Unmanned aerial vehicle flight control and energy deployment method and system based on energy consumption perception

Abstract

The application relates to the technical field of unmanned aerial vehicle flight control, in particular to an unmanned aerial vehicle flight control and energy deployment method and system based on energy consumption perception; the method comprises the following steps: performing abnormal energy consumption detection based on energy consumption data to obtain abnormal energy consumption identification; identifying a flight stage and constructing an energy consumption baseline of each flight stage; calculating an energy efficiency coefficient in combination with operation environment data, and calculating flight parameters based on the energy efficiency coefficient; decomposing residual tasks into task segments, and assigning flight stages and durations to the task segments; calculating total expected energy consumption in combination with the flight parameters and the energy consumption baseline, and constructing an energy consumption map based on the total expected energy consumption; and calculating energy redundancy, when the energy redundancy of any unmanned aerial vehicle is lower than an energy threshold or abnormal energy consumption identification exists, performing task reassignment based on the energy consumption map, and updating residual tasks; the application realizes energy deployment from single-machine energy efficiency optimization to swarm energy balance, and improves the energy utilization efficiency and task reliability of an unmanned aerial vehicle flight control system.

Description

Technical Field

[0001] This application relates to the field of unmanned aerial vehicle (UAV) flight control technology, specifically to a method and system for UAV flight control and energy allocation based on energy consumption perception. Background Technology

[0002] In recent years, the pace of technological iteration and upgrading of drones has continued to accelerate, and they have been widely used in many civilian and industrial fields such as material delivery, agricultural plant protection, and power line inspection. As drone operation modes have gradually evolved from traditional single-drone independent operation to multi-drone swarm collaborative operation, the operation scenarios have also expanded from conventional and simple working conditions to complex, harsh, and dynamically interfered environments. The industry has put forward more stringent technical requirements for the endurance, flight stability, and mission execution reliability of drones.

[0003] Endurance is a core bottleneck determining the operational radius, mission duration, and execution efficiency of drones, directly restricting their adaptability to complex scenarios. The accuracy of energy consumption statistics and the scientific nature of energy allocation strategies are key factors affecting drone endurance, flight status control, and mission completion quality. Therefore, achieving high-precision energy consumption statistics and constructing an intelligent energy allocation mechanism have become key research and development directions for drone flight control systems, and are also critical technical challenges that urgently need to be overcome in the field of drone swarm collaborative operations.

[0004] Existing methods for UAV energy consumption statistics and energy allocation mainly include the following: First, energy management methods based on overall power consumption monitoring and threshold alarms. This method is simple to implement and can meet basic power monitoring needs, but it can only obtain macroscopic overall power consumption information. When power consumption anomalies occur, it cannot locate high-energy-consuming modules or the source of the anomaly. Troubleshooting relies heavily on manual point-by-point inspection, which is inefficient and seriously affects the reliability of flight missions. Second, flight phase division and energy consumption estimation methods based on rule engines. This method has clear logic and low computational load, but its flight phase division relies too heavily on threshold settings. In complex maneuvering flight, wind field disturbances, or state transition scenarios, phase misjudgment is prone to occur. At the same time, the energy consumption differences between different flight phases are significant, and phase misjudgment will directly lead to inaccurate estimation of remaining flight time, affecting mission reliability. Third, task allocation and emergency handling methods based on preset rules. In multi-UAV collaborative missions, task allocation or transfer is usually based on simple rules such as task priority, distance, and remaining power. This method is fast in decision-making and simple in engineering implementation, but it does not take into account the status of the drone itself. As a result, after the task is redistributed, the task recipient may be unable to complete the new task due to insufficient energy, resulting in energy waste and low energy utilization efficiency; or it may be unable to respond immediately due to mismatch in flight status, which will lead to task interruption or repeated redistribution, which not only wastes energy, but also greatly reduces the reliability of multi-drone collaborative operations.

[0005] The information disclosed in this background section is intended only to enhance the understanding of the overall background of this application and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention

[0006] The purpose of this application is to provide a method and system for UAV flight control and energy allocation based on energy consumption perception, which can solve the problems raised in the background technology to a certain extent and improve the energy utilization efficiency and mission reliability of UAV flight control system.

[0007] To achieve the above objectives, this application provides the following technical solution:

[0008] In a first aspect, this application provides a method for unmanned aerial vehicle (UAV) flight control and energy allocation based on energy consumption perception, including the following steps:

[0009] Real-time collection of energy consumption data, flight status data, and operational environment data of each UAV; and abnormal energy consumption detection based on the energy consumption data to generate abnormal energy consumption identifiers.

[0010] Based on the flight status data, the flight phase is identified, and the energy consumption baseline for each flight phase is constructed. The energy efficiency coefficient of each UAV is calculated in combination with the operational environment data, and the flight parameters of each UAV are solved with the goal of maximizing the energy efficiency coefficient. At the same time, the flight status of the corresponding UAV is adjusted based on the flight parameters.

[0011] Obtain the remaining tasks for each UAV and decompose the remaining tasks into consecutive task segments; assign flight phases and durations to each task segment, and calculate the total expected energy consumption of the remaining tasks in combination with the corresponding flight parameters and energy consumption baseline;

[0012] An energy consumption map is constructed based on the total expected energy consumption of each drone, and the energy surplus of each drone is calculated.

[0013] If the energy surplus of any drone is lower than the preset energy threshold, or if any drone has an abnormal energy consumption indicator, then the task will be redistributed based on the energy consumption map, and the remaining tasks of each drone will be updated.

[0014] As a preferred embodiment of the UAV flight control and energy allocation method based on energy consumption perception described in this application, the task reassignment step specifically includes:

[0015] If any drone's energy surplus is less than the preset energy threshold, or if there is an abnormal energy consumption indicator, then the corresponding drone will be designated as the mission initiator.

[0016] The task initiator packages the transferable parts of its remaining tasks into one or more task packages, while querying its own energy consumption map and obtaining the energy consumption status information of all neighboring drones.

[0017] Based on the energy consumption status information of all neighboring drones, neighboring drones that are qualified to receive mission packets are selected to obtain candidate drones, and an interruptibility coefficient is set according to the flight stage of each candidate drone.

[0018] The penalty factor is calculated based on the task type label and the expected energy consumption of the task, and the task acceptance rate of each candidate UAV is calculated based on the interruptibility coefficient and the penalty factor.

[0019] As a preferred embodiment of the UAV flight control and energy allocation method based on energy consumption perception described in this application, the task reassignment step further includes:

[0020] Sort all candidate UAVs by task acceptance in descending order, and select the k candidate UAVs with the highest task acceptance as task receivers; where k is a positive integer.

[0021] Send a task package to the task recipient and receive a confirmation of receipt from the task recipient;

[0022] Count the number of received confirmation messages. If the number is greater than 1, select the task recipient with the highest task acceptance rate as the task executor and send a reconfirmation message to the task executor.

[0023] If the number of entries is equal to 1, then the task recipient is directly designated as the task executor, and a reconfirmation instruction is sent to the task executor.

