A priority-based task scheduling and discrete action event synchronous control method and system for a mother machine flight control

Abstract

This application relates to the field of UAV flight control technology, specifically a priority-based task scheduling and discrete action event synchronization control method and system for mother aircraft flight control. It ensures the timeliness of sensor data through a timeliness guarantee mechanism and switches the flight control target to the highest priority attitude control when an instability risk is detected, while suspending discrete tasks to prioritize airframe stability and prevent crashes caused by forced execution of discrete actions. After attitude recovery, it accurately locates the breakpoints of discrete actions based on real-time feedback from physical sensors and resumes execution, avoiding backtracking or retrying strategies due to irreversible physical actions, thus ensuring the strict temporality and integrity of the separation action sequence. Furthermore, critical point feedforward compensation effectively suppresses instantaneous disturbances caused by sudden quality changes, further improving the smoothness and success rate of the separation process. This method significantly improves the reliability and robustness of the mother-daughter aircraft separation task while ensuring flight safety.

Description

Technical Field

[0001] This application relates to the field of UAV flight control technology, specifically a method and system for priority task scheduling and discrete action event synchronization control for mother aircraft flight control. Background Technology

[0002] With the rapid development of UAV swarm and aerial delivery technologies, unmanned mother-daughter systems are playing an increasingly important role in collaborative reconnaissance, precision delivery, and distributed warfare. The flight control system of these mother-daughter systems must simultaneously undertake two distinct types of control tasks: one is continuous closed-loop control with strong real-time requirements and high frequency (such as attitude / track holding based on IMU / GNSS, typically requiring millisecond-level response); the other is discrete action control with complex logic and strict time-series dependencies (such as unlocking electro-permanent magnet locks, extending electromagnetic pins, detonating explosive bolts, and switching mission phases after separation when the daughter aircraft is mounted).

[0003] For discrete actions, the execution order of their task sequences must be guaranteed. However, current mother-daughter UAVs are highly susceptible to instability when encountering crosswinds or sudden changes in mass during separation. Forcing the execution of discrete actions such as unlocking and locking as planned in such situations would lead to attitude divergence and directly cause a crash. Therefore, attitude control must be prioritized (introducing priority scheduling). However, high-priority attitude tasks preempting the CPU frequently interrupts the discrete action sequence, and physical actions are irreversible, making simple rollback or restarting impossible. Therefore, a collaborative control method is urgently needed: while allowing attitude preemption, real-time status perception through physical sensors and a breakpoint resumption mechanism ensures accurate continuation of execution after the discrete sequence is interrupted, thus balancing flight safety and the reliability of the separation mission. Summary of the Invention

[0004] In view of this, the purpose of this application is to provide a priority-based task scheduling and discrete action event synchronization control method and system for mother machine flight control, so as to solve the problems in the background art.

[0005] To achieve the above objectives, this application adopts the following technical solution:

[0006] This application discloses a priority-based task scheduling and discrete action event synchronization control method for mother machine flight control, comprising the following steps:

[0007] During the flight of the mother and daughter drones to the separation point, sensor data of the mother and daughter drones is acquired, wherein the sensor data is updated based on a pre-configured timeliness guarantee mechanism;

[0008] When the mother and daughter UAVs fly to the preset separation point, the pre-created discrete action task is awakened and placed in the ready state. The discrete action task generates discrete action instructions in sequence according to the step sequence index and sends them out for execution via the message queue.

[0009] During the separation of the mother and daughter UAVs, critical point feedforward compensation to suppress mass mutations is performed, and a comprehensive risk index to characterize the instability risk of the mother UAV is calculated based on the sensor data. When the comprehensive risk index characterizes the instability risk of the mother UAV, the control target of the flight control module is forcibly switched from mission tracking to the highest priority flight attitude control. During the switching period, discrete action tasks are suspended. The feedforward compensation includes a mass feedforward term that takes effect immediately when the electromagnetic plug is energized, and an asynchronous thrust correction term that is smoothly switched in by first-order inertial filtering after the daughter UAV is confirmed to have separated.

[0010] When the machine returns to a stable state, the discrete action task is awakened and the next discrete action instruction to be executed is located based on the feedback in the sensor data. Then, the process returns to calculating the comprehensive risk index to characterize the machine's instability risk based on the sensor data until separation is completed.

[0011] In one embodiment of this application, acquiring sensor data from a mother-daughter unmanned aerial vehicle includes:

[0012] Obtain the current raw sensor data of the mother-daughter UAV and the acquisition timestamp of the raw sensor data, wherein the raw sensor data is updated periodically;

[0013] The data delay is calculated based on the acquisition timestamp of the original sensor data and the current moment.

[0014] When the data delay does not exceed a preset first delay threshold, the original sensor data is used as sensor data;

[0015] When the data delay exceeds a preset first delay threshold, inertial compensation is performed on the sensor data and the PID integral term of the mother-daughter UAV control module is frozen to obtain the compensated sensor data, and the compensated sensor data is used as the sensor data.

[0016] When the data delay exceeds a preset second delay threshold, the sensor link is determined to be faulty, and the data acquisition module is forcibly restarted. After restarting, the module returns to acquiring the current raw sensor data of the mother and daughter UAVs and the acquisition timestamp of the raw sensor data. The second delay threshold is greater than the first delay threshold.

[0017] In one embodiment of this application, it further includes:

[0018] During the flight of the mother-daughter UAV to the separation point, the ready task queue in the mother-daughter mission machine is acquired, wherein the tasks in the ready task queue come from the ground command center or preset automatic triggering tasks.

[0019] Based on a pre-configured task priority scheduling mechanism, the highest priority task is selected from the ready task queue and executed in conjunction with the sensor data.

[0020] In one embodiment of this application, the discrete action instructions in the discrete action task include an unlocking electro-permanent magnet lock instruction, an electromagnetic pin energizing instruction, a microswitch detection instruction and a slave unit thrust data detection instruction, and a mother unit balancing instruction. The execution of critical point feedforward compensation to suppress mass mutations includes:

[0021] When the electromagnetic pin energizing command is executed, the mass feedforward term is calculated. And based on the quality feedforward term Execute the first feedforward compensation;

[0022] When the microswitch detection and submachine thrust data detection commands are executed, the asynchronous thrust correction term is calculated. And based on the asynchronous thrust correction term A second feedforward compensation is performed, wherein the first feedforward compensation and the second feedforward compensation are smoothly switched within a preset window.

