Self-adaptive flight control method and device
Through the adaptive flight control method, the flight control model and PID controller are used to solve the stability and accuracy of UAV flight control in the power inspection scenario, and efficient and precise inspection tasks are achieved under complex conditions. .
Patent Information
- Application Number
- CN202411167363.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-23
- Publication Date
- 2025-01-03
AI Technical Summary
In the complex and changeable power inspection scenarios, there are difficulties in the flight control of drones, making it difficult to maintain stable flight and accurately identify subtle defects in power equipment.
Adaptive flight control method is adopted to obtain the flight status information of the drone, and input the flight control model after preprocessing. The flight control model based on the long and short-term memory network outputs the control parameters, and combines the PID controller to adjust the drone's Position, position and speed ensure accurate and stable inspection tasks execution.
It improves the flight efficiency and accuracy of drones in power inspection scenarios, reduces the risks and costs of human operations, and can maintain stable flight under complex conditions and accurately identify equipment defects.
Smart Images
Figure CN119247738A_ABST
Abstract
Description
Technical Field
[0001] The embodiments of the present application relate to the field of artificial intelligence technology, and in particular, to an adaptive flight control method and device. Background Art
[0002] With the growing demand for intelligent and automated inspections in the power industry, drones are becoming increasingly important as efficient and flexible inspection tools. Power inspections not only require drones to be able to traverse complex terrain environments, such as mountains, canyons, and densely populated urban areas, but also to maintain a stable flight state under changing weather conditions (such as strong winds, lightning, haze, etc.) and accurately identify subtle defects on power equipment. Traditional flight control methods have many difficulties when applied to complex and changing power inspection scenarios. Summary of the invention
[0003] In view of this, the purpose of the embodiments of the present application is to propose an adaptive flight control method and device to solve the flight control problem of UAVs in power inspection scenarios.
[0004] Based on the above objectives, an embodiment of the present application provides an adaptive flight control method, including:
[0005] Get the flight status information of the drone;
[0006] Preprocessing the flight state information to obtain a flight state vector;
[0007] Inputting the flight state vector into a pre-built flight control model, and using the flight control model to output control parameters of the UAV;
[0008] The UAV is controlled to perform the target mission according to the control parameters.
[0009] Optionally, the flight state information includes the attitude, position and speed of the drone, the flight state vector includes the corresponding attitude vector, position vector and speed vector, and the drone is controlled by a PID controller;
[0010] The flight state vector is input into a pre-built flight control model, and the control parameters of the UAV are output using the flight control model, including:
[0011] The attitude vector, position vector and velocity vector are input into the flight control model, and the flight control model is used to output the control parameters of the PID controller.
[0012] Optionally, the flight control model is constructed based on a long short-term memory network.
[0013] Optionally, the training method of the flight control model includes:
[0014] During the training process, the model parameters are adjusted according to the preset reward function.
[0015] Optionally, the reward function is expressed as:
[0016] R=ω 1 R p +ω 2 R s +ω 3 R q (1)
[0017]
[0018] Among them, R p is the accuracy reward, R s is the stability reward, R q is the energy efficiency ratio reward, ω 1 ,ω 2 ,ω 3 is the weight coefficient, k p is the coefficient of the accuracy reward, E thres is the accuracy deviation threshold, and E is the deviation between the actual flight state of the UAV and the target flight state.
[0019] The embodiment of the present application provides an adaptive flight control device, comprising:
[0020] The acquisition module is used to obtain the flight status information of the drone;
[0021] A preprocessing module, used for preprocessing the flight state information to obtain a flight state vector;
[0022] A parameter determination module, used to input the flight state vector into a pre-built flight control model, and output control parameters of the UAV using the flight control model;
[0023] The control module is used to control the UAV to perform the target task according to the control parameters.
[0024] Optionally, the flight state information includes the attitude, position and speed of the drone, the flight state vector includes the corresponding attitude vector, position vector and speed vector, and the drone is controlled by a PID controller;
[0025] The parameter determination module is used to input the attitude vector, position vector and velocity vector into the flight control model, and use the flight control model to output the control parameters of the PID controller.
