102 is used to divide the global flight path into multiple continuous task segments with independent validity calculation windows based on the spatial clustering results and time series of the task objectives. Each task segment is bound to a unique set of target spatiotemporal anchor points. The dynamic effectiveness system construction unit 103 is used to establish a dynamic quantitative evaluation system for task execution effectiveness that is linked to environmental complexity and target importance for two core operation scenarios: inspection and plant protection, and to solidify the core evaluation indicators and dynamic weight calculation logic. The multi-dimensional threshold linkage unit 104 is used to bind the dynamic quantitative evaluation system and link the effective achievement threshold, the maximum number of iterations for single task segment trajectory correction, the minimum power threshold for safe return, and the environmental complexity trigger threshold.

[0031] In this embodiment, the anchor point-based flight path preset unit 101 generates a global flight path based on the actual geographical coordinates of the mission target (such as the GPS coordinates of power transmission towers and the boundary points of plant protection plots) and observation requirements (such as shooting distance and field of view coverage). This global flight path includes the three-dimensional coordinates of waypoints, a reference flight altitude (e.g., 30m-80m), a cruising speed (e.g., 5m / s-15m / s), a payload field of view (e.g., 60°-90°), and exposure triggering logic. The mission segment spatiotemporal division unit 102 divides the global flight path into continuous mission segments of 100m-500m in length based on the spatial distribution density of the mission targets (e.g., targets with an interval <50m are grouped into the same cluster) and the inspection / spraying time sequence. Each mission segment corresponds to an effectiveness calculation sliding window.

[0032] Furthermore, the task segment spatiotemporal partitioning unit 102 includes a target spatiotemporal clustering subunit 1021, a calculation window adaptation subunit 1022, and an anchor point binding subunit 1023; wherein: The target spatiotemporal clustering subunit 1021 is used to use an adaptive clustering method based on spatial density to cluster all task targets covered by the global flight path according to their geographic coordinates, and at the same time match the time series of task execution to output multiple sets of target sets with spatiotemporal boundaries. The calculation window adaptation subunit 1022 is used to adapt the effective calculation sliding window of the task segment length and corresponding duration of 100m-500m according to the spatial span, number of targets and observation requirements of each target set, and output the adapted segmentation rules. The anchor point binding subunit 1023 is used to bind the spatiotemporal anchor points of each target set to a single independent task segment, and output segmented routes with target spatiotemporal anchor points and calculation windows.

[0033] In this embodiment, the calculation window adaptation subunit 1022 adapts the task segment length to 100m-500m according to the clustering space span (maximum target distance) and observation requirements, and sets a sliding window with a corresponding duration (e.g., 10s-50s); the anchor point binding subunit 1023 uses the geometric center of each target set as a spatiotemporal anchor point, binds it to each independent task segment, and outputs segmented route data containing anchor point coordinates, calculation window length, and target list.

[0034] Furthermore, the dynamic validity system construction unit 103 includes a scenario index solidification subunit 1031, a dynamic weight calculation subunit 1032, and a calculation logic solidification subunit 1033; wherein: The scenario index solidification subunit 1031 is used to solidify the target shooting coverage, target feature clarity, unobstructed image ratio, and target feature integrity as core evaluation indicators for inspection-type operation scenarios; and to solidify the target area coverage integrity, droplet adhesion uniformity, missed spray rate, and re-spray rate as core evaluation indicators for plant protection operation scenarios. The dynamic weight calculation subunit 1032 is used to pre-set weight benchmarks based on target importance, environmental complexity, and flight conditions using the analytic hierarchy process, and to adjust the dynamic weight ratio of each core evaluation indicator in real time in conjunction with environmental complexity. The weight adjustment range is 10%-60%. The calculation logic solidification subunit 1033 is used to calculate the frame-level effectiveness within a single calculation window through a weighted summation algorithm, and then output the cumulative task execution effectiveness comprehensive score of the current task segment through a moving average algorithm.