[0024] As a preferred embodiment of the UAV flight control and energy allocation method based on energy consumption perception described in this application, the task reassignment step further includes:

[0025] If the number of confirmed reception instructions is 0, then the candidate drones other than the top k candidate drones with the highest task reception will be regarded as task receivers.

[0026] Send a task packet to the task recipient and recount the number of received confirmation instructions;

[0027] When the task executor receives the reconfirmation receipt instruction, it adds the task waypoint sequence corresponding to the task package to its remaining tasks, while the task initiator removes the task package from its remaining tasks.

[0028] As a preferred embodiment of the UAV flight control and energy allocation method based on energy consumption perception described in this application, the penalty factor is calculated as follows:

[0029] If a candidate drone has an abnormal energy consumption flag, a base penalty value is set according to the abnormal type of the abnormal energy consumption flag.

[0030] The weight of the base penalty value is set according to the task type label, and the base penalty value is multiplied by the weight of the base penalty value to obtain the abnormal penalty item; if the candidate drone does not have an abnormal energy consumption indicator, the abnormal penalty item is set to 0;

[0031] The ratio of the projected energy consumption of the task to the remaining energy of the candidate drone is used to obtain the energy matching degree. If the energy matching degree is greater than a preset matching degree threshold, the difference between the energy matching degree and the matching degree threshold is calculated, and an energy matching penalty term is calculated based on the difference. The energy matching penalty term is proportional to the difference.

[0032] If the energy matching degree is less than or equal to the matching degree threshold, then the energy matching penalty term is 0;

[0033] The sum of the anomaly penalty term and the energy matching penalty term is calculated to obtain the penalty factor.

[0034] As a preferred embodiment of the UAV flight control and energy allocation method based on energy consumption perception described in this application, the energy consumption baseline for each flight phase is constructed as follows:

[0035] Extract energy consumption data from historical data of M complete flight missions, and remove energy consumption data with abnormal energy consumption indicators to obtain reference data; where M is a positive integer;

[0036] The reference data is classified according to flight phase to obtain the reference dataset corresponding to each flight phase;

[0037] For each flight phase, the average instantaneous power consumption of each functional module is calculated, and the arithmetic mean of the average instantaneous power consumption of all functional modules is calculated to obtain the average power consumption of the corresponding flight phase.

[0038] Using the average power consumption as the dependent variable and the vacuum speed, air pressure altitude, angle of attack, pitch angle and engine throttle opening in the reference dataset as independent variables, a linear regression model is established, and the regression coefficients are fitted to obtain the energy consumption baseline for the corresponding flight phase.

[0039] As a preferred embodiment of the UAV flight control and energy allocation method based on energy consumption perception described in this application, the flight parameters are calculated as follows: using the energy efficiency coefficient corresponding to the current flight phase as the objective function, and vacuum speed, angle of attack, and heading as optimization variables, the flight parameters that maximize the energy efficiency coefficient are solved under preset constraints; the steps of solving the flight parameters that maximize the energy efficiency coefficient specifically include:

[0040] Using the current vacuum speed, angle of attack, and heading as the initial points for iteration, iterative optimization is performed.

[0041] In each iteration, the gradient of the objective function with respect to vacuum velocity, angle of attack, and heading is calculated at the current point using numerical difference, thus obtaining the first-order change of the objective function at the current point;

[0042] Based on the first-order transformation, a second-order Taylor expansion is performed on the objective function at the current point, and the constraints are linearized to the first order.

[0043] As a preferred embodiment of the UAV flight control and energy allocation method based on energy consumption perception described in this application, the step of solving the flight parameters that maximize the energy efficiency coefficient further includes:

[0044] Using the second-order Taylor expansion as the objective function and the first-order linearization as the constraint, a quadratic programming subproblem is constructed; the quadratic programming subproblem is solved to obtain the search direction, and a line search is performed along the search direction to obtain the optimal step size;

[0045] The current point is updated based on the optimal step size to obtain new flight parameters, which serve as the starting point for the next iteration. The change in the objective function between two adjacent iterations is also calculated.

[0046] If the change value is less than the preset convergence threshold, or the number of iterations reaches the preset maximum number of iterations, then the flight parameters obtained in the last iteration will be used as the final flight parameters.

[0047] As a preferred embodiment of the UAV flight control and energy allocation method based on energy consumption perception described in this application, the specific method for decomposing the remaining tasks into continuous task segments is as follows:

[0048] Using the current real-time location of the UAV as the starting path point, and taking the first waypoint in the remaining waypoint list as the first target point, two adjacent path points are defined as a task segment in sequence;

[0049] Traverse all remaining waypoints to obtain a sequence of task segments; in the sequence of task segments, calculate the geometric parameters of the corresponding task segment based on the three-dimensional coordinates of the path point corresponding to any task segment, including segment length, altitude change, horizontal projection length and segment orientation angle;

[0050] Based on the geometric parameters and mission type labels, flight phases are assigned to each mission segment;

[0051] The performance parameters of the corresponding UAV, including the maximum rate of climb and the maximum rate of descent, are obtained from the UAV's flight control calculator. Based on the performance parameters and the flight phase, the duration of each mission segment is calculated.

[0052] Secondly, this application provides an unmanned aerial vehicle (UAV) flight control and energy allocation system based on energy consumption perception, including a data acquisition module, an energy consumption analysis module, a mission planning module, and an energy allocation module;

[0053] The data acquisition module is used to collect energy consumption data, flight status data and operating environment data of each UAV in real time, and to detect abnormal energy consumption based on the energy consumption data and generate abnormal energy consumption identifiers.

[0054] The energy consumption analysis module identifies the flight phase based on the flight status data and constructs the energy consumption baseline for each flight phase. The energy consumption analysis module also calculates the energy efficiency coefficient of each UAV based on the flight phase and operating environment data, calculates the flight parameters of each UAV based on the energy efficiency coefficient, and controls the corresponding UAV to adjust its flight status based on the flight parameters.

[0055] The task planning module is used to obtain the remaining tasks of each UAV and decompose the remaining tasks into continuous task segments; allocate flight phases and durations to each task segment, and calculate the total expected energy consumption of the remaining tasks in combination with the corresponding flight parameters and energy consumption baseline; the task planning module also constructs an energy consumption map based on the total expected energy consumption of each UAV and calculates the energy surplus of each UAV.

[0056] The energy allocation module is used to redistribute tasks based on the energy consumption map and update the remaining tasks of each drone when the energy surplus of any drone is lower than the preset energy threshold or when any drone has an abnormal energy consumption indicator.

