A method and system for remote flight control of unmanned aerial vehicles (UAVs)

Abstract

This invention belongs to the field of unmanned aerial vehicle (UAV) flight control technology. It discloses a remote flight control method and system for UAVs, including: collecting multi-source state information of the UAV during flight and performing time alignment and noise reduction; receiving flight command information from a flight control terminal and mapping the processed multi-source state information and flight command information into a symbolic state set; triggering corresponding flight tasks based on the symbolic state set, and performing semantic parsing and priority control on the flight tasks to convert them into a sequence of task commands with priority logic; collecting real-time state information of the control link between the UAV and the flight control terminal, dynamically scheduling the task command sequence based on the real-time state information, and recording the communication scheduling results; improving mission continuity and flight safety, and effectively reducing mission failure rate and flight risk.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) flight control technology, and more specifically, to a method and system for remote flight control of UAVs. Background Technology

[0002] Existing remote flight control methods and systems for unmanned aerial vehicles (UAVs) mainly suffer from the following problems:

[0003] With the widespread application of drones in scenarios such as remote inspection, emergency rescue, environmental monitoring, and logistics transportation, the demand for remote control during their flight is constantly increasing. Drones typically rely on a wireless control link with a flight control terminal to issue mission commands and transmit status information. However, in complex environments or during long-distance flights, the control link often faces unstable factors such as insufficient bandwidth, increased latency, and rising packet loss rates, which can easily affect the reliability of drone mission execution and flight safety.

[0004] Existing remote flight control methods and systems for unmanned aerial vehicles (UAVs) generally suffer from the following shortcomings: When control link bandwidth is insufficient, a common approach is to directly discard low-priority or non-critical task commands. This can easily lead to missing task information, interruption of task continuity, and increased task failure rate. In cases of fluctuating link status, existing technologies lack flexible and effective adjustment mechanisms, often resulting in resource waste or the inability to issue critical commands in a timely manner. Especially under resource constraints, the lack of a clear mechanism to prioritize safety-related tasks can render UAVs unable to respond promptly under extreme conditions.

[0005] Furthermore, existing solutions often employ full transmission or direct data discarding strategies when bandwidth is limited, lacking more targeted partial transmission or compressed transmission mechanisms, thus lacking flexibility and adaptability. On the other hand, traditional action selection mechanisms typically rely solely on task priority or fixed threshold values ​​for arbitration, failing to incorporate real-time status factors such as the drone's remaining energy and communication link quality into the decision-making process. When the drone's battery is low or the link quality deteriorates, the system may still execute high-energy-consuming, high-bandwidth, or high-risk tasks, increasing the risk of flight failure, mission delays, or even drone loss of control. Because existing methods largely rely on static scoring or fixed priorities, they are prone to making unreasonable action selections in complex dynamic environments, such as executing non-critical tasks under conditions of insufficient energy or deteriorated link quality, thereby affecting the drone's flight safety and mission completion rate.

[0006] In view of this, the present invention proposes a remote flight control method for unmanned aerial vehicles (UAVs) to solve the above problems. Summary of the Invention

[0007] To overcome the aforementioned deficiencies of the prior art and to achieve the above objectives, the present invention provides the following technical solution:

[0008] A method for remote flight control of an unmanned aerial vehicle (UAV) includes:

[0009] S1. Collect multi-source state information of the UAV during flight, and perform time alignment and noise reduction; receive flight command information issued by the flight control terminal, and map the processed multi-source state information and flight command information into a symbolic state set.

[0010] S2. Based on the symbolic state set, trigger the corresponding flight mission, and perform semantic parsing and priority control on the flight mission to convert the flight mission into a sequence of mission instructions with priority logic.

[0011] S3. Collect real-time status information of the control link between the UAV and the flight control terminal, dynamically schedule the mission instruction sequence based on the real-time status information, and record the communication scheduling results;

[0012] S4. Based on the communication scheduling results, construct a hierarchical decision table and use the hierarchical decision table as the arbitration basis. Perform action arbitration through a fuzzy gating mechanism and introduce a dynamic threshold drift mechanism to determine the arbitration action plan.

[0013] S5. Based on the arbitration action plan, generate a flight trajectory that meets the preset constraints and send it to the UAV for execution.

[0014] Specifically, the method for collecting multi-source state information of the UAV during flight and performing time alignment and noise reduction includes:

[0015] Multi-source state information of the UAV during flight is collected, including flight attitude parameters, spatial position parameters, environmental perception parameters, and power energy parameters. A unified timestamp is added to the collected multi-source state information, and interpolation compensation and resampling techniques are used to map information with different sampling frequencies to a unified time axis.