[0035] In this implementation, the judgment matrix is ​​constructed offline using the Analytic Hierarchy Process (AHP) to calculate the baseline weights of each evaluation index. Specifically: for high-voltage tower inspection, the baseline weight for target feature clarity is set to 50%, target coverage to 30%, and unobstructed image coverage to 20%. During flight, the baseline weights are fine-tuned online based on the environmental complexity factor (derived from real-time airborne visual detection of light intensity, haze concentration, obstacle density, etc.): when the light intensity is below the threshold, the weight of "clarity" is reduced by 5-10 percentage points, and the weight of "coverage" or "unobstructed image" is increased accordingly. The final dynamic weights are controlled within the range of 10%-60%. The calculation logic solidification subunit 1033 uses weighted summation to calculate the frame-level effectiveness: E_frame=Σ(w_i*s_i), where s_i is the normalized score (0-1) of each index. Then, the cumulative effectiveness comprehensive score is calculated through a sliding window (window size = total number of frames in the task segment): E_segment=(1 / N)ΣE_frame.

[0036] Furthermore, the task effectiveness calculation module 2 includes a spatiotemporal aligned data acquisition unit 201, an airborne lightweight feature extraction unit 202, a real-time effectiveness calculation unit 203, and a threshold linkage comparison unit 204; wherein: The spatiotemporal alignment data acquisition unit 201 is used to synchronously acquire the operational data of the airborne mission payload and the attitude data of the flight control system based on the spatiotemporal anchor point of the mission segment, and uniformly mark all data with millisecond-level timestamps and centimeter-level position stamps; wherein the operational data includes visible light, multispectral data, plant protection nozzle flow and pressure data, and the attitude data includes three-dimensional coordinates, flight altitude, attitude angle, and velocity vector. The airborne lightweight feature extraction unit 202 is used to extract target features and perform semantic segmentation on the collected operation data through a lightweight semantic segmentation model based on deep learning, identify target regions, non-target regions, and occluded regions, and output a standardized feature dataset. The real-time validity calculation unit 203 is used to perform frame-level validity calculation on the standardized feature dataset based on the dynamic quantitative evaluation system, and then calculate the cumulative validity comprehensive score of the current task segment in real time through the sliding window algorithm. The threshold linkage comparison unit 204 is used to compare the real-time calculated comprehensive score of effectiveness with the threshold set in the linkage frame by frame, and output two-level judgment results of compliance and non-compliance, and synchronously output the effectiveness difference data.

[0037] In this embodiment, the spatiotemporal aligned data acquisition unit 201 uses the timestamp of the task segment anchor point as a reference to synchronously trigger the acquisition of payload data (such as visible light images, multispectral data, nozzle flow rate) and flight control attitude data (GPS coordinates, IMU attitude, airspeed), and adds millisecond-level timestamps (accuracy ±10ms) and centimeter-level position stamps (RTK positioning) to each frame of data. The airborne lightweight feature extraction unit 202 adopts a lightweight semantic segmentation model based on deep learning (such as YOLOv8-seg, accelerated by TensorRT and deployed on NVIDIA Jetson OrinNX, with inference time ≤25ms / frame) to identify target areas (such as insulator strings, crop rows), non-target areas, and occluded areas (such as leaves, clouds and fog) pixel by pixel, and outputs a standardized feature dataset (containing the classification label and confidence score of each pixel); the real-time validity calculation unit 203 calls the dynamic evaluation system of claim 4 to calculate the validity score frame by frame, and updates the cumulative validity of the current task segment through a sliding window. The threshold linkage comparison unit 204 compares the cumulative validity with the preset compliance threshold (e.g., 0.85). If it is ≥0.85, it outputs "compliant"; otherwise, it outputs "not compliant" and the validity difference ΔE = compliance threshold - current score.

[0038] Furthermore, the task effectiveness calculation module 2 also includes a precise positioning unit 205 for areas that do not meet the standards; The sub-unit for precise location of non-compliant areas is used to locate the three-dimensional geographic boundaries, area, non-compliance type and cause of non-compliant areas for frame-level data within the non-compliant calculation window through pixel-level semantic segmentation and inverse geographic coordinate calculation, and output a feature dataset of non-compliant areas with spatiotemporal labels.