[0057] Compared with the prior art, the beneficial effects achieved by the present invention are as follows:

[0058] Based on the collected energy consumption data, abnormal energy consumption is detected for each UAV, and abnormal energy consumption identifiers are obtained, realizing the identification and location of energy consumption anomalies. Flight status data is used to identify the flight phase of each UAV and construct energy consumption baselines for each flight phase. The energy efficiency coefficient of each UAV is calculated based on the flight phase and operational environment data, and flight parameters are calculated based on the energy efficiency coefficient. Simultaneously, the flight status of the corresponding UAV is adjusted based on the flight parameters, reducing single-aircraft cruise energy consumption and extending mission endurance. The remaining tasks are decomposed into continuous task segments, and flight phases and durations are assigned to each task segment. The total expected energy consumption is calculated by combining flight parameters and energy consumption baselines, and an energy consumption map is constructed based on the total expected energy consumption, improving the energy utilization efficiency and endurance of the UAVs. The energy surplus of each UAV is calculated. When the energy surplus of any UAV is lower than the energy threshold or an abnormal energy consumption identifier exists, tasks are redistributed based on the energy consumption map, and the remaining tasks are updated, effectively balancing the energy load of the swarm, avoiding mission interruptions caused by local energy depletion, and enhancing the overall mission reliability and robustness of the swarm system. Attached Figure Description

[0059] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0060] Figure 1 A flowchart of the UAV flight control and energy allocation method based on energy consumption perception provided in this application;

[0061] Figure 2 The architecture diagram of the UAV flight control and energy allocation system based on energy consumption perception provided in this application. Detailed Implementation

[0062] The technical solution of this application will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the embodiments and specific features in the embodiments are detailed descriptions of the technical solution of this application, rather than limitations thereof. In the absence of conflict, the embodiments and technical features in the embodiments can be combined with each other.

[0063] Example 1

[0064] like Figure 1 As shown, this embodiment introduces a method for UAV flight control and energy allocation based on energy consumption perception, including the following steps:

[0065] Real-time collection of energy consumption data, flight status data, and operational environment data of each UAV; abnormal energy consumption detection based on the energy consumption data; generation of abnormal energy consumption identifiers; and storage of each collection of energy consumption data, flight status data, operational environment data, and corresponding abnormal energy consumption identifiers as historical data.

[0066] Specifically, the energy consumption data is collected as follows: The power monitoring units built into each functional module collect the instantaneous power consumption of the corresponding functional module at a preset fixed frequency, such as 200Hz. This includes the power consumption of the flight core, the main inertial measurement unit, the backup inertial measurement unit, the operational environment data system, the global navigation satellite system receiver, the communication link, and the mission payload. Simultaneously, the bus load rate and data retransmission count are recorded. The flight core power consumption includes at least the total power consumption of the flight control computer's processor and memory. The flight status data is collected as follows: the main inertial measurement unit collects three-axis angular rates, three-axis accelerations, pitch angles, and roll angles; the engine control system collects engine throttle opening; and the landing gear sensors collect landing gear status. The operational environment data is collected as follows: the operational environment data system outputs vacuum speed, angle of attack, sideslip angle, and barometric altitude in real time; and the global navigation satellite system receiver outputs position and velocity in real time.

[0067] The specific method for detecting abnormal energy consumption and generating abnormal energy consumption identifiers based on the energy consumption data is as follows: Calculate the absolute value of the difference between the power consumption of the primary inertial measurement unit (IMU) and the power consumption of the backup IMU to obtain the IMU power consumption deviation; if the IMU power consumption deviation is greater than a preset deviation threshold (e.g., 0.15W) for a preset continuous time (e.g., within 1 second), the IMU is determined to have abnormal energy consumption, and an abnormal energy consumption identifier is generated; respectively, the mean and standard deviation of the power consumption of the operational environment data system, the receiver power consumption of the global navigation satellite system, the communication link power consumption, and the mission payload power consumption in the most recent N flight missions are statistically analyzed; where N is a positive integer (e.g., N=10); if any functional module among the operational environment data system power consumption, the receiver power consumption of the global navigation satellite system, the communication link power consumption, and the mission payload power consumption is greater than the sum of the corresponding mean and three times the standard deviation, the corresponding functional module is determined to have abnormal energy consumption, and an abnormal energy consumption identifier is generated; if the bus load rate is greater than a preset load rate threshold for a continuous period of time, or the number of data retransmissions increases and the communication link power consumption increases, it is marked as abnormal communication energy consumption, and an abnormal energy consumption identifier is generated.

[0068] The flight phase is identified based on flight status data, and an energy consumption baseline for each flight phase is constructed. The energy efficiency coefficient of each UAV is calculated based on the flight phase, energy consumption data, and operating environment data. The flight parameters of each UAV are calculated based on the energy efficiency coefficient, and the flight status of the corresponding UAV is adjusted based on the flight parameters.

[0069] The specific method for identifying flight status is as follows: A flight feature vector is constructed based on the flight status data, and the flight feature vector within a preset sliding time window is obtained. The flight feature vector consists of barometric altitude and rate of change, vacuum speed and rate of change, pitch angle, roll angle, engine throttle opening, and landing gear status. An LSTM model, i.e., a Long Short-Term Memory network model, is constructed and trained. Using the flight feature vector within the sliding time window as input, the LSTM model outputs the flight stage, and the flight stage is associated with the corresponding energy consumption data. In this embodiment, the LSTM model consists of one input layer, two LSTM layers, three Dropout layers, one fully connected layer, and one output layer.

[0070] For example, the sliding time window is the most recent 10 seconds, that is, the sequence of flight feature vectors of the most recent 50 consecutive moments is used as the input vector; the dimension of the input vector is 50×8; after the input vector is input to the input layer, it passes through the first LSTM layer and the first Dropout layer to obtain the first output vector; the first LSTM layer is set to 64 memory units, and the dimension of the first output vector is 50×64; the first output vector passes through the second LSTM layer and the second Dropout layer to obtain the second output vector; the second LSTM layer is set to 64 memory units, and the dimension of the second output vector is 64; both the first LSTM layer and the second LSTM layer use the tanh function as the activation function and the sigmoid function as the gating function; the second output vector passes through the fully connected layer and the third Dropout layer. The dropout layer yields the third output vector; the fully connected layer has 16 neurons and uses the ReLU function as the activation function; the dropout rate of the first, second, and third dropout layers is all set to 0.3; the dimension of the third output vector is 16; after inputting the third output vector into the output layer, the final output vector is obtained; the output layer has 8 neurons and uses the sigmoid function as the activation function; the dimension of the final output vector is 8; the training method of the LSTM model is as follows: historical data from 50 complete flight missions are extracted, containing a total flight time of approximately 500 hours, and three different engineers independently manually annotate the flight stages corresponding to the historical data, and the final annotation results of each historical data segment are determined through a voting mechanism; the annotation set S of the flight stages is... The historical data was sliced ​​into 10-second windows with a 1-second step size, resulting in 1.8 million samples, each of which was... ;in Let be the historical data of the window corresponding to the i-th sample, where i is a positive integer, and ; The final annotation result of the historical data at the center moment of the window corresponding to the i-th sample is given. The 1.8 million samples are divided into training set, validation set and test set in a ratio of 7:2:1. During the division, it is ensured that the historical data of the same flight mission will not appear in the training set and validation set at the same time to avoid data leakage. The sparse classification cross-entropy is used as the loss function, and the Adam optimizer is used to train the LSTM model. The initial learning rate is set to 0.001. The decay strategy during the training process is that if the validation set loss does not decrease for 3 consecutive rounds, the learning rate is multiplied by 0.5. During the training process, the batch size is set to 54 and the number of training rounds is set to 100.