[0023] In one embodiment of this application, the mass feedforward term is calculated. And based on the quality feedforward term Performing the first feedforward compensation includes:

[0024] Calculate the quality feedforward term The mass feedforward term The mathematical expression is:

[0025]

[0026] In the formula, For the quality of the submachine, At the current speed, The longitudinal lever arm is the distance from the daughter unit's mounting point to the mother unit's center of gravity. The moment of inertia of the mother machine's pitch axis. This is the gain coefficient;

[0027] Extract the feedback control quantity at the current moment from the mother aircraft's flight control system. And based on the feedback control quantity and the aforementioned quality feedforward term Calculate the final control quantity , .

[0028] In one embodiment of this application, the asynchronous thrust correction term is calculated. And based on the asynchronous thrust correction term Performing the second feedforward compensation includes:

[0029] Calculate the asynchronous thrust correction term The asynchronous thrust correction term The mathematical expression is:

[0030]

[0031] In the formula, The thrust value transmitted back by the slave unit. The lever arm of the daughter unit's thrust relative to the mother unit's center of gravity. Angular acceleration-rudder surface conversion gain, The moment of inertia of the mother machine's pitch axis. The time constant of the first-order inertial filter is... The execution time of the command to energize the electromagnetic latch. The current moment;

[0032] Extract the feedback control quantity at the current moment from the mother aircraft's flight control system. And based on the feedback control quantity and the asynchronous thrust correction term Calculate the final control quantity , .

[0033] In one embodiment of this application, a comprehensive risk index for characterizing the instability risk of the mother machine is calculated based on the sensor data, including:

[0034] Obtain the measured three-axis attitude vectors and measured three-axis acceleration vectors of the machine at the current moment. Control surface command vector And obtain the three-axis attitude target vector of the mother machine at the current moment;

[0035] The attitude tracking error is obtained by subtracting the measured three-axis attitude vector from the target three-axis attitude vector. Based on the attitude tracking error The measured vector of the triaxial acceleration and the control surface command vector Calculate the comprehensive risk index used to characterize the risk of machine instability. Among them, the comprehensive risk index The mathematical expression is:

[0036]

[0037] In the formula, For Chebyshev norm, To the maximum attitude tracking error, The first derivative of the attitude tracking error. For the maximum angular divergence rate, This indicates taking the maximum value. This represents the maximum physical deflection angle of the control surface. This represents the maximum excess capacity of the control surface saturation. For standard triaxial acceleration vectors, For the maximum dynamic overload increment, All of these are weight parameters.

[0038] In one embodiment of this application, when the control objective of the flight control module is switched to flight attitude control, the flight control module performs attitude stabilization control by calling a pre-configured flight control stabilization algorithm, wherein the flight control stabilization algorithm includes an active disturbance rejection and robust control algorithm, an incremental and nonlinear control algorithm, and a gain scheduling control method.

[0039] In one embodiment of this application, when the mother and daughter UAVs fly to a preset separation point, the mother UAV maintains its cruising speed and controls the start of the daughter UAV. When the thrust of the daughter UAV is greater than a preset thrust threshold, it enters a separation execution state and performs discrete actions based on the wake-up discrete action task.

[0040] This application also provides a priority-based task scheduling and discrete action event synchronization control system for mother aircraft flight control, including:

[0041] The acquisition module is used to acquire sensor data of the mother and daughter drones during their flight to the separation point. The sensor data is updated based on a pre-configured timeliness guarantee mechanism.

[0042] The separation task wake-up module is used to wake up the pre-created discrete action task and put it into a ready state when the mother and daughter UAVs fly to the preset separation point. The discrete action task generates discrete action instructions in sequence according to the step sequence index and sends them out for execution through the message queue.

[0043] The separation and priority scheduling module is used to perform critical point feedforward compensation to suppress mass mutations during the separation process of the mother and daughter UAVs, and to calculate a comprehensive risk index to characterize the instability risk of the mother UAV based on the sensor data. When the comprehensive risk index indicates that the mother UAV has an instability risk, the control target of the flight control module is forcibly switched from task tracking to the highest priority flight attitude control. During the switching period, the discrete action task is suspended. The feedforward compensation includes a mass feedforward term that takes effect immediately when the electromagnetic plug is energized, and an asynchronous thrust correction term that is smoothly switched in by first-order inertial filtering after the daughter UAV is confirmed to have separated.

[0044] The loop execution module is used to wake up the discrete action task when the machine recovers to a stable state, locate the next discrete action instruction to be executed based on the feedback in the sensor data, and return to calculate the comprehensive risk index to characterize the machine's instability risk based on the sensor data, until the separation is completed.

[0045] The beneficial effects of this application are as follows: This application provides a priority-based task scheduling and discrete action event synchronization control method and system for mother aircraft flight control. On the one hand, it ensures the timeliness of sensor data through a timeliness guarantee mechanism, and switches the flight control target to the highest priority attitude control when an instability risk is detected, while suspending discrete tasks to prioritize airframe stability and avoid crashes caused by forced execution of discrete actions. On the other hand, after attitude recovery, it accurately locates the breakpoints of discrete actions based on real-time feedback from physical sensors and wakes them up to resume execution, avoiding the need for backtracking or retrying strategies due to irreversible physical actions, thereby ensuring the strict temporality and integrity of the separation action sequence. Furthermore, critical point feedforward compensation effectively suppresses instantaneous disturbances caused by sudden quality changes, further improving the smoothness and success rate of the separation process. This method significantly improves the execution reliability and robustness of the mother-daughter aircraft separation task while ensuring flight safety. Attached Figure Description

[0046] The present application will be further described below with reference to the accompanying drawings and embodiments:

[0047] Figure 1 This is a schematic diagram of the structure of a mother-daughter unmanned aerial vehicle (UAV) according to one embodiment of this application;

[0048] Figure 2 This is a hardware architecture diagram of a mother-daughter unmanned aerial vehicle in one embodiment of this application;

[0049] Figure 3 This is a flowchart illustrating a priority-based task scheduling and discrete action event synchronization control method for mother machine flight control in one embodiment of this application;

[0050] Figure 4 This is a schematic diagram of the execution flow of the separation preparation mode in one embodiment of this application;

[0051] Figure 5 This is a logical structure diagram of a priority-based task scheduling and discrete action event synchronization control system for mother machine flight control, as shown in one embodiment of this application. Detailed Implementation

[0052] The following specific examples illustrate the implementation of this application. Those skilled in the art can easily understand other advantages and effects of this application from the content disclosed in this specification. This application can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this application. It should be noted that, unless otherwise specified, the following embodiments and features in the embodiments can be combined with each other.

[0053] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of this application. Therefore, the drawings only show the layers related to this application and are not drawn according to the actual number, shape and size of the layers in the actual implementation. In the actual implementation, the form, number and proportion of each layer can be arbitrarily changed, and the layer layout may also be more complex.