[0016] To address environmental noise, sensor jitter, and communication interference in multi-source state information, a Kalman filter method is employed for denoising. After time alignment and denoising are completed, the processing results are verified for consistency, and distorted or missing information is removed.

[0017] Specifically, the method for obtaining the symbolic state set includes:

[0018] The UAV receives flight command information sent by the flight control terminal through the control link between the UAV and the flight control terminal. The flight command information includes target position, speed command, attitude command, heading adjustment command, mission switching command and safety constraint command.

[0019] The received flight command information is decoded and structured. The denoised and time-aligned multi-source state information is linearly normalized to map continuous values ​​of different dimensions to a unified interval. The normalized continuous values ​​are further mapped to discrete symbolic categories. The processed multi-source state information is combined with the structured flight command information through symbolic encoding to generate a unified symbolic state set.

[0020] Specifically, the method for triggering the corresponding flight mission based on the symbolic state set includes:

[0021] For each symbolic element in the generated symbolic state set, a task triggering rule is matched. The symbolic element contains the current flight state information of the UAV, the flight command type, and the preset priority information of the flight mission. According to the predefined task triggering rule library, the flight missions that meet the task triggering conditions are marked as missions to be executed. Each rule in the task triggering rule library corresponds to a combination of state symbols and a task type.

[0022] The mission is triggered when the symbolic elements in the symbolic state set simultaneously satisfy the predefined state symbol conditions and flight command symbol conditions. When the symbolic elements in the symbolic state set simultaneously satisfy the state symbol conditions and flight command symbol conditions, the corresponding flight mission is triggered.

[0023] Specifically, the method for obtaining the task instruction sequence includes:

[0024] Semantic parsing is performed on each flight mission, which is parsed into specific executable flight actions and corresponding target parameters. Executable flight actions include speed adjustment, heading adjustment, position adjustment, return to home, and landing.

[0025] Based on the preset priority information of each flight mission, the priority of the parsed flight missions is adjusted. When there are conflicts between different flight missions, the higher priority mission will override the lower priority mission. According to the adjusted priority order, the parsed flight missions are generated into a sequence of mission instructions.

[0026] Specifically, the method for obtaining the communication scheduling result includes:

[0027] Real-time status information of the control link between the UAV and the flight control terminal is collected. The real-time status information includes the available bandwidth of the link, the link latency, and the packet loss rate. Based on the real-time status information, the generated task instruction sequence is dynamically scheduled for communication through bandwidth scheduling rules, and the communication scheduling results are recorded.

[0028] Specifically, the method of constructing a hierarchical decision table and using the hierarchical decision table as the basis for arbitration includes:

[0029] Extract the required bandwidth, link latency, and packet loss rate of each task instruction from the communication scheduling results; divide the flight missions into different levels according to the functional category and scheduling weight of the task instructions, including safety control tasks, navigation control tasks, and auxiliary information tasks;

[0030] Within each level, the scheduling weight of task instructions, the bandwidth required by each task instruction, the link latency and packet loss rate are weighted and fused to obtain the intra-level priority score. The tasks are then sorted according to the intra-level priority score to obtain an intra-level ordered task sequence.

[0031] Priority rules based on the importance of task categories are established between different levels, prioritizing safety control tasks over navigation control tasks, and navigation control tasks over auxiliary information tasks; and a hierarchical decision table is generated based on the intra-level sorting and priority rules.

[0032] Specifically, the method for determining the arbitration action plan includes:

[0033] Based on the generated hierarchical decision table, executable actions corresponding to task instructions are extracted from each priority level to form an action candidate set; for each candidate action in the action candidate set, symbolic state set, communication scheduling results and power energy parameter information are used as input variables through fuzzy logic reasoning technology.

[0034] A fuzzy mapping function is defined for each input variable, and the fuzzy membership degree of the input variable of the candidate action is obtained through the fuzzy mapping function. According to the preset fuzzy rule library, the input conditions of each rule are matched with the fuzzy membership degree of the input variable of the candidate action, the membership degree of the candidate action is calculated, and a numerical action suitability score is obtained through defuzzification.

[0035] A dynamic threshold drift mechanism is introduced to correct the action suitability score. The suitability score corrected by the dynamic threshold is used as the arbitration basis to select the final arbitration action from the action candidate set and determine the arbitration action plan.

[0036] Specifically, the method for generating a flight trajectory that meets preset constraints and sending it to the UAV for execution includes:

[0037] After determining the arbitration action plan, preset constraints are set for the flight trajectory, including dynamic constraints, energy constraints, no-fly zone constraints, safety constraints, and communication constraints. Under the preset constraints, a flight trajectory sequence that meets the preset constraints is generated based on the arbitration action plan, combined with the current position of the UAV, the position of the target point, and the position of the return point.