[0026] Optionally, the flight control model is constructed based on a long short-term memory network.
[0027] Optionally, the device further comprises:
[0028] The training module is used to adjust the model parameters according to a preset reward function during the training process of the flight control model.
[0029] Optionally, the reward function is expressed as:
[0030] R=ω 1 R p +ω 2 R s +ω 3 R q (1)
[0031]
[0032] Among them, R p is the accuracy reward, R s is the stability reward, R q is the energy efficiency ratio reward, ω 1 ,ω 2 ,ω 3 is the weight coefficient, k p is the coefficient of the accuracy reward, E thres is the accuracy deviation threshold, and E is the deviation between the actual flight state of the UAV and the target flight state.
[0033] From the above description, it can be seen that the adaptive flight control method and device provided in the embodiment of the present application obtains the flight status information of the UAV, pre-processes the flight status information, obtains the flight status vector, inputs the flight status vector into a pre-built flight control model, and uses the flight control model to output the control parameters of the UAV. The UAV is controlled to perform the target task according to the control parameters, which can ensure that the UAV can accurately and stably perform the inspection task in the power inspection scenario. BRIEF DESCRIPTION OF THE DRAWINGS
[0034] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.
[0035] Figure 1 A schematic diagram of a method flow of an embodiment of the present application;
[0036] Figure 2 A schematic diagram of the flight control principle of an embodiment of the present application;
[0037] Figure 3This is a block diagram of the device structure of an embodiment of the present application;
[0038] Figure 4 This is a structural block diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION
[0039] In order to make the objectives, technical solutions and advantages of the present disclosure more clearly understood, the present disclosure is further described in detail below in combination with specific embodiments and with reference to the accompanying drawings.
[0040] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the embodiments of the present application should be the usual meanings understood by people with ordinary skills in the field to which the present disclosure belongs. The "first", "second" and similar words used in the embodiments of the present application do not represent any order, quantity or importance, but are only used to distinguish different components. "Including" or "comprising" and similar words mean that the elements or objects appearing in front of the word cover the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connecting" or "connected" and similar words are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to represent relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.
[0041] As mentioned in the background technology section, the application of drones in power inspection requires solving the problem of flight control under complex conditions. Adopting an adaptive flight control method based on artificial intelligence, using a flight control model to determine the control parameters of the drone according to real-time flight status information, and controlling the drone to perform inspection tasks according to the control parameters can improve the efficiency and accuracy of inspections and reduce the risks and costs of manual operations.
[0042] The technical solution of the present application is further described in detail below through specific embodiments.
[0043] like Figure 1 , 2 As shown, the embodiment of the present application provides an adaptive flight control method, including:
[0044] S101: Obtain the flight status information of the drone;
[0045] S102: Preprocess the flight state information to obtain a flight state vector;
[0046] In some embodiments, the flight status information of the UAV is obtained, including the attitude, position and speed of the UAV, etc., and each item of flight status information can be measured by a measurement unit such as an inertial measurement unit, a positioning unit, and an airspeed meter provided on the UAV. After obtaining each item of flight status information, in order to facilitate the processing of the flight control model, the flight status information needs to be preprocessed to obtain a flight status vector suitable for processing by the flight control model. Among them, the preprocessing of the flight status information includes: data cleaning of each item of flight status information, removing noise and outliers, etc., to ensure the accuracy and reliability of the data; standardizing the cleaned data, converting data of different dimensions and ranges to the same scale, obtaining standardized data, selecting key attitude features, velocity features and position features based on the standardized data, and converting them into corresponding attitude vectors, velocity vectors and position vectors.
[0047] S103: inputting the flight state vector into a pre-built flight control model, and using the flight control model to output control parameters of the UAV;
[0048] In this embodiment, after preprocessing the flight state information, the obtained attitude vector, position vector and velocity vector are input into the flight control model, and the flight control model is used to output the control parameters. Among them, the UAV adopts a PID controller for feedback control, and the flight control model outputs the control parameters of the PID controller, including the proportional coefficient, the integral coefficient and the differential coefficient.