[0054] Numerous details are explored in the following description to provide a more thorough explanation of embodiments of this application; however, it will be apparent to those skilled in the art that embodiments of this application may be practiced without these specific details.

[0055] Figure 1 This is a schematic diagram of the structure of a mother-daughter unmanned aerial vehicle (UAV) according to one embodiment of this application, as shown below. Figure 1 As shown, the mother aircraft adopts a blended wing-body delta wing structure with a smooth overall aerodynamic shape, possessing excellent lift characteristics and stealth capabilities. The mother aircraft's nose is equipped with a daughter aircraft mounting rail and a separation lock, allowing the daughter aircraft to be detachably connected to the mother aircraft via a two-stage separation mechanism.

[0056] The secondary separation mechanism is located between the mother unit's head mounting base and the daughter unit's tail interface, and consists of a guide rail, an opening slider, and a separation lock. The guide rail is fixed to the mother unit's head; the opening slider is fixed to the daughter unit's rear abdomen, and a photoelectric sensor is installed at the bottom of the opening slider to detect whether the daughter unit has completely detached. During mounting, the opening slider on the daughter unit's rear abdomen inserts into the guide rail on the mother unit's head, forming a sliding clearance fit. During separation, the daughter unit accelerates forward, and the opening slider is pulled out of the guide rail.

[0057] The sub-unit adopts a blended wing-body tailless delta wing structure, with a flat and smooth overall shape, small size, and good lift characteristics and stealth capabilities. The sub-unit is designed for single use, and the overall weight is kept within the reasonable range for lightweight unmanned aerial vehicles.

[0058] The daughter unit uses two turbojet engines as propulsion devices, with rated power sufficient for cruise and peak power to support acceleration maneuvers. During separation, the turbojet engines activate to generate thrust, which, in conjunction with the separation mechanism, unlocks the daughter unit and separates it from the mother unit.

[0059] Figure 2 This is a hardware architecture diagram of a mother-daughter unmanned aerial vehicle (UAV) in one embodiment of this application, such as... Figure 2 As shown, the mother aircraft flight control system is used for the mother aircraft's flight control, navigation calculation, mission management, and separation / recovery command generation.

[0060] The flight control system adopts a distributed I / O architecture, consisting of a central computing module and multiple remote I / O nodes, a positioning and navigation module, a data link module, and a power management module.

[0061] The central computing module is the core of the flight control system for computing and decision-making. It provides UART and CAN interfaces, connects to remote I / O nodes and positioning modules via the CAN bus, and connects to the data link module via UART. It includes the main processor and hardware watchdog, runs the operating system in real time, supports multi-task scheduling, and allows high-priority tasks to preempt low-priority tasks to ensure the real-time performance of flight control.

[0062] Remote I / O nodes are located in key areas of the motherboard. Low-cost microcontrollers collect local sensor data and convert instructions from the central computing module into specific parameters to drive actuators. Each node has a built-in independent hardware watchdog. The nose node establishes a physical connection with the flight controller via a magnetic connector, exchanging status and command data when mounted. It also collects data from sensors related to secondary separation of the flight controller (including limit switches and microswitches) and drives electromagnetic latches and electro-permanent magnet locks. Furthermore, it collects thrust telemetry data (current / speed) for asynchronous thrust feedforward compensation calculations after separation. The dorsal node (parachute node) collects parachute-related sensors and drives explosive bolts. The tail node (flight control node) collects data from control surfaces and engine-related sensors and drives servos and throttles. The takeoff node (in coordination with the tail node) collects takeoff-related sensor data and participates in engine throttle control. The power management module collects mains voltage, current, and temperature data through power nodes to power the entire flight control system and provides monitoring and protection functions.

[0063] The positioning module employs an integrated navigation unit, incorporating a multi-mode GNSS receiver, IMU, magnetometer, and barometer. It outputs high-precision, high-frequency position, velocity, attitude, and heading information, with each data point accompanied by a timestamp. The module connects to the central computing module via a CAN bus, providing the flight control system with all the navigation parameters required for autonomous flight, track tracking, and precise recovery.

[0064] The data link module is an independent wireless communication module that connects to the central computing module and is responsible for two-way communication with the ground command and control system.

[0065] The power management module supplies power to the entire flight control system and has monitoring and protection functions.

[0066] In terms of software architecture, the flight control software for the mother-daughter UAV in this application adopts a three-layer architecture: decision layer, control layer, and execution layer. Based on real-time requirements, the priority of each module is divided into interrupt level, real-time level, and non-real-time level. The control layer is further subdivided into the flight control domain and the mission management domain.

[0067] A. Priority Scheduling

[0068] Interruption level: Core flight module (belonging to the flight control domain of the control layer), executes attitude / velocity / trimming closed loop, periodically reads sensor data and its timestamps in the shared data area, and performs timeliness determination and inertial prediction compensation.

[0069] Real-time level: Discrete task module (belonging to the task management domain of the control layer) is responsible for the arrangement of mode state machine and discrete action timing. It generates discrete action instructions in sequence according to the preset step sequence index. It can be preempted by the interrupt level, but the timing of the action sequence is guaranteed. The data acquisition module is promoted to the lowest priority of the real-time level to ensure the freshness of the sensor data timestamp.

[0070] Non-real-time level: Decision layer (task planning module, route generation module) and execution layer (command issuance module), running in the main loop or background thread.

[0071] B. Communication Mechanism

[0072] The modules communicate through a shared data area. Sensor data is written to the shared data area by the data acquisition module and read by the decision-making and control layers; control targets are written to the shared data area by the decision-making layer and read by the control layer. Mode switching, action guidance, and control parameters are transmitted via direct calls or message queues.

[0073] C. Module Responsibilities

[0074] Mission planning module: parses ground commands, determines the overall mission objective, maintains the mission sequence, and decides the next mode to enter.

[0075] The route planning module plans a sequence of waypoints based on the target points given in the mission plan and stores it in the shared data area. The guidance submodule reads the waypoints and current position from the shared data area, performs guidance law calculations, and outputs control targets such as heading and speed.

[0076] The mode state machine receives mode requests from the mission planning module, performs mode switching based on sensor conditions, and sets control targets (attitude, speed, altitude, etc.). It generates mode switching requests (takeoff mode, cruise mode, separation preparation mode, separation preparation mode, separation execution mode, separation completion confirmation mode, return mode, parachute preparation mode, parachute recovery mode) and control targets (target attitude, speed, trim value). During separation, if the comprehensive risk index Λ(t) exceeds the limit, the mode state machine forcibly switches the flight control module's control target from mission tracking to flight attitude control (angle of attack / overload safety protection value), and simultaneously suspends discrete maneuver tasks. After the risk index falls back to the hysteresis recovery threshold, the discrete maneuver tasks are resumed, and the breakpoint is located based on physical sensor feedback.