[0038] All flight trajectories are subjected to trajectory constraint verification. If a violation of the preset constraint conditions is found, the flight trajectory is iteratively corrected until the preset constraint conditions are met. After the trajectory constraint verification is completed, the flight trajectory is converted into flight trajectory control commands that can be executed by the UAV and sent to the UAV for execution through the control link.

[0039] A remote flight control system for unmanned aerial vehicles (UAVs) includes:

[0040] The perception preprocessing unit collects multi-source state information of the UAV during flight and performs time alignment and noise reduction; it receives flight command information from the flight control terminal and maps the processed multi-source state information and flight command information into a symbolic state set.

[0041] The task parsing and control unit triggers the corresponding flight task based on the symbolic state set, and performs semantic parsing and priority control on the flight task, converting the flight task into a sequence of task instructions with priority logic.

[0042] The link communication scheduling unit collects real-time status information of the control link between the UAV and the flight control terminal, performs dynamic communication scheduling of the mission instruction sequence based on the real-time status information, and records the communication scheduling results.

[0043] The hierarchical decision-making arbitration unit constructs a hierarchical decision table based on communication scheduling results and uses the hierarchical decision table as the arbitration basis. It conducts action arbitration through a fuzzy gating mechanism and introduces a dynamic threshold drift mechanism to determine the arbitration action plan.

[0044] The trajectory generation and execution unit generates a flight trajectory that meets preset constraints according to the arbitration action plan, and sends it to the UAV for execution.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] This invention dynamically adjusts the bandwidth allocation for task commands by real-time collection of available link bandwidth, latency, and packet loss rate. It can adaptively schedule commands based on changes in link status, avoiding link overload and congestion. By introducing scheduling weights and critical task thresholds, when the importance of a task command reaches the threshold, it is automatically identified as a critical task and prioritized for bandwidth allocation, ensuring safety and reliability in remote flight control. Through a bandwidth compression mechanism for non-critical tasks, limited link resources are prioritized for critical tasks while minimizing their usage while ensuring the transmission of non-critical tasks, achieving efficient utilization of link resources. In environments with fluctuating link conditions, it can minimize the risk of delay and loss of critical control commands, thereby ensuring real-time response and stable control of the UAV during remote flight.

[0047] By using a residual energy mapping function, the UAV's battery percentage is converted into a threshold correction value in real time. This allows the arbitration result to automatically suppress high-energy-consuming tasks and prioritize safety and navigation control actions when the battery is low. A link quality mapping function maps real-time communication metrics such as bandwidth, latency, and packet loss rate to threshold correction values. When the link deteriorates, the arbitration threshold is dynamically adjusted upwards to avoid allocating high-bandwidth, low-reliability actions, thereby improving system robustness. Fuzzy logic reasoning is employed to jointly model state information, scheduling results, and energy parameters. Action suitability scores are obtained through fuzzy membership and defuzzification calculations, enabling comprehensive judgment of complex, multi-source information. In cases of low battery or degraded link quality, high-risk actions are automatically suppressed, and safety-critical tasks are prioritized, reducing flight mission failure rates and achieving more robust and reliable flight control. Attached Figure Description

[0048] Figure 1 This is a schematic diagram of a remote flight control method for unmanned aerial vehicles (UAVs) according to the present invention.

[0049] Figure 2 This is a schematic diagram of the structure of a remote flight control system for unmanned aerial vehicles (UAVs) according to the present invention. Detailed Implementation

[0050] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0051] Example 1: Please refer to Figure 1 As shown, this embodiment provides a remote flight control method for unmanned aerial vehicles (UAVs), which specifically includes the following steps:

[0052] S1. Collect multi-source state information of the UAV during flight, and perform time alignment and noise reduction; receive flight command information issued by the flight control terminal, and map the processed multi-source state information and flight command information into a symbolic state set.

[0053] S2. Based on the symbolic state set, trigger the corresponding flight mission, and perform semantic parsing and priority control on the flight mission to convert the flight mission into a sequence of mission instructions with priority logic.

[0054] S3. Collect real-time status information of the control link between the UAV and the flight control terminal, dynamically schedule the mission instruction sequence based on the real-time status information, and record the communication scheduling results;

[0055] S4. Based on the communication scheduling results, construct a hierarchical decision table and use the hierarchical decision table as the arbitration basis. Perform action arbitration through a fuzzy gating mechanism and introduce a dynamic threshold drift mechanism to determine the arbitration action plan.