[0049] In some embodiments, the flight control model is constructed based on a long short-term memory network (Long Short-Term Memory, LSTM). The LSTM network includes an input layer, a hidden layer, an output layer and other structures, wherein the attitude vector, the position vector and the velocity vector are input by the input layer, and the hidden layer is composed of multiple LSTM units, which are used to extract the time series characteristics of the input data. The LSTM unit realizes effective memory and update of historical information through mechanisms such as forget gates, input gates and output gates. The output layer is used to output the predicted control parameters of the PID controller according to the extracted time series characteristics. Since the LSTM network can capture long-term dependencies, the PID controller can predict future control parameters based on past flight status information to adjust the flight trend, thereby responding to possible flight changes in advance and improving the predictability and stability of the control.
[0050] In view of the characteristics of flight status information such as fast dynamic changes, high dimension, and strong temporal sequence, the network's ability to capture complex flight states is enhanced by increasing the network depth, adjusting the number of LSTM units, and introducing the attention mechanism. A multi-level feature extraction module is designed, and combined with pre-processing layers such as convolutional neural networks (CNN), the input flight state vector is reduced in dimension and features are extracted to improve the quality of input data and processing efficiency.
[0051] In some embodiments, the training method of the flight control model includes: adjusting the model parameters according to a preset reward function during the training process. In this embodiment, during the model training process, the control accuracy, flight stability and energy consumption of the UAV are comprehensively considered, and a reward function with accuracy, stability and energy consumption as the goals is established, and the flight control of the UAV is optimized according to the goals.
[0052] Among them, control accuracy is the degree of deviation between the actual flight state of the drone and the set target flight state. The smaller the deviation, the higher the control accuracy. For control accuracy, the reward mechanism set includes positive rewards and negative penalties. When the deviation between the actual flight state of the drone and the target flight state is less than the preset threshold, a positive reward is given. The reward value can increase as the deviation decreases, which is used to encourage the control system to further reduce the deviation; when the deviation exceeds the threshold, a negative penalty is given, and the severity of the penalty increases as the deviation increases, so that the control system can adjust as soon as possible to reduce the deviation.
[0053] Flight stability refers to the stability of the attitude and speed of the drone during flight. A drone with better stability will not experience severe turbulence or sudden changes in speed during flight. For stability, stability evaluation indicators are set, such as attitude angle change rate, speed change rate, position change rate, etc., and the stability evaluation indicators are used to reflect the stability of the drone during flight. Rewards or penalties are given according to the value of the stability index. When the stability index remains at a low level, that is, the flight is more stable, positive rewards are given; when the stability index increases, that is, the flight stability decreases, negative penalties are given.
[0054] Energy consumption is the total energy consumed by the drone during flight, which is usually directly related to the battery life and endurance of the drone. For energy consumption, an energy efficiency ratio is set, which is the ratio of the energy required by the drone to complete a specific task to its flight distance or time. The higher the energy efficiency ratio, the higher the energy utilization efficiency. The energy efficiency ratio is introduced as a reward factor in the reward function. When the energy efficiency ratio increases, that is, energy consumption decreases, a positive reward is given; when the energy efficiency ratio decreases, that is, energy consumption increases, a negative penalty is given or the reward remains unchanged.
[0055] In some ways, the reward function R is expressed as:
[0056] R=ω 1 R p +ω 2 R s +ω 3 R q (1)
[0057]
[0058] Among them, R p is the accuracy reward, R s is the stability reward, R q is the energy efficiency ratio reward, ω 1 ,ω 2 ,ω 3 is the weight coefficient, which is used to adjust the relative importance of different goals in the reward function. Each weight coefficient can be adjusted according to specific tasks and needs; k p is the coefficient of the precision reward, which is used to adjust the magnitude of the precision reward; E represents the deviation between the actual flight state of the UAV (i.e., the real-time attitude, position, and speed) and the target flight state (the preset target attitude, target position, and target speed); E thresh is the precision deviation threshold, which is used to define when to start reducing the reward. When the deviation E exceeds the precision deviation threshold, the reward will begin to decrease linearly until it is reduced to zero. If the deviation is less than or equal to the precision deviation threshold, the precision reward is calculated based on the size of the deviation. The smaller the deviation, the greater the precision reward. The maximum precision reward is k. p , if the deviation is greater than the precision deviation threshold, the precision reward is zero.