[0039] In this implementation, when a certain calculation window is determined to be "unqualified," the unqualified area precise positioning subunit performs pixel-level semantic segmentation on all frame-level data within that window, extracting all pixel clusters marked as "target areas" but with "insufficient feature completeness" or "occlusion." Using a camera intrinsic parameter model (pinhole camera model) and UAV pose data (altitude, pitch angle, yaw angle), the pixel coordinates are inversely calculated to the WGS-84 geographic coordinate system, yielding the three-dimensional geographic boundary (latitude, longitude, and altitude range), area (㎡), unqualified type (e.g., "insufficient clarity," "missing coverage," "occlusion"), and cause (e.g., "backlighting," "target offset," "obstacle occlusion") of the unqualified area. A feature dataset with spatiotemporal labels is output for use by the trajectory correction module.

[0040] Furthermore, the trajectory iteration correction module 3 includes a trajectory parameter inverse calculation unit 301, a multi-constraint supplementary flight trajectory generation unit 302, and a closed-loop execution scheduling unit 303; wherein: The inverse calculation unit 301 for track parameters is used to establish an inverse calculation mapping model between the effective difference and track parameters, with the effective difference and the non-compliant area feature dataset as the core input, and to output the optimal track correction parameters by matching the non-compliant type. The correction parameters include flight altitude, flight speed, lateral offset of the route, load field of view, and route overlap rate. The multi-constraint flight path generation unit 302 is used to generate an iterative flight path that satisfies the global optimum based on the optimal path correction parameters, combined with UAV flight dynamics constraints, remaining battery power constraints, field of view coverage constraints, and safety obstacle avoidance constraints, through a heuristic path search algorithm. The closed-loop execution scheduling unit 303 is used to receive the validity comparison and judgment results and generate closed-loop flight action commands: for the task segment with the validity meets the standard, it generates a continuous flight command that extends to the next task segment; for the task segment with the validity does not meet the standard, it generates a control command to pause the global route and execute iterative supplementary flight trajectory.

[0041] In this embodiment, regarding the "establishment of an inverse calculation mapping model between effectiveness difference and trajectory parameters" in the trajectory parameter inverse calculation unit 301: a rule mapping table is pre-constructed using actual flight calibration and simulation data. The non-compliance types (insufficient clarity, insufficient coverage, image occlusion, composite non-compliance) and effectiveness difference ΔE (divided into intervals such as ≤0.10, >0.10, ≤0.15, >0.15) are directly mapped to flight altitude adjustment (reduced by 10%-20%), flight speed adjustment (reduced by 40%-50%), lateral offset of the flight path (5-10m), load field of view adjustment (narrowed or widened by 10°-15°), and flight path overlap adjustment (increased by 15%-30%). For composite non-compliance, the above parameters are weighted and fused according to the proportion of each sub-item ΔE. This mapping table is stored in the onboard computer, and the lookup time is <1ms. Meanwhile, the heuristic path search algorithm specifically adopts HybridA* and quantifies flight dynamics constraints (minimum turning radius ≥ 15m, maximum roll angle ≤ 30°), remaining power constraints (replacement flight energy consumption ≤ 60% of remaining power), field of view coverage constraints (overlap rate ≥ 20%), and safety obstacle avoidance constraints (distance from obstacles ≥ 5m), outputting a feasible replacement flight track within 200ms. Upon receiving a "failed to meet" signal, the closed-loop execution scheduling unit 303 sequentially executes the complete time sequence of interrupting the current track, saving the context, calling 301 / 302 to generate a replacement flight track, switching execution, and re-triggering the validity calculation after replacement. Through the above specific implementation, those skilled in the art can implement this inverse calculation mapping model and replacement flight track generation logic without creative effort. The inverse calculation unit 301 for track parameters employs an inverse calculation method based on a rule mapping table to establish a mapping rule of "non-compliance type + effectiveness difference → track correction parameter". The mapping table pre-sets the following typical rules: Insufficient target clarity ΔE≤0.10 Flight altitude reduced by 10%, flight speed reduced by 40%. Insufficient target clarity ΔE>0.10 Flight altitude reduced by 20%, flight speed reduced by 50%, payload field of view narrowed by 10° Insufficient target coverage ΔE≤0.15 Lateral overlap of flight paths increased by 15%, and flight speed decreased by 20%. Insufficient target coverage ΔE>0.15 The lateral overlap of the flight path is increased by 30%, the load field of view is widened by 15°, and the flight path offset is adjusted. Image occlusion any Lateral deviation of 5m-10m (towards the unobstructed side), flight altitude increased by 15%. For composite data that fails to meet the standards (simultaneous insufficient sharpness and insufficient coverage), a weighted fusion method is used: Correction parameter = α·Parameter_Sharpness + β·Parameter_Coverage, where α and β are dynamically allocated based on the proportion of the effective difference (α = ΔE_Sharpness / (ΔE_Sharpness + ΔE_Coverage)). This mapping table is stored in the onboard computer, and the lookup time is <1ms.