[0056] S5. Based on the arbitration action plan, generate a flight trajectory that meets the preset constraints and send it to the UAV for execution.

[0057] Methods for collecting multi-source state information of UAVs during flight and performing time alignment and noise reduction include:

[0058] Multi-source state information of the UAV during flight is collected, including flight attitude parameters, spatial position parameters, environmental perception parameters, and power energy parameters. A unified timestamp is added to the collected multi-source state information, and interpolation compensation and resampling techniques are used to map information with different sampling frequencies to a unified time axis.

[0059] To address environmental noise, sensor jitter, and communication interference in multi-source state information, a Kalman filter method is employed for denoising. After time alignment and denoising are completed, the processing results are verified for consistency, and distorted or missing information is removed.

[0060] Flight attitude parameters are collected by an inertial measurement unit (IMU), which includes a three-axis accelerometer, a three-axis gyroscope, and a three-axis magnetometer, to obtain the UAV's pitch angle, roll angle, yaw angle, and angular velocity. Spatial position parameters are collected by a global navigation satellite system (GNSS) to obtain latitude, longitude, and flight altitude, and are combined with a barometric altimeter and optical flow sensor to obtain relative altitude and velocity information.

[0061] Environmental sensing parameters are collected by wind speed sensors, temperature and humidity sensors, and magnetic field strength sensors to obtain information on external wind speed and direction, ambient temperature and humidity, and magnetic field interference. Power energy parameters are collected by current sensors, voltage sensors, power sensors, and motor speed sensors to obtain information on battery current, voltage, remaining power, power consumption rate, and motor operating status.

[0062] Methods for obtaining symbolic state sets include:

[0063] The UAV receives flight command information sent by the flight control terminal through the control link between the UAV and the flight control terminal. The flight command information includes target position, speed command, attitude command, heading adjustment command, mission switching command and safety constraint command.

[0064] The received flight command information is decoded and structured. The denoised and time-aligned multi-source state information is linearly normalized, mapping continuous values ​​of different dimensions to a unified interval. The normalized continuous values ​​are further mapped to discrete symbolic categories (e.g., altitude "low / medium / high", speed "slow / medium / fast", battery "low / normal / fully charged"). Using symbolic encoding, the processed multi-source state information is combined with the structured flight command information to generate a unified symbolic state set. This symbolic state set simultaneously describes the UAV's current flight state and mission execution requirements; each symbolic element contains the corresponding state category, command type, and priority information.

[0065] The methods for triggering the corresponding flight mission based on the symbolic state set include:

[0066] For each symbolic element in the generated symbolic state set, a task triggering rule is matched. The symbolic element contains the current flight state information of the UAV, the flight command type, and the preset priority information of the flight mission. According to the predefined task triggering rule library, the flight missions that meet the task triggering conditions are marked as missions to be executed. Each rule in the task triggering rule library corresponds to a combination of state symbols and a task type.

[0067] The mission is triggered when the symbolic elements in the symbolic state set simultaneously satisfy the predefined state symbol conditions and flight command symbol conditions. When the symbolic elements in the symbolic state set simultaneously satisfy the state symbol conditions and flight command symbol conditions, the corresponding flight mission is triggered.

[0068] The status symbol conditions include the symbolic categories of the UAV's current altitude, speed, battery level, and attitude status parameters; the flight command symbol conditions include the symbolic representation of speed adjustment, heading adjustment, position adjustment, and mission switching command types.

[0069] For example, when the drone's altitude is symbolized as high, its speed as fast, its battery level as low, and the received flight command symbol is a position adjustment command, the corresponding deceleration mission, return-to-home mission, landing mission, and position adjustment mission are triggered according to the preset rules in the mission triggering rule base.

[0070] Methods for obtaining task instruction sequences include:

[0071] Semantic parsing is performed on each flight mission, which is parsed into specific executable flight actions and corresponding target parameters. Executable flight actions include speed adjustment, heading adjustment, position adjustment, return to home, and landing.

[0072] Based on the preset priority information of each flight mission, the priority of the parsed flight missions is adjusted, with safety-related missions having the highest priority, navigation and position adjustment missions having the second highest priority, and non-critical auxiliary missions having the lowest priority. When there are conflicts between different flight missions, the higher priority missions will override the lower priority missions. According to the adjusted priority order, the parsed flight missions are generated into a sequence of mission instructions.

[0073] The mission instruction sequence can be dynamically updated as the UAV's flight status changes or as a new symbolic state set is generated. High-priority tasks can be inserted into the instruction sequence in real time, while low-priority tasks can be delayed or have their execution order adjusted, thereby enabling multi-task closed-loop control and dynamic response of the UAV in complex environments.