[0059] In some embodiments, during the training of the model, an adaptive learning rate adjustment algorithm is used to dynamically adjust the learning rate according to the gradient changes during the training process, accelerate the convergence process and avoid falling into a local optimal solution. Batch training is adopted to process multiple flight state vectors in each iteration to accelerate the training process. At the same time, the flight control model can improve performance and adaptability through continuous learning and optimization.
[0060] S104: Control the UAV to perform the target mission according to the control parameters.
[0061] In this embodiment, after the control parameters of the UAV are determined by the flight control model, the control parameters are input into the PID controller, so that the PID controller controls the action of the actuator of the UAV according to the determined control parameters, and performs specific tasks by adjusting the attitude, speed, position and other states of the UAV in real time. At the same time, as the state of the UAV changes, the real-time control parameters are determined by the flight control model based on the real-time UAV flight state information, and the optimal control effect of the UAV is achieved through closed-loop control.
[0062] In some ways, when drones are performing power inspection tasks, they can plan flight paths, avoid obstacles and dangerous areas, and use high-precision sensors and image processing technology to accurately identify defects and hidden dangers on power equipment.
[0063] The adaptive flight control method provided in the embodiment of the present application combines the feedback control of the UAV with artificial intelligence, determines the current control parameters based on the flight status information acquired in real time using the flight control model, and dynamically adjusts the UAV according to the control parameters using a PID controller, thereby achieving adaptive adjustment and optimized control of the UAV, ensuring that the UAV can maintain precise and stable performance and meet application requirements under complex conditions.
[0064] The scheme and effect of this application are explained below through experimental data.
[0065] In this embodiment, the open virtual experiment software platform CoppeliaSim is used to carry out the UAV hovering control experiment and the full drive control experiment.
[0066] In the hovering virtual experiment, the initial position of the UAV is set to [0m, 0m, 1m], the expected position is [2m, 2m, 2m], and the Euler angles between the initial position and the expected position are [0°, 0°, 0°]. The method of the present application is compared with the finite-time sliding mode algorithm. As shown in Table 1, compared with the comparative method, the method of the present application has a smaller average control law (a control performance index in UAV flight control, the smaller the value, the better the algorithm performance). As shown in Table 2, the average tracking error of the present application is significantly smaller than that of the comparative method. It can be seen that the overall performance of the present application is more excellent and can meet the working conditions of the hovering operation of the flying platform.
[0067] Table 1 Statistics of average control law of attitude system
[0068] This application Comparison Method Pitch Angle 80.52 87.83 Roll Angle 30.55 49.68 Yaw angle 5.39 6.32
[0069] Table 2 Statistics of average tracking error of attitude system
[0070] This application Comparison Method Pitch Angle 0.01122 0.03135 Roll Angle 0.00356 0.01235 Yaw angle 0.02357 0.05623
[0071] In the full-drive control virtual experiment, the present application tests the control performance during flight. Assuming the initial position of the drone is [0m, 0m, 2m], the expected position is [3m, 3m, 3m], the initial Euler angle is [5°, 5°, 5°], and the expected Euler angle is [0°, 0°, 0°]. Compare the method of the present application with the linear self-disturbance rejection method. As shown in Tables 3 and 4, it can be seen that the results of the present application are better than the comparison method. It can be seen that the overall performance of the present application is better when performing aerial work tasks in motion.
[0072] Table 3 Statistics of average control law of position system
[0073]
[0074]
[0075] Table 4 Statistics of average tracking error of position system
[0076] method This application Comparison Method x 0.2435 0.3565 y 0.2397 0.3125 z 0.0879 0.2123
[0077] It should be noted that the method of the embodiment of the present application can be performed by a single device, such as a computer or server. The method of this embodiment can also be applied to a distributed scenario and completed by multiple devices cooperating with each other. In the case of such a distributed scenario, one of the multiple devices can only perform one or more steps in the method of the embodiment of the present application, and the multiple devices will interact with each other to complete the described method.