[0077] The core flight module is always running, periodically reading sensor data and their timestamps from the shared data area, and performing the following timeliness checks: If the data delay does not exceed the first delay threshold, the PID / ADRC control law is directly executed using the data; if the data delay exceeds the first delay threshold but not the second delay threshold, inertial prediction compensation is enabled and the PID integral term is frozen; if the data delay exceeds the second delay threshold, the data acquisition module is restarted. After the timeliness check passes, the control target issued by the decision-making layer is read, the PID / ADRC control law is executed, feedforward compensation is added, and the control surface deflection angle and throttle value are output.

[0078] Discrete Action Module: Receives the mode state and trigger conditions from the mode state machine, arranges the action sequence sequentially according to the preset step index (unlocking the electro-permanent magnet lock → energizing the electromagnetic latch → detecting the microswitch and slave thrust data → master balancing), generates discrete instructions with step numbers, and sends them to the execution layer via a message queue. Synchronization Mechanism: The discrete action module and the mode state machine synchronize through event flag groups. After the mode state machine issues an action request, it waits for the corresponding event to complete; if a timeout occurs, a fault is triggered. If preempted by an interrupt-level task, upon recovery, the next discrete action instruction to be executed is located based on a comprehensive comparison of the real-time feedback status from physical sensors and the current step index. When the electromagnetic latch energizing instruction is executed, the mass feedforward compensation value calculated based on the slave mass and current speed is directly written into the shared data area. After confirming slave disengagement, the asynchronous thrust correction term is smoothly introduced into the shared data area using a first-order inertial filter.

[0079] Data acquisition module: Receives sensor data reported by each remote I / O node via CAN bus polling or interrupt. Simultaneously records the data acquisition timestamp upon reception, and writes it along with the timestamp into the shared data area after verification, providing a basis for data delay determination for the flight control module.

[0080] Command Issuance Module: Monitors the control layer message queue and direct calls from the core flight module, converting control surface / throttle PWM values ​​or discrete action commands into CAN frames and sending them to the corresponding nodes. The control surface / throttle control frames are bound to the smallest CAN ID (highest priority) to ensure that continuous control takes precedence over discrete action commands in bus arbitration; control surface outputs strictly adhere to the servo rate limiting scheduling strategy to prevent step commands from causing servo burnout.

[0081] D. Data Flow and Control Flow

[0082] Sensor data uplink: Remote I / O node → CAN bus → data acquisition module (records timestamp) → shared data area (including timestamp) → control layer / decision layer (verifies timeliness when reading).

[0083] Continuous control downlink: Decision layer → Shared data area (target value) → Core flight module (reading, including timeliness judgment and feedforward superposition) → Direct call → Execution layer (command issuance, including CAN highest priority arbitration and control surface rate limit) → Tail node.

[0084] Discrete control downlink: Decision layer (pattern triggering) → Pattern state machine → Discrete action module (step sequence orchestration) → Message queue → Execution layer (instruction issuance) → Corresponding node.

[0085] Feedforward compensation data stream: Discrete action module (electromagnetic pin energized) → mass feedforward written to shared data area → core flight module reads and superimposes to total control quantity; after slave unit disengagement confirmation → asynchronous thrust correction term smoothed by first-order inertial filter and written to shared data area → core flight module reads and superimposes.

[0086] Figure 3 This is a flowchart illustrating a priority-based task scheduling and discrete action event synchronization control method for mother machine flight control, as shown in one embodiment of this application. Figure 3 As shown, the priority-based task scheduling and discrete action event synchronization control method for mother aircraft flight control in this application mainly includes the following steps:

[0087] S310, during the process of the mother-daughter UAV flying to the separation point, sensor data of the mother-daughter UAV is acquired, wherein the sensor data is updated based on a pre-configured timeliness guarantee mechanism;

[0088] During the mission, the mother and daughter UAVs fly from the takeoff point to the separation point, and are treated as a single unit. At this time, the mode state machine receives instructions from the ground command center to switch the UAVs to takeoff mode, and then to cruise mode after takeoff. During cruise, the mother UAV periodically acquires sensor data and stores it in a shared data area. The correspondence between sensor data types and sources is shown in Table 1.

[0089] Table 1. Sensor Data Types and Sources

[0090]

[0091] During cruise, the aforementioned sensor data is acquired by the data acquisition module via CAN bus polling or interrupt. After CRC verification, it is written to the shared data area along with the acquisition timestamp, allowing the decision-making layer (mission planning, route generation) and the control layer (core flight module, mode state machine, discrete action module) to read it as needed. Because the system employs a priority-based task scheduling mechanism: interrupt-level core flight control tasks prioritize ensuring flight attitude stability, real-time discrete action tasks are responsible for executing separation timings, and non-real-time data acquisition tasks are responsible for transferring sensor data from various remote I / O nodes to the shared data area. However, when the UAV encounters extreme conditions such as strong wind disturbances or sudden changes in quality, interrupt-level flight control tasks and real-time discrete action tasks continuously occupy CPU resources, causing the lowest-priority data acquisition task (which is awakened via CAN receive interrupts or timed polling) to be unable to obtain execution opportunities for an extended period, i.e., data starvation.

[0092] The direct consequence of data starvation is that sensor data in the shared data area remains unupdated for extended periods, timestamps become severely outdated, and the core flight control module, despite operating at peak frequency, can only perform control law calculations based on stale data. This stale data leads to the continuous accumulation of errors in the PID integral term, causing integral saturation. Simultaneously, it causes the ADRC Extended State Observer (ESO) to estimate disturbances significantly behind actual physical changes, ultimately resulting in control surface oscillations or even control divergence. Therefore, this application employs the following method to perform data acquisition to ensure the timeliness of sensor data:

[0093] Obtain the current raw sensor data of the mother-daughter UAV and the acquisition timestamp of the raw sensor data, wherein the raw sensor data is updated periodically;

[0094] S311, Calculate the data delay based on the acquisition timestamp of the original sensor data and the current moment;

[0095] Data latency , For the current moment, For collection timestamps.

[0096] S312, when the data delay does not exceed a preset first delay threshold, the original sensor data is used as sensor data;

[0097] Data latency A smaller timeframe (e.g., less than 5ms) indicates that the data acquisition task is running normally and the sensor data is being updated in a timely manner. Sensor data is time-sensitive.

[0098] S313, when the data delay exceeds the preset first delay threshold, perform inertial compensation on the sensor data and freeze the PID integral term of the mother-daughter UAV control module to obtain the compensated sensor data, and use the compensated sensor data as the sensor data.