[0074] Methods for obtaining communication scheduling results include:

[0075] Real-time status information of the control link between the UAV and the flight control terminal is collected. The real-time status information includes the available bandwidth of the link, the link latency, and the packet loss rate. Based on the real-time status information, the generated task instruction sequence is dynamically scheduled for communication through bandwidth scheduling rules, and the communication scheduling results are recorded.

[0076] The bandwidth scheduling rules are as follows: ;in, Indicates a point in time The available bandwidth of the control link is the total bandwidth resource that the control link can allocate to task instructions at that point in time. This means that the total bandwidth after all tasks are assigned cannot exceed the available bandwidth of the control link at that point in time; Indicates task instructions The required bandwidth is the amount of bandwidth required for the task to transmit under normal conditions. Indicates the first One task instruction; Indicates task instructions The scheduling weight; This represents the critical task threshold, which is set when any task instruction... When the scheduling weight is greater than or equal to the critical task threshold, the task instruction is... This instruction has been identified as a critical mission instruction. This indicates the actual task instructions given after scheduling and allocation. bandwidth; This indicates the bandwidth compression ratio for non-critical tasks; according to expert experience, the value ranges from 0 to 1. Indicates the index of the task instruction; This indicates the total number of task instructions that need to be scheduled.

[0077] Methods for constructing hierarchical decision tables and using them as arbitration criteria include:

[0078] Extract the required bandwidth, link latency, and packet loss rate of each task instruction from the communication scheduling results; divide the flight missions into different levels according to the functional category and scheduling weight of the task instructions, including safety control tasks, navigation control tasks, and auxiliary information tasks;

[0079] Within each level, the scheduling weight of task instructions, the bandwidth required by each task instruction, the link latency and packet loss rate are weighted and fused to obtain the intra-level priority score. The tasks are then sorted according to the intra-level priority score to obtain an intra-level ordered task sequence.

[0080] Priority rules based on task category importance are established across different levels, prioritizing safety control tasks over navigation control tasks, and navigation control tasks over auxiliary information tasks. A hierarchical decision table is generated based on the intra-level sorting and priority rules. The hierarchical decision table records task category, task priority, task scheduling status, and recommended execution order in a structured format.

[0081] Methods for determining arbitration action plans include:

[0082] Based on the generated hierarchical decision table, executable actions corresponding to task instructions are extracted from each priority level to form an action candidate set; for each candidate action in the action candidate set, symbolic state set, communication scheduling results and power energy parameter information are used as input variables through fuzzy logic reasoning technology.

[0083] A fuzzy mapping function is defined for each input variable, and the fuzzy membership degree of the input variable of the candidate action is obtained through the fuzzy mapping function. According to the preset fuzzy rule library, the input conditions of each rule are matched with the fuzzy membership degree of the input variable of the candidate action to calculate the membership degree of the candidate action, and a numerical action suitability score is obtained through defuzzification. (For example, if height = high, speed = fast, and battery = low, then action priority = medium; if bandwidth satisfaction = low and task category = safe, then action priority = high; if link latency = high and battery = low, then action priority = low).

[0084] The fuzzy mapping function is: ;in, Indicates candidate actions Fuzzy membership degree; Represents fuzzy membership functions; Represents a symbolic set of states; Indicates the communication scheduling result; Indicates kinetic energy parameter information; Indicates the first One candidate action; Indicates the index of the candidate action;

[0085] A dynamic threshold drift mechanism is introduced to correct the action suitability score. The suitability score corrected by the dynamic threshold is used as the arbitration basis to select the final arbitration action from the action candidate set and determine the arbitration action plan.

[0086] The dynamic threshold drift mechanism is as follows: ;in, Indicates a point in time The dynamic threshold value is used to correct the action suitability score, ensuring that safer actions are prioritized when energy is insufficient or the link deteriorates. The function representing the remaining energy mapping of the drone at a given time point... Remaining energy Mapped to the corresponding threshold increment; This represents the energy adjustment factor, used to adjust the weight of the impact of the drone's remaining energy on the threshold value; Indicates the time point of the drone The remaining energy, such as the battery percentage; This represents the link quality mapping function, which maps link quality metrics. Mapped to the corresponding threshold increment; This represents the communication link adjustment factor, used to adjust the weight of the link quality on the threshold value. This indicates the link quality metrics, reflecting the communication status of the control link between the UAV and the flight control terminal, including the required bandwidth, link latency, and packet loss rate for each mission command.