[0071] The energy consumption baseline for each flight phase is constructed as follows: Energy consumption data is extracted from historical data of M complete flight missions, and data with abnormal energy consumption indicators is removed to obtain reference data; where M is a positive integer, such as M=50; the reference data is categorized by flight phase, and reference data with the same flight phase are selected to obtain the reference dataset corresponding to the flight phase; for each flight phase, the mean instantaneous power consumption of each functional module is calculated, and the arithmetic mean of the mean instantaneous power consumption of all functional modules is calculated to obtain the average power consumption of the corresponding flight phase; using the average power consumption as the dependent variable, and using the energy-related operational environment data and flight state parameters from the historical data corresponding to the reference dataset as independent variables, a linear regression model is established, and the regression coefficients are fitted using least squares to obtain the energy consumption baseline for the corresponding flight phase; the energy-related operational environment data and flight state parameters include: vacuum speed, barometric altitude, angle of attack, pitch angle, and engine throttle opening; the linear regression model is specifically as follows: Where P represents the expected energy consumption for the corresponding flight phase. Here are the regression coefficients, where V is the vacuum velocity, h is the atmospheric pressure altitude, and... For the angle of attack, For pitch angle, This refers to the engine throttle opening.

[0072] The energy efficiency coefficient is a comprehensive indicator that measures the energy utilization efficiency of a drone in its current flight state, and it is calculated as follows:

[0073] The local speed of sound is obtained, and the ratio of the speed of vacuum to the local speed of sound is calculated to obtain the corresponding Mach number. Using the angle of attack and Mach number as indices, the corresponding lift coefficient and drag coefficient are retrieved from the aerodynamic coefficient table pre-stored in the UAV's flight control computer, and the ratio of the lift coefficient to the drag coefficient is calculated to obtain the lift-drag ratio. The aerodynamic coefficient table covers the entire flight envelope of the UAV and is obtained in advance through wind tunnel experiments or computational fluid dynamics simulations. The aerodynamic coefficient table is stored in the UAV's flight control computer in the form of a two-dimensional table, with rows corresponding to the angle of attack and columns corresponding to the Mach number.

[0074] In this embodiment, the local speed of sound is taken as 340 m / s; during the UAV design finalization stage, the lift coefficient and drag coefficient at different angles of attack and Mach numbers are measured through wind tunnel experiments, and the measured experimental data are smoothed to form a two-dimensional table; wherein, the angle of attack range of the wind tunnel test is -10° to 10° with a step size of 0.1°, and the Mach number range is 0 to 2.0 with a step size of 0.01;

[0075] The sum of the instantaneous power consumption of each functional module is calculated to obtain the instantaneous total power consumption. The ratio of vacuum speed to the instantaneous total power consumption is calculated to obtain the range per unit energy consumption. Weighting coefficients are set according to the flight phase, and the lift-to-drag ratio and range per unit energy consumption are weighted and summed according to the weighting coefficients to obtain the corresponding energy efficiency coefficient.

[0076] In this embodiment, the weighting coefficients are set as follows: initial weights are set for each flight phase as shown in Table 1.

[0077] Table 1 Initial weights for each flight phase

[0078]

[0079] Wherein, the first weight is the weight coefficient of lift-to-drag ratio, and the second weight is the weight coefficient of range per unit energy consumption; to avoid sudden changes in flight parameters due to abrupt changes in weight coefficients during flight phase switching, the initial weights are subjected to first-order low-pass filtering smoothing to obtain the corresponding weight coefficients; the first-order low-pass filtering smoothing specifically involves: obtaining the weight value of the previous data acquisition cycle, and weighting and summing the initial weight value with the weight value of the previous data acquisition cycle to obtain the corresponding weight coefficients; wherein, the weight coefficient of the initial weight value is a smoothing coefficient, such as 0.1, and the weight coefficient of the weight value of the previous data acquisition cycle is the difference between 1 and the smoothing coefficient;

[0080] The flight parameters are calculated as follows: using the energy efficiency coefficient corresponding to the current flight phase as the objective function, and vacuum speed, angle of attack, and heading as optimization variables, under preset constraints, a sequential quadratic programming algorithm is used to solve for the flight parameters that maximize the energy efficiency coefficient; the flight parameters include vacuum speed, angle of attack, and heading.

[0081] The constraints include: vacuum velocity constraint is... Among them, V min V is the minimum vacuum velocity. max The maximum vacuum speed is specified in the UAV flight manual, for example, V. min =15m / s, V max =100m / s; Angle of attack constraint is The α min For the minimum angle of attack, α max The maximum angle of attack is obtained in advance through wind tunnel experiments and flight tests, for example, α min =-5°, α max =15°; heading constraint is Where Φ is the optimized heading, Φ p For the course of the flight mission, The maximum heading difference is set based on the UAV's own flight performance, such as minimum turning radius, maximum roll rate, and the mission's heading accuracy requirements; for example, The instantaneous total power consumption is 10°; the instantaneous total power consumption is less than or equal to the preset maximum instantaneous power consumption; the maximum instantaneous power consumption is derived from the maximum allowable output power of the power system specified in the UAV design manual, and in this embodiment, the maximum instantaneous power consumption is 500W;

[0082] The specific method for solving the flight parameters that maximize the energy efficiency coefficient using a sequential quadratic programming algorithm is as follows: An iterative optimization process is performed using the current vacuum speed, angle of attack, and heading as the initial iteration points; in each iteration, the gradient of the objective function with respect to vacuum speed, angle of attack, and heading is calculated at the current point using numerical difference, obtaining the first-order change of the objective function at the current point; based on the first-order change, a second-order Taylor expansion is performed on the objective function at the current point, and the constraints are linearized to the first order; using the second-order Taylor expansion as the objective function and the first-order linearization as the constraints, a... The quadratic programming subproblem is solved to obtain the search direction. A line search is performed along the search direction to obtain the optimal step size, which ensures that the value of the objective function is sufficiently improved while satisfying the constraints. The current point is updated based on the optimal step size to obtain new flight parameters, which serve as the starting point for the next iteration. The change in the objective function between two adjacent iterations is calculated. If the change is less than a preset convergence threshold, such as 0.001, or the number of iterations reaches a preset maximum number of iterations, such as 20, then the flight parameters obtained in the last iteration are used as the final flight parameters.