[0099] when If the time exceeds the first threshold (e.g., 5ms) but has not yet deteriorated significantly, it indicates that the data acquisition task is experiencing mild starvation due to CPU resources being preempted by higher-priority tasks. In this case, a dual protection strategy is employed:

[0100] Inertial compensation: The estimated state at the current moment is calculated by integrating the effective attitude data of the previous step with the gyroscope angular velocity, and used as a substitute input for the old data, so that the flight control can still obtain relatively reliable feedback values ​​in the short term.

[0101] Freeze the PID integral term: freeze the integral coefficient Temporarily set to zero to prevent the controller from continuously accumulating integral terms based on expired large error signals, thus avoiding severe oscillations caused by integral saturation during data recovery.

[0102] S314, when the data delay exceeds the preset second delay threshold, the sensor link is determined to be faulty, and the data acquisition module is forcibly restarted. After restarting, the system returns to acquiring the current raw sensor data of the mother and daughter UAVs and the acquisition timestamp of the raw sensor data, wherein the second delay threshold is greater than the first delay threshold.

[0103] when If the situation worsens further, exceeding the second threshold (e.g., 50ms), it indicates that the data acquisition task has not received CPU execution rights for an extended period, exceeding the effective range of inertial compensation. At this point, the core flight controller determines that the sensor data link is in a state of substantial failure. In this situation, relying solely on predictive compensation is insufficient to guarantee control accuracy; the data flow must be restored at its source. The system forcibly triggers a restart of the data acquisition module (e.g., resetting the CAN communication stack or reinitializing the DMA buffer). After the data flow is restored, the timeliness determination steps S311~S313 are re-executed, ensuring that the system retains self-recovery capabilities even in extreme preemptive scenarios, rather than relying indefinitely on outdated data for blind control.

[0104] In addition, during the data acquisition module restart, the core flight module freezes all control law updates, outputs the last valid control surface command value (keeping locked) or the preset safety trim value, and gradually releases the integral term after the data is restored to avoid sudden changes in control surfaces due to data interruption.

[0105] During the cruise flight of the mother and daughter UAVs to the separation point, the mission planning module continuously acquires the ready task queue from the mother and daughter mission aircraft. The tasks in the ready task queue come from the ground command center or preset automatic triggering tasks. Then, based on the pre-configured task priority scheduling mechanism, the highest priority task is selected from the ready task queue and executed in conjunction with the sensor data.

[0106] In this application, task priorities, from highest to lowest, are: interrupt level, real-time level, and non-real-time level. Real-time and non-real-time tasks are in the task queue, with real-time tasks executed first, and non-real-time tasks executed only when no real-time tasks are available. Interrupt level is the highest priority; when an interrupt-level task is ready, computing resources are forcibly preempted to run the interrupt-level task, ensuring normal flight attitude.

[0107] S320, when the mother and daughter UAVs fly to the preset separation point, the pre-created discrete action task is awakened and placed in the ready state, wherein the discrete action task generates discrete action instructions in sequence according to the step sequence index and sends them out for execution via the message queue;

[0108] When the mother and daughter drones fly to the preset separation point, the mode state machine switches the drones to separation mode. In separation mode, the preset discrete action tasks (real-time level) inside the drones are ready and begin to be executed if there is no risk to the drones' flight attitude.

[0109] The discrete action instructions in the discrete action task include sequentially executed instructions such as unlocking the electro-permanent magnet lock, energizing the electromagnetic pin, detecting the micro switch and detecting the thrust data of the slave unit, and balancing the mother unit.

[0110] Figure 4 This is a schematic diagram of the execution flow of the separation preparation mode in one embodiment of this application, as shown below. Figure 4 As shown, when the mother and daughter UAVs fly to the preset separation point, the separation preparation mode begins, the mother UAV maintains the cruise speed, and the daughter UAV is started (the daughter UAV performs the separation process), the daughter UAV thrust is established, and when the daughter UAV thrust is greater than the preset thrust threshold, the separation execution mode is entered and discrete actions are performed based on the wake-up discrete action task, and the separation preparation mode ends.

[0111] S330, during the separation process of the mother and daughter UAVs, performs critical point feedforward compensation to suppress mass mutations, and calculates a comprehensive risk index to characterize the instability risk of the mother UAV based on the sensor data. When the comprehensive risk index characterizes the instability risk of the mother UAV, the control target of the flight control module is forcibly switched from mission tracking to the highest priority flight attitude control. During the switching period, the discrete action mission is suspended. The feedforward compensation includes a mass feedforward term that takes effect immediately when the electromagnetic pin is energized, and an asynchronous thrust correction term that is smoothly switched in by first-order inertial filtering after the daughter UAV is confirmed to have separated.

[0112] During the separation of the mother and daughter aircraft, two key discrete actions will significantly perturb the attitude of the mother aircraft:

[0113] Firstly, the moment the electromagnetic pin is energized, the daughter unit is accelerated out, and the mother unit suddenly loses tens of kilograms of load weight, causing the center of gravity to change instantaneously, generating a strong additional pitching moment, which causes the nose of the aircraft to suddenly pitch up or down.

[0114] Secondly, after the engine detaches, its engine jet changes the flow field at the tail of the mother machine, generating an additional pitching aerodynamic torque on the mother machine.

[0115] The aforementioned disturbances are all sudden and non-continuous physical abrupt changes. If only passive responses are relied upon for feedback control (ADRC / PID), the control lag of tens of milliseconds will occur due to the delay between sensor sampling and control cycles, leading to attitude overshoot or oscillation. Therefore, it is necessary to actively inject feedforward compensation at the same time as executing discrete actions to preemptively offset known deterministic disturbances and achieve smooth attitude transition.

[0116] (1) When the electromagnetic pin is energized, calculate the mass feedforward term. The mass feedforward term The mathematical expression is:

[0117]

[0118] In the formula, For the quality of the submachine, At the current speed, The longitudinal lever arm is the distance from the daughter unit's mounting point to the mother unit's center of gravity. The moment of inertia of the mother machine's pitch axis. This is the gain coefficient;

[0119] in, This represents the angular momentum carried away by the slave unit upon detachment. The sudden detachment of the slave unit breaks the conservation of angular momentum in the system, causing the mother unit to generate a reverse angular acceleration. Divide this by the moment of inertia. Calculate the angular acceleration and multiply it by the gain factor. The feedforward rudder input is obtained.

[0120] Gain coefficient , which is the angular velocity-control surface conversion gain. Its physical meaning is: the proportional coefficient that converts the step of angular velocity into elevator deflection angle. It is determined by the control effect (the aerodynamic torque generated per unit angle of elevator deflection) and the flight state (dynamic pressure), and is obtained through wind tunnel testing or flight testing.