[0078] It should be noted that the above is a description of a specific embodiment of the present specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0079] like Figure 3 As shown, an embodiment of the present application provides an adaptive flight control device, comprising:
[0080] The acquisition module is used to obtain the flight status information of the drone;
[0081] A preprocessing module, used for preprocessing the flight status information to obtain a flight status vector;
[0082] A parameter determination module is used to input the flight state vector into a pre-built flight control model and output the control parameters of the UAV using the flight control model;
[0083] The control module is used to control the UAV to perform the target mission according to the control parameters.
[0084] In some embodiments, the flight state information includes the attitude, position and speed of the drone, the flight state vector includes the corresponding attitude vector, position vector and speed vector, and the drone is controlled by a PID controller;
[0085] The parameter determination module is used to input the attitude vector, position vector and velocity vector into the flight control model, and output the control parameters of the PID controller using the flight control model.
[0086] In some embodiments, the flight control model is constructed based on a long short-term memory network.
[0087] In some embodiments, the adaptive flight control device further comprises:
[0088] The training module is used to adjust the model parameters according to a preset reward function during the training process of the flight control model.
[0089] In some embodiments, the reward function is expressed as:
[0090] R=ω 1 R p +ω 2 R s +ω 3 R q (1)
[0091]
[0092] Among them, R p is the accuracy reward, R s is the stability reward, R q is the energy efficiency ratio reward, ω 1 ,ω 2 ,ω 3 is the weight coefficient, k p is the coefficient of the accuracy reward, E thresh is the accuracy deviation threshold.
[0093] For the convenience of description, the above devices are described in terms of functions divided into various modules. Of course, when implementing the embodiments of the present application, the functions of each module can be implemented in the same or multiple software and / or hardware.
[0094] The apparatus of the above-mentioned embodiment is used to implement the corresponding method in the above-mentioned embodiment, and has the beneficial effects of the corresponding method embodiment, which will not be described in detail here.
[0095] Figure 4 A more specific schematic diagram of the hardware structure of an electronic device provided in this embodiment is shown, and the device may include: a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. The processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040 are connected to each other through the bus 1050 in the device.
[0096] The processor 1010 can be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.
[0097] The memory 1020 may be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 1020 may store an operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented by software or firmware, the relevant program codes are stored in the memory 1020 and are called and executed by the processor 1010.
[0098] The input / output interface 1030 is used to connect the input / output module to realize information input and output. The input / output module can be configured in the device as a component (not shown in the figure), or it can be externally connected to the device to provide corresponding functions. The input device may include a keyboard, a mouse, a touch screen, a microphone, various sensors, etc., and the output device may include a display, a speaker, a vibrator, an indicator light, etc.
[0099] The communication interface 1040 is used to connect a communication module (not shown) to realize communication interaction between the device and other devices. The communication module can realize communication through a wired mode (such as USB, network cable, etc.) or a wireless mode (such as mobile network, WIFI, Bluetooth, etc.).
[0100] The bus 1050 includes a path that transmits information between the various components of the device (eg, the processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040).
[0101] It should be noted that, although the above device only shows the processor 1010, the memory 1020, the input / output interface 1030, the communication interface 1040 and the bus 1050, in the specific implementation process, the device may also include other components necessary for normal operation. In addition, it can be understood by those skilled in the art that the above device may also only include the components necessary for implementing the embodiments of the present specification, and does not necessarily include all the components shown in the figure.
[0102] The electronic device of the above embodiment is used to implement the corresponding method in the above embodiment, and has the beneficial effects of the corresponding method embodiment, which will not be described in detail here.
[0103] The computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, read-only compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device.
[0104] Those skilled in the art should understand that the discussion of any of the above embodiments is merely illustrative and is not intended to imply that the scope of the present disclosure (including the claims) is limited to these examples. Based on the concept of the present disclosure, the technical features in the above embodiments or different embodiments may be combined, the steps may be implemented in any order, and there are many other variations of the different aspects of the embodiments of the present application as described above, which are not provided in detail for the sake of simplicity.