[0083] In this embodiment, the specific implementation of solving the quadratic programming subproblem to obtain the search direction is as follows: the quadratic programming subproblem is transformed into a standard quadratic form; the standard quadratic form includes: a vector of the search direction to be solved, a linear coefficient matrix composed of gradients, a Hessian matrix composed of second-order Taylor expansion, and a coefficient matrix composed of first-order linearization; the constraints that are effective at the current point are identified, and the constraints that are not effective are ignored; the standard quadratic form is solved in the subspace defined by the effective constraints to obtain candidate solutions for the search direction; whether the candidate solutions satisfy all the ignored constraints is checked. If not, the effective constraints are adjusted and the solution is re-solved until a vector of the search direction that simultaneously satisfies all constraints and makes the objective function optimal is obtained;

[0084] Optionally, the specific method of the line search is as follows: set an initial step size, such as 1; and check whether the next point satisfies all constraints based on the initial step size; if not, multiply the initial step size by a preset reduction factor to obtain a new step size, and check whether the next point satisfies all constraints based on the new step size; if not, continue to multiply the new step size by the reduction factor until the next point corresponding to the new step size satisfies all constraints, thus obtaining the optimal step size; the reduction factor is less than 1, used to reduce the step size when the step size does not satisfy the constraints to ensure the constraint satisfaction, and its setting is based on a trade-off between search efficiency and constraint satisfaction accuracy, such as 0.5.

[0085] The specific method for controlling the UAV to adjust its flight state based on the flight parameters is as follows: the flight parameters are input as commands to the inner loop controller of the flight control computer system. The inner loop controller generates control commands for each actuator based on the current flight stage and flight parameters. In this embodiment, the inner loop controller includes a speed controller, a pitch controller, and a heading controller. The speed controller calculates the throttle adjustment amount by calculating the deviation between the vacuum speed and the current vacuum speed, and outputs the throttle adjustment amount to the engine control system. The pitch controller calculates the elevator deflection command by calculating the deviation between the angle of attack and the current angle of attack, and combines it with the current pitch rate, and outputs the elevator deflection command to the servo control system. The heading controller calculates the aileron and rudder deflection commands by calculating the deviation between the heading and the current heading, and combines it with the current roll rate, and outputs the aileron and rudder deflection commands to the corresponding servo control systems.

[0086] Obtain the remaining tasks for each human-machine interface and decompose them into consecutive task segments; assign flight phases and durations to each task segment, and calculate the total expected energy consumption of the remaining tasks based on the corresponding flight parameters and energy consumption baseline; the remaining tasks include at least a list of remaining waypoints. The current real-time position, airspeed, and heading of the UAV; where n is the total number of remaining waypoints; each waypoint W j Each includes three-dimensional coordinates, a passing speed requirement, and a corresponding task type label, where j is a positive integer and j∈[1,n]; the task type label ∈{cruising, inspection and monitoring, fixed-point operation, hovering and waiting, landing}; the passing speed requirement is the speed specified by the task.

[0087] The specific method for decomposing the remaining tasks into continuous task segments is as follows: taking the current real-time position of the UAV as the starting path point P0, and taking the first waypoint W1 in the remaining waypoint list as the first target point, and sequentially decomposing adjacent pathpoints (P0, W1 ... j-1 ,P j ) is defined as a task segment R j ;where P j-1 The (j-1)th waypoint corresponds to the (j-1)th waypoint W. j-1 The three-dimensional coordinates of P j For the j-th waypoint, the corresponding waypoint W j The three-dimensional coordinates of the UAV are given, and when j is 1, P0 is the current real-time position of the UAV; all remaining waypoints are traversed to obtain a sequence of task segments consisting of n task segments; in the sequence of task segments, according to the path point (P) corresponding to any task segment... j-1 ,P j The three-dimensional coordinates of the given coordinates are used to calculate the geometric parameters of the corresponding task segment, including segment length, height variation, horizontal projection length, and segment orientation angle. The segment length is the straight-line distance between two corresponding path points; the height variation is the height difference between two corresponding path points; the horizontal projection length is the projected distance between two corresponding path points in the horizontal plane; and the segment orientation angle is the distance from path point P. j-1 Point to path point P j The azimuth angle;

[0088] For example, if path point P j-1 The three-dimensional coordinates are (3000, 2500, 500), and the path point P j If the three-dimensional coordinates are (3500, 3000, 300), then the segment length is... Meters, with an altitude change of -200 meters, and a horizontal projection length of... meters, segment direction angle is .

[0089] Based on the geometric parameters and mission type labels, flight phases are assigned to each mission segment: if any mission segment R j Corresponding waypoint W j If the task type label is inspection and monitoring, fixed-point operation, or hovering and waiting, then the task segment R will be... jThe flight phase is set as a special operation; if W j If the task type label is "landing", then R j The flight phase is set to landing; when W j When the task type label is "cruising through", if the altitude change of the corresponding task segment exceeds a preset first altitude change threshold, then R... j The flight phase is set to climb; if the altitude change of the corresponding mission segment is less than the preset second altitude change threshold, then R... j The flight phase is set to descent; if the absolute value of the altitude change in the corresponding mission segment is less than the preset second altitude change threshold, then R... j The flight phase is set to cruise; wherein, both the first altitude change threshold and the third altitude change threshold are greater than 0, the first altitude change threshold is greater than the third altitude change threshold, and the second altitude change threshold is less than 0; when R j Corresponding waypoint W j The height is less than a preset height threshold, such as 100 meters, and W j When the last waypoint is reached, if the altitude change is less than the preset fourth altitude change threshold, then R will be... j Set the flight phase to approach, otherwise set R. j The flight phase is set to landing; the altitude threshold, the first altitude change threshold, the second altitude change threshold, the third altitude change threshold, and the fourth altitude change threshold are all set according to the engineering definition of the UAV flight phase division; for example, the altitude threshold is 100 meters, the first altitude change threshold is 50 meters, the second altitude change threshold is -50 meters, the third altitude change threshold is 20 meters, and the fourth altitude change threshold is -20 meters.

[0090] Obtain the corresponding performance parameters of the UAV from the UAV's flight control calculator, including maximum rate of climb, maximum rate of descent, cruise speed range, and maximum turn rate; if mission segment R... j If the flight phase is climb, then the duration of the corresponding mission segment is calculated based on the corresponding altitude change and the maximum climb rate; the duration of the climb is the ratio of the altitude change to the maximum climb rate; if mission segment R... j If the flight phase is descent, then the duration of the corresponding mission segment is calculated based on the corresponding altitude change and the maximum descent rate; the duration of descent is the ratio of the absolute value of the altitude change to the maximum descent rate; if mission segment R j If the flight phase is cruise, then the duration of the corresponding mission segment is calculated based on the corresponding segment length and the required speed at the pass point; the duration of cruise is the ratio of the segment length to the required speed at the pass point; if mission segment R jIf the flight phase is a special operation, the duration is set according to the task type label; for example, the duration for a task type label of inspection and monitoring is 120 seconds, the duration for a task type label of fixed-point operation is 60 seconds, and the duration for a task type label of circling and waiting is set according to the required number of laps and speed; if the task segment R j If the flight phase is approach or landing, the duration is set according to the standard approach procedure, such as 180 seconds.