[0042] Based on the above-mentioned modified parameters, the multi-constraint flight path generation unit 302 combines UAV flight dynamics constraints (maximum roll angle 30°, maximum climb rate 5m / s), remaining power constraints (total energy consumption of the flight path ≤ 60% of the remaining power), field of view coverage constraints (overlap rate of adjacent paths ≥ 20%) and safety obstacle avoidance constraints (distance from the nearest obstacle ≥ 5m). It uses a heuristic path search algorithm (such as HybridA*) to generate iterative flight paths in the local planning space, ensuring that replanning is completed within 5 seconds.

[0043] After receiving the "not meeting the standard" judgment, the closed-loop execution scheduling unit 303 immediately interrupts the global route execution, saves the current task segment context, loads the supplementary flight track and executes it; after the supplementary flight is completed, it re-enters the validity calculation process. If it meets the standard, it restores the global route and enters the next task segment. If it still does not meet the standard, it triggers the supplementary flight again (not exceeding the preset maximum number of iterations).

[0044] Furthermore, the trajectory parameter inverse calculation unit 301 includes a single-factor precise correction subunit 3011, a multi-factor coupled correction subunit 3012, and an iteration over-limit termination subunit 3013; wherein: The single-factor precision correction subunit 3011 is used to make precise corrections to the corresponding flight path parameters for a single type of non-compliance problem: when the target clarity is not up to standard, the reference flight altitude is lowered by 10%-30% and the flight speed is reduced by 40%; when the target coverage is not up to standard, the lateral overlap rate of the flight path is increased by 15%-30% and the payload field of view is adjusted; when the image occlusion is not up to standard, the lateral offset of the flight path is adjusted and the flight altitude is optimized. The multi-factor coupled correction subunit 3012 is used to solve multi-parameter coupled correction schemes for multiple types of composite non-compliance problems through multi-objective optimization algorithms, and generate optimal trajectory parameters that take into account both correction effect and energy consumption. The iteration over-limit termination subunit 3013 is used to terminate the current mission segment's supplementary flight process when the number of supplementary flight iterations reaches the preset maximum correction number and the mission segment's effectiveness still fails to meet the standard. It pushes the full feature data of the non-compliant area to the ground control station, triggers an anomaly alarm, and generates an instruction to restore the execution of the global flight path.

[0045] In this embodiment, the single-factor precise correction subunit 3011 strictly follows the mapping table to correct for a single type of non-compliance. For example, if it is determined that "target clarity is non-compliant" and ΔE=0.08, then the following actions are taken: the flight altitude is reduced from 50m to 45m (reduction of 10%), and the flight speed is reduced from 10m / s to 6m / s (reduction of 40%). The multi-factor coupled correction subunit 3012, for composite non-compliance, uses the Pareto multi-objective optimization algorithm to search for the parameter combination that maximizes the effectiveness of each indicator in the parameter space (altitude, speed, overlap rate, offset), and adds an energy consumption penalty term (energy consumption increment ≤ 15%). The iteration over-limit termination subunit 3013 monitors the number of supplementary flights. When the preset maximum number of flights (e.g., 3 times) is reached and the validity is still <0.85, the supplementary flight is terminated. The feature dataset of the non-compliant area (including geographical location, image snapshot, and cause analysis) is pushed to the ground control station through the 4G / 5G link, triggering an audible and visual alarm. At the same time, an instruction to "skip the current task segment and continue to execute the subsequent task segment" is generated to avoid the interruption of the entire task due to local problems.