[0087] The final arbitration action was as follows: ;in, This indicates that the final selected arbitration action is the optimal action under the current state, link conditions, and energy constraints. Represents the set of action candidates;

[0088] It should be noted that the introduction of a dynamic threshold drift mechanism to correct action suitability scores aims to address the following technical issues: Traditional action arbitration typically uses fixed thresholds or relies solely on task priority to select actions. When the drone's battery is low or the link quality deteriorates, fixed thresholds may lead to the selection of high-risk actions, increasing the risk of flight failure or mission interruption. Existing methods lack dynamic awareness of the drone's energy consumption and the real-time status of the communication link. In complex or harsh environments, fixed scores or static arbitration can easily lead to the system selecting inappropriate actions, such as executing high-bandwidth, high-energy-consuming tasks, which may cause mission delays or drone loss of control. Dynamic thresholds drift with energy and link status, suppressing high-risk actions when battery is low or the link is deteriorating, ensuring that safety-critical actions are executed first, reducing the flight mission failure rate, and enabling the action arbitration system to adapt to environmental changes, achieving more reliable and robust flight control.

[0089] Methods for generating flight trajectories that meet preset constraints and sending them to the drone for execution include:

[0090] After determining the arbitration action plan, preset constraints are set for the flight trajectory, including dynamic constraints, energy constraints, no-fly zone constraints, safety constraints, and communication constraints. Under the preset constraints, a flight trajectory sequence that meets the preset constraints is generated based on the arbitration action plan, combined with the current position of the UAV, the position of the target point, and the position of the return point.

[0091] Dynamic constraints include the UAV’s maximum speed, maximum acceleration, and maximum rate of climb; energy constraints include the maximum flight distance and mission range that the UAV can cover with its current remaining energy; safety constraints include the minimum safe altitude and minimum safe distance; and communication constraints include maintaining a link connection with the flight control terminal at all times.

[0092] All flight trajectories are subjected to trajectory constraint verification. If a violation of the preset constraint conditions is found, the flight trajectory is iteratively corrected until the preset constraint conditions are met. After the trajectory constraint verification is completed, the flight trajectory is converted into flight trajectory control commands that can be executed by the UAV and sent to the UAV for execution through the control link.

[0093] The preset critical task threshold is set by staff based on historical data analysis results. This historical analysis process includes the system collecting the scheduling weights of multiple task instructions and calculating their average value as a reference to obtain the preset critical task threshold, which can be adjusted by staff during system operation according to the actual situation.

[0094] This embodiment dynamically adjusts the bandwidth allocation for task commands by real-time collection of available link bandwidth, latency, and packet loss rate. It can adaptively schedule commands based on changes in link status, avoiding link overload and congestion. By introducing scheduling weights and critical task thresholds, when the importance of a task command reaches the threshold, it is automatically identified as a critical task and prioritized for bandwidth allocation, ensuring safety and reliability in remote flight control. Through a bandwidth compression mechanism for non-critical tasks, limited link resources are prioritized for critical tasks, while reducing their usage while ensuring transmission capability for non-critical tasks, achieving efficient utilization of link resources. In environments with fluctuating link conditions, it can minimize the risk of delay and loss of critical control commands, thereby ensuring real-time response and stable control of the UAV during remote flight.

[0095] By using a residual energy mapping function, the UAV's battery percentage is converted into a threshold correction value in real time. This allows the arbitration result to automatically suppress high-energy-consuming tasks and prioritize safety and navigation control actions when the battery is low. A link quality mapping function maps real-time communication metrics such as bandwidth, latency, and packet loss rate to threshold correction values. When the link deteriorates, the arbitration threshold is dynamically adjusted upwards to avoid allocating high-bandwidth, low-reliability actions, thereby improving system robustness. Fuzzy logic reasoning is employed to jointly model state information, scheduling results, and energy parameters. Action suitability scores are obtained through fuzzy membership and defuzzification calculations, enabling comprehensive judgment of complex, multi-source information. In cases of low battery or degraded link quality, high-risk actions are automatically suppressed, and safety-critical tasks are prioritized, reducing flight mission failure rates and achieving more robust and reliable flight control.

[0096] Example 2: Please refer to Figure 2 As shown, parts not described in detail in this embodiment are described in Embodiment 1. A remote flight control system for unmanned aerial vehicles is provided, including:

[0097] The perception preprocessing unit collects multi-source state information of the UAV during flight and performs time alignment and noise reduction; it receives flight command information from the flight control terminal and maps the processed multi-source state information and flight command information into a symbolic state set.

[0098] The task parsing and control unit triggers the corresponding flight task based on the symbolic state set, and performs semantic parsing and priority control on the flight task, converting the flight task into a sequence of task instructions with priority logic.