[0091] The total expected energy consumption is calculated as follows: based on task segment R j The task type is used to call the corresponding energy consumption baseline and calculate the task segment R. j The instantaneous expected energy consumption is calculated; the instantaneous expected energy consumption is multiplied by the corresponding duration to obtain the task segment R. j The expected energy consumption is calculated; the sum of the expected energy consumption of all task segments is calculated, and the sum is added to the preset safety margin to obtain the total expected energy consumption of the remaining tasks of the corresponding UAV; in this embodiment, the safety margin is taken as 1.5 times the root mean square of the difference between the actual total energy consumption and the total expected energy consumption in all historical tasks.

[0092] An energy consumption map is constructed based on the total expected energy consumption of each drone, and the energy surplus of each drone is calculated.

[0093] Specifically, each UAV periodically broadcasts its own energy consumption status information through an inter-UAV communication link; for example, broadcasting its own energy consumption status information once every 1 second. The energy consumption status information includes: UAV identifier, remaining energy, total expected energy consumption, abnormal energy consumption identifier, flight stage, instantaneous total power, current real-time location, and timestamp; the remaining energy is obtained from the UAV's battery management system; after receiving the broadcast, each UAV stores the energy consumption status information of the corresponding UAV in the received broadcast in the UAV's flight control computer to form an energy consumption map, and regards the corresponding UAV in the received broadcast as a neighboring UAV, that is, all UAVs within the communication range of the UAV are neighboring UAVs; the energy consumption map is stored in the form of a two-dimensional table, including the energy consumption status information of all neighboring UAVs of the UAV; for example, Table 2 shows the energy consumption map formed by the broadcast received by UAV004 at 14:23:05 on a certain day, where UAV001, UAV002, and UAV003 are all neighboring UAVs of UAV004; in Table 2, 1 for abnormal energy consumption identifier indicates the presence of an abnormal energy consumption identifier, and 0 indicates the absence of an abnormal energy consumption identifier.

[0094] The energy consumption map is updated over time. If the drone does not receive a broadcast from any neighboring drone within a preset time, such as 5 seconds, it is determined that the communication between the neighboring drone and the drone is interrupted, and the energy consumption status information corresponding to the neighboring drone is removed from the energy consumption map corresponding to the drone. For example, if UAV004 does not receive a broadcast from UAV001 for 5 consecutive seconds, it is determined that the communication between UAV001 and UAV004 is interrupted, and the energy consumption status information corresponding to UAV001 is removed from the energy consumption map of UAV004.

[0095] Table 2 Example of an energy consumption map

[0096]

[0097] The energy surplus of each UAV is calculated as follows: the difference between the remaining energy of each UAV and the corresponding total expected energy consumption is calculated, that is, the remaining energy minus the corresponding total expected energy consumption, to obtain the energy surplus of the corresponding UAV.

[0098] If the energy surplus of any drone is lower than the preset energy threshold, or if any drone has an abnormal energy consumption indicator, then the task will be redistributed based on the energy consumption map, and the remaining tasks of each drone will be updated.

[0099] Specifically, the system monitors the energy surplus and abnormal energy consumption indicators of each drone in real time. If the energy surplus of any drone is less than the energy threshold, it is determined that the corresponding drone has insufficient energy reserves, posing a risk of mission interruption. The energy threshold is proportional to the total expected energy consumption of the corresponding drone, for example, the energy threshold is 10% of the total expected energy consumption of the corresponding drone. If any drone has an abnormal energy consumption indicator, it is determined that the corresponding drone has an abnormal energy consumption module, which may affect the reliability or energy efficiency of the mission, and the mission should be transferred first to reduce the risk of abnormality. If the energy surplus of any drone is less than the energy threshold, or has an abnormal energy consumption indicator, the corresponding drone is designated as the mission initiator, and mission reassignment is triggered. The specific method of mission reassignment is as follows:

[0100] The task initiator packages the transferable portions of its remaining tasks into one or more task packages, while simultaneously querying its own energy consumption map and obtaining the energy consumption status information of all neighboring drones. Each task package includes: a task package identifier, a task waypoint sequence, and the estimated energy consumption of the task. The transferable portion refers to tasks from the task initiator's remaining tasks that can be divided and are suitable for execution by neighboring drones, and must meet the following conditions: continuous waypoints, not containing currently executed task segments, not containing task types dependent on a specific platform, and the transfer will not disrupt the original task's time window constraints. In this embodiment, divisible tasks include: cruise passage, area inspection and monitoring, circling and waiting, etc., which do not depend on a specific platform or do not require continuous execution. Indivisible tasks include fixed-point operations, landing, and special operations. The estimated energy consumption of the task is the sum of the expected energy consumption and safety margin of all task segments contained in the task waypoint sequence.

[0101] Based on the current energy consumption status information of all neighboring drones, candidate drones are selected based on their eligibility to receive the mission package. Specifically, the selection method for candidate drones is as follows: if a neighboring drone is in an uninterrupted flight phase, it is excluded; the uninterrupted flight phase includes landing, approach, takeoff, taxiing, and fixed-point operations or locking during special operations; if a neighboring drone's energy surplus is less than the mission package's expected energy consumption, it is excluded; if a neighboring drone has an abnormal energy consumption indicator, it is excluded.

[0102] An interruptibility coefficient is set according to the flight phase of each candidate UAV; the value range of the interruptibility coefficient is [0,1], and the setting method is as follows: the interruptibility coefficient of the flight phase that cannot be interrupted is set to 0, and the interruptibility coefficient of the interruption phase is set according to the difficulty of interrupting the task; the interruption phase includes climb, cruise, descent, and other sub-phases of special operations except for fixed-point operation and locking; in this embodiment, the flight state of the cruise phase is stable and easy to interrupt, so the interruptibility coefficient of cruise is set to 0.1; the climb and descent phases have altitude changes and the interruption cost is moderate, so the interruptibility coefficient of climb and descent is set to 0.3; the special operation phase has strong task criticality and drastic state changes, so the interruption cost is the highest, so the interruptibility coefficient of other sub-phases of special operations except for fixed-point operation and locking is set to 0.5;

[0103] A penalty factor is calculated based on the task type label and the expected energy consumption of the task. The calculation method for the penalty factor is as follows: if a candidate UAV has an abnormal energy consumption indicator, a basic penalty value is set according to the abnormal type corresponding to the abnormal energy consumption indicator, and a weight is set according to the task type label. For example, if the task type label is inspection and monitoring, the task duration is long and easily replaceable, so the weight is set to 0.5; if the task type label is fixed-point operation, the task is critical, so the weight is set to 1.2; if the task type label is landing, the task cannot be interrupted and has high safety requirements, so the weight is set to 1.5. The basic penalty value is multiplied by its weight to obtain the abnormal penalty item. If the candidate UAV has an abnormal energy consumption indicator, a basic penalty value is set according to the expected energy consumption of the task. If no abnormal energy consumption indicator is found in the human-machine interface, the abnormal penalty item is set to 0. The ratio of the expected energy consumption of the task to the remaining energy of the candidate drone is calculated to obtain the energy matching degree. If the energy matching degree is greater than a preset matching degree threshold, such as 30%, the difference between the energy matching degree and the matching degree threshold is calculated, i.e., the energy matching degree minus the matching degree threshold, and an energy matching penalty item is calculated based on the difference. The energy matching penalty item is proportional to the difference; the larger the difference, the larger the value of the energy matching penalty item. If the energy matching degree is less than or equal to the matching degree threshold, the energy matching penalty item is 0. The sum of the abnormal penalty item and the energy matching penalty item is calculated to obtain the penalty factor.