[0046] Furthermore, the flight closed-loop execution module 4 also includes a power emergency dispatch unit 401; The power emergency dispatch unit 401 is used to compare the remaining airborne power with the preset minimum power threshold for safe return in real time, and establish a multi-objective optimization model for the completion of tasks with remaining power. When the remaining power is lower than the threshold, the execution order of task segments is dynamically adjusted according to the target weight priority in the effectiveness evaluation system. The task segments corresponding to high-weight targets are completed first to maximize the overall task execution effectiveness. After completing the high-priority tasks, an automatic return command is generated, and the effectiveness calculation record and trajectory correction link data are retained.

[0047] In this embodiment, the remaining battery charge (SOC) is read in real time and compared with a preset safe return threshold (e.g., 20%). When the SOC is lower than the threshold, emergency dispatch is initiated: First, the list of currently incomplete task segments and their target weights are obtained (derived from the importance weights preset by AHP in claim 4, such as "0.6 for core equipment in substations and 0.3 for general lines"); second, a multi-objective optimization model is established: maxΣ(w_j*x_j*E_segment_j), with the constraint that Σ(energy_j*x_j)≤SOC_remaining, where x_j∈{0,1} indicates whether task segment j is executed, and energy_j is the estimated energy consumption; the optimal task subset is solved using a greedy algorithm or dynamic programming; finally, the execution order of task segments is dynamically adjusted according to the solution results, prioritizing the execution of task segments corresponding to high-weight targets. After each segment is completed, the remaining battery charge is recalculated. If the battery charge is still sufficient, the next highest weight segment is executed until the battery charge is insufficient to complete the next minimum energy consumption task segment, at which point a return command is automatically generated. Before returning to base, the validity calculation records of the completed mission segments, the supplementary flight track link data, and the location information of the unfinished mission segments are packaged and stored in the airborne black box, and exported through the ground station after returning to base for subsequent analysis or secondary operations.

[0048] The present invention also provides an unmanned aerial vehicle (UAV), including the aforementioned autonomous mission flight control system for the UAV, wherein: The system includes an airframe, a power system, a flight control computer, a mission payload (such as a high-definition camera, a multispectral sensor, and an agricultural spray nozzle), and an autonomous mission flight control system as described in any one of claims 1 to 9. The flight control computer is connected to the mission payload via a high-speed bus (such as GMSL or CANFD) and runs the software of each module of the control system. This UAV can autonomously complete iterative trajectory correction and mission quality closed-loop assurance in tasks such as power line inspection and agricultural plant protection, using mission execution effectiveness as the sole core control input.

[0049] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the disclosure herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein.

[0050] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope.

Claims

1. An autonomous mission flight control system for an unmanned aerial vehicle (UAV), characterized in that, It includes a mission planning module, a mission effectiveness calculation module, a trajectory iterative correction module, and a flight closed-loop execution module; among which: The task planning module is used to preset the global autonomous flight route with task target spatiotemporal anchor points, divide the independent task segments with accounting windows, establish a dynamic quantitative evaluation system for the effectiveness of task execution adapted to the scenario, and set multi-dimensional linkage control thresholds. The task effectiveness calculation module is used to collect spatiotemporally aligned task payload operation data and flight attitude data. Relying on the dynamic quantitative evaluation system, the airborne terminal calculates the task execution effectiveness of the current task segment in real time and outputs the linkage comparison results with the control threshold. The trajectory iterative correction module is used to establish an inverse solution mapping model between the effectiveness difference and trajectory parameters, with the mission execution effectiveness comparison result as the core control input, to generate an iteratively corrected trajectory that satisfies multiple constraints, and to trigger closed-loop flight action commands. The flight closed-loop execution module is used to control the UAV to execute a preset global route or iteratively correct the trajectory. After completing the closed-loop verification of the effectiveness of the entire mission segment, it generates and transmits back the mission execution key feature data with spatiotemporal tags and the trajectory correction link record.