[0121] Extract the feedback control quantity at the current moment from the mother aircraft's flight control system. And based on the feedback control quantity and the aforementioned quality feedforward term Calculate the final control quantity , .

[0122] The aforementioned compensation is injected in full instantaneously because the mass loss is a step-like physical abrupt change. The feedforward must take effect synchronously with the discrete action to offset the upward tendency of the nose caused by the sudden weight reduction in advance, thus compensating for the inherent defect of the ADRC observer lagging behind the physical abrupt change.

[0123] (2) When the microswitch detection and submachine thrust data command are executed, calculate the asynchronous thrust correction term. The asynchronous thrust correction term The mathematical expression is:

[0124]

[0125] In the formula, The thrust value transmitted back by the slave unit. The lever arm of the daughter unit's thrust relative to the mother unit's center of gravity. Angular acceleration-rudder surface conversion gain, The moment of inertia of the mother machine's pitch axis. The time constant of the first-order inertial filter is... The execution time of the command to energize the electromagnetic latch. The current moment;

[0126] The additional pitching moment generated by the daughter engine's thrust on the mother engine, divided by the inertia. Calculate the angular acceleration. Then multiply it by the angular acceleration minus the control surface conversion gain. The feedforward compensation is obtained, but the propulsion build-up of the slave unit is a gradual and slow process (as the engine speed increases). Its core mechanism lies in the exponential term. This achieves first-order inertial filtering soft start, which avoids suddenly forcing the daughter machine's thrust data onto the mother machine's control surface at the moment the microswitch is triggered, thus preventing additional angular acceleration step impact on the servo motor.

[0127] Angular acceleration-controller surface conversion gain Its physical meaning is "the proportional coefficient that converts angular acceleration disturbances into elevator deflection angles". It is determined by both rudder effectiveness and flight conditions, and is obtained through wind tunnel testing or flight testing.

[0128] Extract the feedback control quantity at the current moment from the mother aircraft's flight control system. And based on the feedback control quantity and the asynchronous thrust correction term Calculate the final control quantity , .

[0129] (3) Smooth switching: Within the overlapping window of the two feedforward compensations mentioned above, in order to avoid sudden cancellation and cut in At this time, the total feedforward quantity drops and then rises instantaneously, causing a secondary disturbance. This application calculates the total feedforward quantity during the overlapping transition period (overlapping window). The calculation formula is:

[0130]

[0131] In the formula, This refers to the moment when the microswitch is triggered.

[0132] During the overlap transition period, the quality feedforward term According to the index Gradually withdrawing, thrust term according to Gradually integrate to achieve a smooth switching of the total feedforward quantity.

[0133] The feedforward compensation and smooth switching in the two stages mentioned above are both aimed at maintaining the stability of the mother aircraft during separation. This compensation process is not performed as an independent task, but rather embedded within the discrete action task. During separation, in addition to performing discrete action tasks, the core flight module also needs to periodically calculate the mother aircraft's comprehensive risk index based on sensor data to determine if there is a risk of instability. The calculation process includes:

[0134] S331, obtain the current measured vectors of the machine's three-axis attitude and three-axis acceleration. Control surface command vector And obtain the three-axis attitude target vector of the mother machine at the current moment;

[0135] S332, the difference between the measured three-axis attitude vector and the target three-axis attitude vector is calculated to obtain the attitude tracking error. Based on the attitude tracking error The measured vector of the triaxial acceleration and the control surface command vector Calculate the comprehensive risk index used to characterize the risk of machine instability. Among them, the comprehensive risk index The mathematical expression is:

[0136]

[0137] In the formula, For Chebyshev norm, To the maximum attitude tracking error, The first derivative of the attitude tracking error. For the maximum angular divergence rate, This indicates taking the maximum value. This represents the maximum physical deflection angle of the control surface. This represents the maximum excess capacity of the control surface saturation. For standard triaxial acceleration vectors, For the maximum dynamic overload increment, All of these are weight parameters.

[0138] in, The Chebyshev norm (i.e., the largest absolute value) of the attitude tracking errors from the pitch, roll, and yaw channels is used to reflect the axis of greatest deviation from the target at the current attitude. This is then divided by the maximum permissible error. Normalization is achieved to characterize the relative degree of danger of attitude deviation.

[0139] Taking the first derivative and norm of the attitude error reflects the rate at which the attitude is deteriorating. Even if the current deviation is small, if the divergence rate is too high (e.g., ...), the attitude will deteriorate at a rate that is not large enough to prevent the attitude from worsening. This also foreshadows an impending loss of control.

[0140] This indicates that when the flight control calculates the control surface commands... Exceeding the physical maximum deflection angle At this point, there is a gap representing the flight control system being at full capacity but still insufficient. A non-zero value for this value indicates that the control torque is insufficient to counteract the current disturbance, and the linear control range has failed.

[0141] IMU measured triaxial acceleration Deducting reference gravity The norm is then taken to obtain the purely dynamic additional overload. This is used to capture sudden changes in structural stress caused by separation impacts or gusts of wind, preventing the fuselage from disintegrating before the attitude becomes controllable.

[0142] Each of the four items is multiplied by its weight. Summing the results yields a single scalar. ,when This indicates that the risk has exceeded the limit, triggering the suspension of discrete tasks and the switching of control targets.

[0143] Specifically, when When the host aircraft is deemed to have an instability risk, the control objective of the flight control module is forcibly switched from mission tracking to the highest priority flight attitude control. At this time, when the control objective of the flight control module switches to flight attitude control, the flight control module executes attitude stabilization control by invoking a pre-configured flight control stabilization algorithm. This flight control stabilization algorithm includes active disturbance rejection and robust control algorithms, incremental and nonlinear control algorithms, and gain scheduling control methods.

[0144] Specifically, the core flight module detected Then, the control target is immediately switched to the angle of attack / overload safety protection value, the ADRC main algorithm is called and the incremental INDI is used to compensate for angular acceleration, and the response stiffness is adjusted through gain scheduling. Finally, the safe control surface command is output to ensure attitude stability.

[0145] S340, when the machine returns to a stable state, wake up the discrete action task and locate the next discrete action instruction to be executed based on the feedback in the sensor data, and return to calculate the comprehensive risk index used to characterize the machine's instability risk based on the sensor data, until separation is completed.

[0146] After the discrete task is woken up, the real-time status of physical sensors such as limit switches and micro switches at the head node is read and compared with the current step sequence target: if the action has been completed, the step sequence index of the discrete action is advanced; if it has not been completed, the current instruction is resent, so as to achieve precise breakpoint positioning based on physical facts rather than software memory.