[0099] The link communication scheduling unit collects real-time status information of the control link between the UAV and the flight control terminal, performs dynamic communication scheduling of the mission instruction sequence based on the real-time status information, and records the communication scheduling results.

[0100] The hierarchical decision-making arbitration unit constructs a hierarchical decision table based on communication scheduling results and uses the hierarchical decision table as the arbitration basis. It conducts action arbitration through a fuzzy gating mechanism and introduces a dynamic threshold drift mechanism to determine the arbitration action plan.

[0101] The trajectory generation and execution unit generates a flight trajectory that meets preset constraints according to the arbitration action plan, and sends it to the UAV for execution.

[0102] Since the electronic device described in this embodiment is the electronic device used in implementing the UAV remote flight control method and system described in this application, those skilled in the art can understand the specific implementation methods and various variations of the electronic device in this embodiment based on the UAV remote flight control method and system described in this application. Therefore, how the electronic device implements the method in this application will not be described in detail here. Any electronic device used by those skilled in the art in implementing the UAV remote flight control method and system described in this application falls within the scope of protection of this application.

[0103] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters and thresholds in the formulas are set by those skilled in the art according to the actual situation.

[0104] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A method for remote flight control of an unmanned aerial vehicle (UAV), characterized in that, The method comprises the following steps: S1, collecting multi-source state information of the unmanned aerial vehicle in the flight process, and performing time alignment and denoising; receiving flight instruction information sent by the flight control terminal, and uniformly mapping the processed multi-source state information and the flight instruction information into a symbolic state set; S2, triggering a corresponding flight task according to the symbolic state set, and performing semantic analysis and priority control on the flight task, and converting the flight task into a task instruction sequence with priority logic; S3, collecting real-time state information of the control link between the unmanned aerial vehicle and the flight control terminal, dynamically scheduling the task instruction sequence based on the real-time state information, and recording the communication scheduling result; The method for obtaining the communication scheduling result comprises: collecting real-time state information of the control link between the unmanned aerial vehicle and the flight control terminal, wherein the real-time state information comprises link available bandwidth, link delay and packet loss rate; based on the real-time state information, dynamically scheduling the generated task instruction sequence through bandwidth scheduling rules, and recording the communication scheduling result; S4, based on the communication scheduling result, constructing a hierarchical decision table, and taking the hierarchical decision table as an arbitration basis, performing action arbitration through a fuzzy gating mechanism, introducing a dynamic threshold drift mechanism, and determining an arbitration action scheme; The method for determining the arbitration action scheme comprises: extracting executable actions corresponding to the task instructions from each priority layer to form an action candidate set according to the generated hierarchical decision table; for each candidate action in the action candidate set, taking the symbolic state set, the communication scheduling result and the power energy parameter information as input variables through fuzzy logic reasoning technology; defining a fuzzy mapping function for each input variable, obtaining the fuzzy membership degree of the input variable of the candidate action through the fuzzy mapping function; according to a preset fuzzy rule base, matching the input condition of each rule with the fuzzy membership degree of the input variable of the candidate action, calculating the membership degree of the candidate action, and obtaining a numerical action suitability score through defuzzification; introducing a dynamic threshold drift mechanism to correct the action suitability score, taking the suitability score corrected by the dynamic threshold as an arbitration basis, selecting a final arbitration action from the action candidate set, and determining an arbitration action scheme; S5, generating a flight trajectory meeting the preset constraint condition according to the arbitration action scheme, and delivering it to the unmanned aerial vehicle for execution.

2. The method of claim 1, wherein, The method for collecting multi-source state information of the unmanned aerial vehicle in the flight process and performing time alignment and denoising comprises: collecting multi-source state information of the unmanned aerial vehicle in the flight process, wherein the multi-source state information comprises flight attitude parameter information, spatial position parameter information, environmental perception parameter information and power energy parameter information; adding a uniform format timestamp to the collected multi-source state information, and mapping information with different sampling frequencies to a unified time axis by using interpolation compensation and resampling technology; for the environmental noise, sensor jitter and communication interference existing in the multi-source state information, a Kalman filtering method is used for denoising processing; and after completing the time alignment and denoising, the processing result is subjected to consistency checking to eliminate distorted or missing information.

3. The method of claim 2, wherein, The method for obtaining the symbolic state set comprises: Receiving flight instruction information sent by the flight control terminal through a control link between the unmanned aerial vehicle and the flight control terminal, the flight instruction information including target position, speed instruction, attitude instruction, heading adjustment instruction, task switching instruction and safety constraint instruction; Decoding and structuring the received flight instruction information, linearly normalizing the denoised and time-aligned multi-source state information, mapping continuous numerical values of different dimensions to a unified interval, and further mapping the normalized continuous numerical values to discrete symbolic categories; combining the processed multi-source state information and the structured flight instruction information through a symbolic coding method to generate a unified symbolic state set.