2. The autonomous mission flight control system for an unmanned aerial vehicle according to claim 1, characterized in that, The task planning module includes an anchor point route preset unit, a task segment spatiotemporal division unit, a dynamic effectiveness system construction unit, and a multi-dimensional threshold linkage unit; wherein: The anchor point preset unit is used to plan a global flight path with target spatiotemporal anchor points according to the geographical coordinates, physical size and observation requirements of the mission target, and to fix the waypoint three-dimensional coordinates, reference flight altitude, cruise speed, payload field of view and exposure trigger logic. The task segment spatiotemporal partitioning unit is used to divide the global flight path into multiple continuous task segments with independent validity calculation windows based on the spatial clustering results and time series of the task objectives. Each task segment is bound to a unique set of target spatiotemporal anchor points. The dynamic effectiveness system construction unit is used to establish a dynamic quantitative evaluation system for task execution effectiveness that is linked to environmental complexity and target importance for two core operation scenarios: inspection and plant protection, and to solidify the core evaluation indicators and dynamic weight calculation logic. The multi-dimensional threshold linkage unit is used to bind the dynamic quantitative evaluation system and link the effective achievement threshold, the maximum number of iterations for single-task segment trajectory correction, the minimum power threshold for safe return, and the environmental complexity trigger threshold.

3. The autonomous mission flight control system for an unmanned aerial vehicle according to claim 2, characterized in that, The task segment spatiotemporal partitioning unit includes a target spatiotemporal clustering subunit, a calculation window adaptation subunit, and an anchor point binding subunit; wherein: The target spatiotemporal clustering subunit is used to use an adaptive clustering method based on spatial density to cluster all task targets covered by the global flight path according to their geographic coordinates, and at the same time match the time series of task execution to output multiple sets of target sets with spatiotemporal boundaries. The calculation window adaptation subunit is used to adapt the effective calculation sliding window of the task segment length and corresponding duration of 100m-500m according to the spatial span, number of targets and observation requirements of each target set, and output the adapted segmentation rules. The anchor point binding subunit is used to bind the spatiotemporal anchor points of each target set to a single independent task segment, and output segmented routes with target spatiotemporal anchor points and calculation windows.

4. The autonomous mission flight control system for an unmanned aerial vehicle according to claim 2, characterized in that, The dynamic validity system construction unit includes a scenario indicator solidification subunit, a dynamic weight calculation subunit, and a calculation logic solidification subunit; wherein: The scenario index solidification subunit is used to solidify the target shooting coverage, target feature clarity, unobstructed image ratio, and target feature integrity as core evaluation indicators for inspection-type operation scenarios; and to solidify the target area coverage integrity, droplet adhesion uniformity, missed spray rate, and re-spray rate as core evaluation indicators for plant protection operation scenarios. The dynamic weight calculation subunit is used to pre-set weight benchmarks based on target importance, environmental complexity, and flight conditions using the analytic hierarchy process, and to adjust the dynamic weight ratio of each core evaluation indicator in real time in conjunction with environmental complexity. The weight adjustment range is 10%-60%. The calculation logic solidification subunit is used to calculate the frame-level effectiveness within a single calculation window through a weighted summation algorithm, and then output the cumulative task execution effectiveness comprehensive score of the current task segment through a moving average algorithm.

5. The autonomous mission flight control system for an unmanned aerial vehicle according to claim 1, characterized in that, The task effectiveness calculation module includes a spatiotemporal aligned data acquisition unit, an airborne lightweight feature extraction unit, a real-time effectiveness calculation unit, and a threshold-linked comparison unit; wherein: The spatiotemporal alignment data acquisition unit is used to synchronously acquire the operational data of the airborne mission payload and the attitude data of the flight control system based on the spatiotemporal anchor point of the mission segment, and uniformly mark all data with millisecond-level timestamps and centimeter-level position stamps; among which, the operational data includes visible light, multispectral data, and plant protection nozzle flow and pressure data, and the attitude data includes three-dimensional coordinates, flight altitude, attitude angle, and velocity vector. The airborne lightweight feature extraction unit is used to extract target features and perform semantic segmentation on the collected operation data through a lightweight semantic segmentation model based on deep learning, identify target regions, non-target regions, and occluded regions, and output a standardized feature dataset. The real-time validity calculation unit is used to perform frame-level validity calculation on the standardized feature dataset based on the dynamic quantitative evaluation system, and then calculate the cumulative validity comprehensive score of the current task segment in real time through the sliding window algorithm. The threshold linkage comparison unit is used to compare the real-time calculated comprehensive score of effectiveness with the threshold set in the linkage frame by frame, and output two-level judgment results of compliance and non-compliance, and synchronously output the effectiveness difference data.