[0104] Specifically, if the inertial measurement unit (IMU) has an abnormal power consumption, the base penalty value is set to the first base penalty value; if the communication link has an abnormal power consumption, the base penalty value is set to the second base penalty value; otherwise, the base penalty value is set to the third base penalty value. The first base penalty value is greater than the second base penalty value, and the second base penalty value is greater than the third base penalty value. The first, second, and third base penalty values ​​are all set according to the severity of the abnormality type. For example, if the IMU abnormality affects flight safety, the impact is the most severe, so the first base penalty value is set to 0.3; if the communication link abnormality affects coordination and data transmission, the impact is secondary, so the second base penalty value is set to 0.2; and if other abnormalities, such as sensor performance degradation, have a relatively small impact, the third base penalty value is set to 0.1.

[0105] The task acceptance score of each candidate UAV is calculated based on the interruptibility coefficient and the penalty factor. The task acceptance score is directly proportional to the energy surplus of the candidate UAV, directly proportional to the interruptibility coefficient, inversely proportional to the expected energy consumption of the task, and inversely proportional to the penalty factor. In this embodiment, the product of the energy surplus and the interruptibility coefficient is taken as the first product, the sum of 1 and the penalty factor is calculated, and the product of the sum and the expected energy consumption of the task is taken as the second product. The ratio of the first product to the second product is calculated to obtain the task acceptance score.

[0106] The task acceptance scores of all candidate UAVs corresponding to the task initiator are sorted in descending order to obtain a task acceptance score sequence. The top k candidate UAVs with the highest task acceptance scores are designated as task receivers; where k is a positive integer, such as k=2. A task packet is sent to the task receiver, and a confirmation acceptance command is received from the task receiver. The number of received confirmation acceptance commands is counted. If the number is greater than 1, the task receiver with the highest task acceptance score is designated as the task executor, and a reconfirmation acceptance command is sent to the task executor. If the number is equal to 1, the task receiver is directly designated as the task executor, and a reconfirmation acceptance command is sent to the task executor. If the number is 0, all candidate UAVs in the task acceptance score sequence other than the top k candidate UAVs with the highest task acceptance scores are designated as task receivers. A task packet is sent to the task receiver, and the number of received confirmation acceptance commands is counted again. When the task executor receives the reconfirmation acceptance command, the task waypoint sequence corresponding to the task packet is added to its remaining tasks, while the task initiator removes the task packet from its remaining tasks.

[0107] Example 2

[0108] This embodiment is the second embodiment of this application; it is based on the same inventive concept as Embodiment 1, and refers to... Figure 2 This embodiment introduces an unmanned aerial vehicle (UAV) flight control and energy allocation system based on energy consumption perception, including a data acquisition module, an energy consumption analysis module, a mission planning module, and an energy allocation module.

[0109] The data acquisition module is used to collect energy consumption data, flight status data and operating environment data of each UAV in real time, and to detect abnormal energy consumption based on the energy consumption data, generate abnormal energy consumption identifiers, and store the energy consumption data, flight status data and operating environment data collected each time and the corresponding abnormal energy consumption identifiers as historical data.

[0110] The energy consumption analysis module identifies the flight phase based on the flight status data and constructs the energy consumption baseline for each flight phase. The energy consumption analysis module also calculates the energy efficiency coefficient of each UAV based on the flight phase and operating environment data, and solves the flight parameters of each UAV with the goal of maximizing the energy efficiency coefficient. At the same time, it controls the corresponding UAV to adjust its flight status based on the flight parameters.

[0111] The task planning module is used to obtain the remaining tasks of each UAV and decompose the remaining tasks into continuous task segments; allocate flight phases and durations to each task segment, and calculate the total expected energy consumption of the remaining tasks in combination with the corresponding flight parameters and energy consumption baseline; the task planning module also constructs an energy consumption map based on the total expected energy consumption of each UAV and calculates the energy surplus of each UAV; the energy consumption map is stored in the form of a two-dimensional table, including the energy consumption status information of all neighboring UAVs of the UAV.

[0112] The energy allocation module is used to redistribute tasks based on the energy consumption map and update the remaining tasks of each drone when the energy surplus of any drone is lower than the preset energy threshold or when any drone has an abnormal energy consumption indicator.

[0113] The specific functions of each step described above are explained in the relevant content of the UAV flight control and energy allocation method based on energy consumption perception described in Example 1, and will not be repeated here.

[0114] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0115] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of protection of this application, and these forms are all within the protection scope of this application.

Claims

1. A method for UAV flight control and energy allocation based on energy consumption perception, characterized in that, Includes the following steps: Real-time collection of energy consumption data, flight status data, and operational environment data of each UAV; and abnormal energy consumption detection based on the energy consumption data to generate abnormal energy consumption identifiers. Based on the flight status data, the flight phase is identified, and the energy consumption baseline for each flight phase is constructed. The energy efficiency coefficient of each UAV is calculated in combination with the operational environment data, and the flight parameters of each UAV are solved with the goal of maximizing the energy efficiency coefficient. At the same time, the flight status of the corresponding UAV is adjusted based on the flight parameters. Obtain the remaining tasks for each UAV and decompose the remaining tasks into consecutive task segments; assign flight phases and durations to each task segment, and calculate the total expected energy consumption of the remaining tasks in combination with the corresponding flight parameters and energy consumption baseline; An energy consumption map is constructed based on the total expected energy consumption of each drone, and the energy surplus of each drone is calculated. If the energy surplus of any drone is lower than the preset energy threshold, or if any drone has an abnormal energy consumption indicator, then the task will be redistributed based on the energy consumption map, and the remaining tasks of each drone will be updated.

2. The method for UAV flight control and energy allocation based on energy consumption perception as described in claim 1, characterized in that, The specific steps of the task reassignment include: If any drone's energy surplus is less than the preset energy threshold, or if there is an abnormal energy consumption indicator, then the corresponding drone will be designated as the mission initiator. The task initiator packages the transferable parts of its remaining tasks into one or more task packages, while querying its own energy consumption map and obtaining the energy consumption status information of all neighboring drones. Based on the energy consumption status information of all neighboring drones, neighboring drones that are qualified to receive mission packets are selected to obtain candidate drones, and an interruptibility coefficient is set according to the flight stage of each candidate drone. The penalty factor is calculated based on the task type label and the expected energy consumption of the task, and the task acceptance rate of each candidate UAV is calculated based on the interruptibility coefficient and the penalty factor.