4. The method of claim 3, wherein, The method of triggering corresponding flight tasks according to the symbolic state set includes: Matching each symbolic element in the generated symbolic state set with a task triggering rule, the symbolic element containing the current flight state information of the unmanned aerial vehicle, the flight instruction type and the preset priority information of the flight task; marking the flight task that meets the task triggering condition as a to-be-executed task according to a pre-defined task triggering rule library, each rule in the task triggering rule library corresponding to a state symbol combination and a task type; The task triggering condition is that the symbolic elements in the symbolic state set meet the pre-defined state symbol condition and flight instruction symbol condition at the same time; when the symbolic elements in the symbolic state set meet the state symbol condition and the flight instruction symbol condition at the same time, the corresponding flight task is triggered.

5. The method of claim 4, wherein, The method of obtaining the task instruction sequence includes: Performing semantic analysis on each flight task to parse the flight task into a specific executable flight action and a corresponding target parameter, the executable flight action including speed adjustment, heading adjustment, position adjustment, return action and landing action; According to the preset priority information of each flight task, the parsed flight task is prioritized, and when there is a conflict between different flight tasks, a high-priority task is used to override a low-priority task; the parsed flight task is sequentially generated into a task instruction sequence according to the prioritized order.

6. The method of claim 5, wherein, The method of constructing a hierarchical decision table and using the hierarchical decision table as an arbitration basis includes: Extracting the required bandwidth, link delay and packet loss rate of each task instruction from the communication scheduling result; dividing the flight tasks into different levels according to the function category and scheduling weight of the task instruction, the different levels including safety control type tasks, navigation control type tasks and auxiliary information type tasks; Within each level, the scheduling weight of the task instruction, the required bandwidth, the link delay and the packet loss rate of each task instruction are weighted and fused to obtain an intra-level priority score, and the tasks are sorted according to the intra-level priority score to obtain an intra-level ordered task sequence; Between different levels, a priority rule based on the importance of the task category is established, which makes the safety control type task prior to the navigation control type task, and the navigation control type task prior to the auxiliary information type task; and a hierarchical decision table is generated according to the intra-level sorting and the priority rule.

7. The method of claim 6, wherein, The method of generating a flight trajectory that meets the preset constraint condition and delivering it to the unmanned aerial vehicle for execution includes: After determining the arbitration action scheme, preset constraint conditions of the flight trajectory are set, and the preset constraint conditions include dynamic constraints, energy constraints, no-fly zone constraints, safety constraints and communication constraints; under the preset constraint conditions, a flight trajectory sequence conforming to the preset constraint conditions is generated based on the arbitration action scheme in combination with the current position of the unmanned aerial vehicle, the target point position and the return point position; Trajectory constraint verification is performed on all flight trajectories, and if it is found that the preset constraint conditions are violated, the flight trajectories are iteratively modified until the preset constraint conditions are met; after the trajectory constraint verification is completed, the flight trajectories are converted into flight trajectory control instructions executable by the unmanned aerial vehicle, and are issued to the unmanned aerial vehicle through a control link for execution.

8. A remote flight control system for a UAV, for implementing the remote flight control method of any one of claims 1 to 7, characterized in that, Comprise: A perception preprocessing unit acquires multi-source state information of the unmanned aerial vehicle in the flight process, and performs time alignment and denoising; Receiving flight instruction information issued by the flight control terminal, the processed multi-source state information and the flight instruction information are uniformly mapped into a symbolized state set; A task analysis and regulation control unit triggers corresponding flight tasks according to the symbolized state set, and performs semantic analysis and priority regulation on the flight tasks, and converts the flight tasks into a task instruction sequence with priority logic; A link communication scheduling unit acquires real-time state information of the control link between the unmanned aerial vehicle and the flight control terminal, dynamically schedules the task instruction sequence based on the real-time state information, and records the communication scheduling result; A hierarchical decision arbitration unit constructs a hierarchical decision table based on the communication scheduling result, and uses the hierarchical decision table as an arbitration basis to perform action arbitration through a fuzzy gating mechanism, introduces a dynamic threshold drift mechanism, and determines an arbitration action scheme; A trajectory generation and execution unit generates a flight trajectory conforming to the preset constraint conditions according to the arbitration action scheme, and issues the flight trajectory to the unmanned aerial vehicle for execution.

Images