[0147] At this point, the core flight module still needs to periodically calculate the comprehensive risk index of the mother machine based on sensor data. When there is a risk, it switches the control target. When there is no risk and it continues for several cycles (e.g., 5 control cycles, about 10-15ms), it continues to execute discrete action tasks, repeating the above process until separation is completed.

[0148] In this application, the step sequence index is stored in the shared data area to ensure that the index value is not lost even if the task is preempted and suspended. When the task resumes, the program first reads the current physical feedback and compares it with the step sequence index in the shared data area: if the feedback status is consistent with the current step sequence target and the action is confirmed to be completed, the index is incremented by 1 and the process moves forward; if the feedback status is inconsistent with the current step sequence target and no timeout has occurred, the current step sequence instruction is resent; if a timeout occurs, a fault is triggered.

[0149] After separation is complete, the mother machine is balancing, then enters return-to-home mode, switching the data link between the slave unit and the mother machine to a wireless data link. The separation mode execution ends. During the subsequent return-to-home process, a data acquisition task is defined and created to collect slave unit status information; this data acquisition task is executed via the wireless data link.

[0150] like Figure 5 As shown, this application also provides a priority-based task scheduling and discrete action event synchronization control system for mother machine flight control, including:

[0151] The acquisition module is used to acquire sensor data of the mother and daughter drones during their flight to the separation point. The sensor data is updated based on a pre-configured timeliness guarantee mechanism.

[0152] The separation task wake-up module is used to wake up the pre-created discrete action task and put it into a ready state when the mother and daughter UAVs fly to the preset separation point. The discrete action task generates discrete action instructions in sequence according to the step sequence index and sends them out for execution through the message queue.

[0153] The separation and priority scheduling module is used to perform critical point feedforward compensation to suppress mass mutations during the separation process of the mother and daughter UAVs, and to calculate a comprehensive risk index to characterize the instability risk of the mother UAV based on the sensor data. When the comprehensive risk index indicates that the mother UAV has an instability risk, the control target of the flight control module is forcibly switched from task tracking to the highest priority flight attitude control. During the switching period, the discrete action task is suspended. The feedforward compensation includes a mass feedforward term that takes effect immediately when the electromagnetic plug is energized, and an asynchronous thrust correction term that is smoothly switched in by first-order inertial filtering after the daughter UAV is confirmed to have separated.

[0154] The loop execution module is used to wake up the discrete action task when the machine recovers to a stable state, locate the next discrete action instruction to be executed based on the feedback in the sensor data, and return to calculate the comprehensive risk index to characterize the machine's instability risk based on the sensor data, until the separation is completed.

[0155] This application presents a priority-based task scheduling and discrete action event synchronization control method and system for mother aircraft flight control. On one hand, it ensures the timeliness of sensor data through a timeliness guarantee mechanism and switches the flight control target to the highest priority attitude control when an instability risk is detected, while suspending discrete tasks to prioritize airframe stability and prevent crashes caused by forced execution of discrete actions. On the other hand, after attitude recovery, it accurately locates the breakpoints of discrete actions based on real-time feedback from physical sensors and resumes execution, avoiding the need for backtracking or retrying strategies due to irreversible physical actions, thus ensuring the strict temporality and integrity of the separation action sequence. Furthermore, critical point feedforward compensation effectively suppresses instantaneous disturbances caused by sudden quality changes, further improving the smoothness and success rate of the separation process. This method significantly improves the reliability and robustness of the mother-daughter aircraft separation task while ensuring flight safety.

[0156] This embodiment also provides an electronic terminal, including: a processor and a memory;

[0157] The memory is used to store computer programs, and the processor is used to execute the computer programs stored in the memory so that the terminal performs any of the methods in this embodiment.

[0158] As will be understood by those skilled in the art, the computer-readable storage medium described in this embodiment allows for the implementation of all or part of the steps in the above method embodiments by computer program-related hardware. The aforementioned computer program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0159] The electronic terminal provided in this embodiment includes a processor, a memory, a transceiver, and a communication interface. The memory and the communication interface are connected to the processor and the transceiver and complete communication between them. The memory is used to store computer programs, the communication interface is used to perform communication, and the processor and the transceiver are used to run the computer programs, so that the electronic terminal performs the steps of the above method.

[0160] In this embodiment, the memory may include random access memory (RAM) and may also include non-volatile memory, such as at least one disk storage device.

[0161] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0162] In the above embodiments, although the present application has been described in conjunction with specific embodiments thereof, many substitutions, modifications, and variations of these embodiments will be apparent to those skilled in the art based on the foregoing description. The embodiments of the present application are intended to cover all such substitutions, modifications, and variations falling within the broad scope of the appended claims.

[0163] The above embodiments are merely illustrative of the principles and effects of this application and are not intended to limit this application. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of this application. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in this application should still be covered by the claims of this application.

Claims

1. A priority-based task scheduling and discrete action event synchronization control method for mother machine flight control, characterized in that, include: During the flight of the mother and daughter drones to the separation point, sensor data of the mother and daughter drones is acquired, wherein the sensor data is updated based on a pre-configured timeliness guarantee mechanism; When the mother and daughter UAVs fly to the preset separation point, the pre-created discrete action task is awakened and placed in the ready state. The discrete action task generates discrete action instructions in sequence according to the step sequence index and sends them out for execution via the message queue. During the separation of the mother and daughter UAVs, critical point feedforward compensation to suppress mass mutations is performed, and a comprehensive risk index to characterize the instability risk of the mother UAV is calculated based on the sensor data. When the comprehensive risk index characterizes the instability risk of the mother UAV, the control target of the flight control module is forcibly switched from mission tracking to the highest priority flight attitude control. During the switching period, discrete action tasks are suspended. The feedforward compensation includes a mass feedforward term that takes effect immediately when the electromagnetic plug is energized, and an asynchronous thrust correction term that is smoothly switched in by first-order inertial filtering after the daughter UAV is confirmed to have separated. When the machine returns to a stable state, the discrete action task is awakened and the next discrete action instruction to be executed is located based on the feedback in the sensor data. Then, the process returns to calculating the comprehensive risk index to characterize the machine's instability risk based on the sensor data until separation is completed.