6. The autonomous mission flight control system for an unmanned aerial vehicle according to claim 5, characterized in that, The task effectiveness calculation module also includes a precise positioning unit for areas that do not meet the standards. The sub-unit for precise location of non-compliant areas is used to locate the three-dimensional geographic boundaries, area, non-compliance type and cause of non-compliant areas for frame-level data within the non-compliant calculation window through pixel-level semantic segmentation and inverse geographic coordinate calculation, and output a feature dataset of non-compliant areas with spatiotemporal labels.

7. The autonomous mission flight control system for an unmanned aerial vehicle according to claim 1, characterized in that, The trajectory iteration correction module includes a trajectory parameter inverse calculation unit, a multi-constraint supplementary flight trajectory generation unit, and a closed-loop execution scheduling unit; wherein: The inverse calculation unit for track parameters is used to establish an inverse calculation mapping model between the effective difference and track parameters, with the effective difference and the non-compliant area feature dataset as the core inputs. It matches the non-compliant type and outputs the optimal track correction parameters, which include flight altitude, flight speed, lateral offset of the flight path, load field of view, and flight path overlap rate. The multi-constraint flight path generation unit is used to generate an iterative flight path that satisfies global optimum based on the optimal path correction parameters, combined with UAV flight dynamics constraints, remaining battery power constraints, field of view coverage constraints, and safety obstacle avoidance constraints, through a heuristic path search algorithm. The closed-loop execution scheduling unit is used to receive the effectiveness comparison and judgment results and generate closed-loop flight action commands: for task segments that meet the effectiveness criteria, it generates continuous flight commands that are extended to the next task segment; for task segments that do not meet the effectiveness criteria, it generates control commands to pause the global flight path and execute iterative supplementary flight paths.

8. The autonomous mission flight control system for an unmanned aerial vehicle according to claim 7, characterized in that, The trajectory parameter inverse calculation unit includes a single-factor precise correction subunit, a multi-factor coupled correction subunit, and an iteration over-limit termination subunit; wherein: The single-factor precision correction subunit is used to precisely correct the corresponding flight path parameters for a single type of non-compliance problem: when the target clarity is not up to standard, the reference flight altitude is lowered by 10%-30% and the flight speed is reduced by 40%; when the target coverage is not up to standard, the lateral overlap rate of the flight path is increased by 15%-30% and the payload field of view is adjusted; when the image occlusion is not up to standard, the lateral offset of the flight path is adjusted and the flight altitude is optimized. The multi-factor coupled correction subunit is used to solve multi-parameter coupled correction schemes for multiple types of composite non-compliance problems through multi-objective optimization algorithms, and generate optimal trajectory parameters that take into account both correction effect and energy consumption. The iteration over-limit termination subunit is used to terminate the current mission segment's supplementary flight process when the mission segment's effectiveness still fails to meet the target after the number of supplementary flight iterations reaches the preset maximum correction number. It pushes the full feature data of the non-targeted area to the ground control station, triggers an anomaly alarm, and generates an instruction to restore the execution of the global flight path.

9. The autonomous mission flight control system for an unmanned aerial vehicle according to claim 1, characterized in that, The flight closed-loop execution module also includes a power emergency dispatch unit; The power emergency dispatch unit is used to compare the remaining airborne power with the preset minimum power threshold for safe return in real time, and establish a multi-objective optimization model for the completion of tasks with remaining power. When the remaining power is lower than the threshold, the execution order of task segments is dynamically adjusted according to the target weight priority in the effectiveness evaluation system. The task segments corresponding to high-weight targets are completed first to maximize the global task execution effectiveness. After completing the high-priority tasks, an automatic return command is generated, and the effectiveness calculation record and trajectory correction link data are retained.

10. A drone, characterized in that, Including the autonomous mission flight control system of the UAV as described in any one of claims 1-9.

Images