3. The method for UAV flight control and energy allocation based on energy consumption perception as described in claim 2, characterized in that, The task reassignment steps also include: Sort all candidate UAVs by task acceptance in descending order, and select the k candidate UAVs with the highest task acceptance as task receivers; where k is a positive integer. Send a task package to the task recipient and receive a confirmation of receipt from the task recipient; Count the number of received confirmation messages. If the number is greater than 1, select the task recipient with the highest task acceptance rate as the task executor and send a reconfirmation message to the task executor. If the number of entries is equal to 1, then the task recipient is directly designated as the task executor, and a reconfirmation instruction is sent to the task executor.

4. The method for UAV flight control and energy allocation based on energy consumption perception as described in claim 3, characterized in that, The task reassignment steps also include: If the number of confirmed reception instructions is 0, then the candidate drones other than the top k candidate drones with the highest task reception will be regarded as task receivers. Send a task packet to the task recipient and recount the number of received confirmation instructions; When the task executor receives the reconfirmation receipt instruction, it adds the task waypoint sequence corresponding to the task package to its remaining tasks, while the task initiator removes the task package from its remaining tasks.

5. The method for UAV flight control and energy allocation based on energy consumption perception as described in claim 4, characterized in that, The penalty factor is calculated as follows: If a candidate drone has an abnormal energy consumption flag, a base penalty value is set according to the abnormal type of the abnormal energy consumption flag. The weight of the base penalty value is set according to the task type label, and the base penalty value is multiplied by the weight of the base penalty value to obtain the abnormal penalty item; If no abnormal energy consumption indicator is found in the candidate drone, the abnormal penalty item will be set to 0. The energy matching degree is obtained by calculating the ratio of the expected energy consumption of the task to the remaining energy of the candidate UAV. If the energy matching degree is greater than the preset matching degree threshold, the difference between the energy matching degree and the matching degree threshold is calculated, and an energy matching penalty term is calculated based on the difference. The energy matching penalty term is proportional to the difference; If the energy matching degree is less than or equal to the matching degree threshold, then the energy matching penalty term is 0; The sum of the anomaly penalty term and the energy matching penalty term is calculated to obtain the penalty factor.

6. The method for UAV flight control and energy allocation based on energy consumption perception as described in claim 5, characterized in that, The energy consumption baselines for each flight phase are constructed as follows: Extract energy consumption data from historical data of M complete flight missions, and remove energy consumption data with abnormal energy consumption indicators to obtain reference data; where M is a positive integer; The reference data is classified according to flight phase to obtain the reference dataset corresponding to each flight phase; For each flight phase, the average instantaneous power consumption of each functional module is calculated, and the arithmetic mean of the average instantaneous power consumption of all functional modules is calculated to obtain the average power consumption of the corresponding flight phase. Using the average power consumption as the dependent variable and the vacuum speed, air pressure altitude, angle of attack, pitch angle and engine throttle opening in the reference dataset as independent variables, a linear regression model is established, and the regression coefficients are fitted to obtain the energy consumption baseline for the corresponding flight phase.

7. The method for UAV flight control and energy allocation based on energy consumption perception as described in claim 6, characterized in that, The flight parameters are calculated as follows: using the energy efficiency coefficient corresponding to the current flight phase as the objective function, and vacuum speed, angle of attack, and heading as optimization variables, the flight parameters that maximize the energy efficiency coefficient are solved under preset constraints. The steps for solving the flight parameters that maximize the energy efficiency coefficient specifically include: Using the current vacuum speed, angle of attack, and heading as the initial points for iteration, iterative optimization is performed. In each iteration, the gradient of the objective function with respect to vacuum velocity, angle of attack, and heading is calculated at the current point using numerical difference, thus obtaining the first-order change of the objective function at the current point; Based on the first-order transformation, a second-order Taylor expansion is performed on the objective function at the current point, and the constraints are linearized to the first order.

8. The method for UAV flight control and energy allocation based on energy consumption perception as described in claim 7, characterized in that, The step of solving for the flight parameters that maximize the energy efficiency coefficient further includes: Using the second-order Taylor expansion as the objective function and the first-order linearization as the constraint, a quadratic programming subproblem is constructed; the quadratic programming subproblem is solved to obtain the search direction, and a line search is performed along the search direction to obtain the optimal step size; The current point is updated based on the optimal step size to obtain new flight parameters, which serve as the starting point for the next iteration. The change in the objective function between two adjacent iterations is also calculated. If the change value is less than the preset convergence threshold, or the number of iterations reaches the preset maximum number of iterations, then the flight parameters obtained in the last iteration will be used as the final flight parameters.

9. The method for UAV flight control and energy allocation based on energy consumption perception as described in claim 8, characterized in that, The specific method for decomposing the remaining tasks into consecutive task segments is as follows: Using the current real-time location of the UAV as the starting path point, and taking the first waypoint in the remaining waypoint list as the first target point, two adjacent path points are defined as a task segment in sequence; Traverse all remaining waypoints to obtain a sequence of task segments; in the sequence of task segments, calculate the geometric parameters of the corresponding task segment based on the three-dimensional coordinates of the path point corresponding to any task segment, including segment length, altitude change, horizontal projection length and segment orientation angle; Based on the geometric parameters and mission type labels, flight phases are assigned to each mission segment; The performance parameters of the corresponding UAV, including the maximum rate of climb and the maximum rate of descent, are obtained from the UAV's flight control calculator. Based on the performance parameters and the flight phase, the duration of each mission segment is calculated.

10. An energy consumption-aware UAV flight control and energy allocation system, used to implement the energy consumption-aware UAV flight control and energy allocation method as described in any one of claims 1-9, characterized in that, It includes a data acquisition module, an energy consumption analysis module, a task planning module, and an energy allocation module; The data acquisition module is used to collect energy consumption data, flight status data and operating environment data of each UAV in real time, and to detect abnormal energy consumption based on the energy consumption data and generate abnormal energy consumption identifiers. The energy consumption analysis module identifies the flight phase based on the flight status data and constructs the energy consumption baseline for each flight phase. The energy consumption analysis module also calculates the energy efficiency coefficient of each UAV based on the flight phase and operating environment data, calculates the flight parameters of each UAV based on the energy efficiency coefficient, and controls the corresponding UAV to adjust its flight status based on the flight parameters. The task planning module is used to obtain the remaining tasks of each UAV and decompose the remaining tasks into continuous task segments; allocate flight phases and durations to each task segment, and calculate the total expected energy consumption of the remaining tasks in combination with the corresponding flight parameters and energy consumption baseline; the task planning module also constructs an energy consumption map based on the total expected energy consumption of each UAV and calculates the energy surplus of each UAV. The energy allocation module is used to redistribute tasks based on the energy consumption map and update the remaining tasks of each drone when the energy surplus of any drone is lower than the preset energy threshold or when any drone has an abnormal energy consumption indicator.

Images