2. The priority-based task scheduling and discrete action event synchronization control method for mother machine flight control according to claim 1, characterized in that, Acquire sensor data from the mother-daughter drone, including: Obtain the current raw sensor data of the mother-daughter UAV and the acquisition timestamp of the raw sensor data, wherein the raw sensor data is updated periodically; The data delay is calculated based on the acquisition timestamp of the original sensor data and the current moment. When the data delay does not exceed a preset first delay threshold, the original sensor data is used as sensor data; When the data delay exceeds a preset first delay threshold, inertial compensation is performed on the sensor data and the PID integral term of the mother-daughter UAV control module is frozen to obtain the compensated sensor data, and the compensated sensor data is used as the sensor data. When the data delay exceeds a preset second delay threshold, the sensor link is determined to be faulty, and the data acquisition module is forcibly restarted. After restarting, the module returns to acquiring the current raw sensor data of the mother and daughter UAVs and the acquisition timestamp of the raw sensor data. The second delay threshold is greater than the first delay threshold.

3. The priority-based task scheduling and discrete action event synchronization control method for mother machine flight control according to claim 1, characterized in that, Also includes: During the flight of the mother-daughter UAV to the separation point, the ready task queue in the mother-daughter mission machine is acquired, wherein the tasks in the ready task queue come from the ground command center or preset automatic triggering tasks. Based on a pre-configured task priority scheduling mechanism, the highest priority task is selected from the ready task queue and executed in conjunction with the sensor data.

4. The priority-based task scheduling and discrete action event synchronization control method for mother machine flight control according to claim 1, characterized in that, The discrete action commands in the discrete action task include commands to unlock the electro-permanent magnet lock, energize the electromagnetic pin, detect the microswitch and detect the thrust data of the slave unit, and balance the main unit. Among these, the critical point feedforward compensation for suppressing mass mutations includes: When the electromagnetic pin energizing command is executed, the mass feedforward term is calculated. And based on the quality feedforward term Execute the first feedforward compensation; When the microswitch detection and submachine thrust data detection commands are executed, the asynchronous thrust correction term is calculated. And based on the asynchronous thrust correction term A second feedforward compensation is performed, wherein the first feedforward compensation and the second feedforward compensation are smoothly switched within a preset window.

5. The priority-based task scheduling and discrete action event synchronization control method for mother machine flight control according to claim 4, characterized in that, Calculate the quality feedforward term And based on the quality feedforward term Performing the first feedforward compensation includes: Calculate the quality feedforward term The mass feedforward term The mathematical expression is: In the formula, For the quality of the submachine, At the current speed, The longitudinal lever arm is the distance from the daughter unit's mounting point to the mother unit's center of gravity. The moment of inertia of the mother machine's pitch axis. This is the gain coefficient; Extract the feedback control quantity at the current moment from the mother aircraft's flight control system. And based on the feedback control quantity and the aforementioned quality feedforward term Calculate the final control quantity , .

6. The priority-based task scheduling and discrete action event synchronization control method for mother machine flight control according to claim 4, characterized in that, Calculate the asynchronous thrust correction term And based on the asynchronous thrust correction term Performing the second feedforward compensation includes: Calculate the asynchronous thrust correction term The asynchronous thrust correction term The mathematical expression is: In the formula, The thrust value transmitted back by the slave unit. The lever arm of the daughter unit's thrust relative to the mother unit's center of gravity. Angular acceleration-rudder surface conversion gain, The moment of inertia of the mother machine's pitch axis. The time constant of the first-order inertial filter is... The execution time of the command to energize the electromagnetic latch. The current moment; Extract the feedback control quantity at the current moment from the mother aircraft's flight control system. And based on the feedback control quantity and the asynchronous thrust correction term Calculate the final control quantity , .

7. The priority-based task scheduling and discrete action event synchronization control method for mother machine flight control according to claim 1, characterized in that, Based on the sensor data, a comprehensive risk index is calculated to characterize the machine's instability risk, including: Obtain the measured three-axis attitude vectors and measured three-axis acceleration vectors of the machine at the current moment. Control surface command vector And obtain the three-axis attitude target vector of the mother machine at the current moment; The attitude tracking error is obtained by subtracting the measured three-axis attitude vector from the target three-axis attitude vector. Based on the attitude tracking error The measured vector of the triaxial acceleration and the control surface command vector Calculate the comprehensive risk index used to characterize the risk of machine instability. Among them, the comprehensive risk index The mathematical expression is: In the formula, For Chebyshev norm, To the maximum attitude tracking error, The first derivative of the attitude tracking error. For the maximum angular divergence rate, This indicates taking the maximum value. This represents the maximum physical deflection angle of the control surface. This represents the maximum excess capacity of the control surface saturation. For standard triaxial acceleration vectors, For the maximum dynamic overload increment, All of these are weight parameters.

8. The method for priority-based task scheduling and discrete action event synchronization control for mother machine flight control according to claim 1, characterized in that, When the control objective of the flight control module is forcibly switched from mission tracking to the highest priority flight attitude control, the flight control module performs attitude stabilization control by calling a pre-configured flight control stabilization algorithm. The flight control stabilization algorithm includes an active disturbance rejection and robust control algorithm, an incremental and nonlinear control algorithm, and a gain scheduling control method.

9. A priority-based task scheduling and discrete action event synchronization control method for mother machine flight control according to claim 1, characterized in that, When the mother drone flies to the preset separation point, the mother drone maintains its cruise speed and controls the start of the daughter drone. When the thrust of the daughter drone is greater than the preset thrust threshold, it enters the separation execution state and performs discrete actions based on the wake-up discrete action task.

10. A priority-based task scheduling and discrete action event synchronization control system for mother machine flight control, characterized in that, include: The acquisition module is used to acquire sensor data of the mother and daughter drones during their flight to the separation point. The sensor data is updated based on a pre-configured timeliness guarantee mechanism. The separation task wake-up module is used to wake up the pre-created discrete action task and put it into a ready state when the mother and daughter UAVs fly to the preset separation point. The discrete action task generates discrete action instructions in sequence according to the step sequence index and sends them out for execution through the message queue. The separation and priority scheduling module is used to perform critical point feedforward compensation to suppress mass mutations during the separation process of the mother and daughter UAVs, and to calculate a comprehensive risk index to characterize the instability risk of the mother UAV based on the sensor data. When the comprehensive risk index indicates that the mother UAV has an instability risk, the control target of the flight control module is forcibly switched from task tracking to the highest priority flight attitude control. During the switching period, the discrete action task is suspended. The feedforward compensation includes a mass feedforward term that takes effect immediately when the electromagnetic plug is energized, and an asynchronous thrust correction term that is smoothly switched in by first-order inertial filtering after the daughter UAV is confirmed to have separated. The loop execution module is used to wake up the discrete action task when the machine recovers to a stable state, locate the next discrete action instruction to be executed based on the feedback in the sensor data, and return to calculate the comprehensive risk index to characterize the machine's instability risk based on the sensor data, until the separation is completed.

Images