[0105] In addition, to simplify the description and discussion, and in order not to make the embodiments of the present application difficult to understand, the known power supply / ground connection with the integrated circuit (IC) chip and other components may or may not be shown in the provided drawings. In addition, the device can be shown in the form of a block diagram to avoid making the embodiments of the present application difficult to understand, and this also takes into account the fact that the details of the implementation of these block diagram devices are highly dependent on the platform to be implemented in the embodiments of the present application (that is, these details should be fully within the scope of understanding of those skilled in the art). In the case of elaborating specific details (e.g., circuits) to describe exemplary embodiments of the present disclosure, it is obvious to those skilled in the art that the embodiments of the present application can be implemented without these specific details or when these specific details are changed. Therefore, these descriptions should be considered illustrative rather than restrictive.
[0106] Although the present disclosure has been described in conjunction with specific embodiments of the present disclosure, many replacements, modifications and variations of these embodiments will be apparent to those skilled in the art from the foregoing description. For example, other memory architectures (e.g., dynamic RAM (DRAM)) may use the embodiments discussed.
[0107] The embodiments of the present application are intended to cover all such substitutions, modifications and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the embodiments of the present application should be included in the scope of protection of the present disclosure.
Claims
1. An adaptive flight control method, characterized in that: include: Get the flight status information of the drone; Preprocessing the flight state information to obtain a flight state vector; Inputting the flight state vector into a pre-built flight control model, and using the flight control model to output control parameters of the UAV; The UAV is controlled to perform the target mission according to the control parameters.
2. The method according to claim 1, characterized in that The flight state information includes the attitude, position and speed of the UAV, the flight state vector includes the corresponding attitude vector, position vector and speed vector, and the UAV is controlled by a PID controller; The flight state vector is input into a pre-built flight control model, and the control parameters of the UAV are output using the flight control model, including: The attitude vector, position vector and velocity vector are input into the flight control model, and the flight control model is used to output the control parameters of the PID controller.
3. The method according to claim 1, characterized in that The flight control model is constructed based on a long short-term memory network.
4. The method according to claim 3, characterized in that The training method of the flight control model comprises: During the training process, the model parameters are adjusted according to the preset reward function.
5. The method according to claim 4, characterized in that The reward function is expressed as: R=ω1R p +ω2R s +ω3R q (1) Among them, R p is the accuracy reward, R s is the stability reward, R q is the energy efficiency ratio reward, ω1, ω2, ω3 are weight coefficients, k p is the coefficient of the accuracy reward, E thres is the accuracy deviation threshold, and E is the deviation between the actual flight state of the UAV and the target flight state.
6. An adaptive flight control device, characterized in that: include: The acquisition module is used to obtain the flight status information of the drone; A preprocessing module, used for preprocessing the flight state information to obtain a flight state vector; A parameter determination module, used to input the flight state vector into a pre-built flight control model, and output control parameters of the UAV using the flight control model; The control module is used to control the UAV to perform the target task according to the control parameters.
7. The device according to claim 6, characterized in that The flight state information includes the attitude, position and speed of the UAV, the flight state vector includes the corresponding attitude vector, position vector and speed vector, and the UAV is controlled by a PID controller; The parameter determination module is used to input the attitude vector, position vector and velocity vector into the flight control model, and use the flight control model to output the control parameters of the PID controller.
8. The device according to claim 5, characterized in that The flight control model is constructed based on a long short-term memory network.
9. The device according to claim 8, characterized in that Also includes: The training module is used to adjust the model parameters according to a preset reward function during the training process of the flight control model.
10. The device according to claim 9, characterized in that The reward function is expressed as: R=ω1R p +ω2R s +ω3R q (1) Among them, R p is the accuracy reward, R s is the stability reward, R q is the energy efficiency ratio reward, ω1, ω2, ω3 are weight coefficients, k p is the coefficient of the accuracy reward, E thres is the accuracy deviation threshold, and E is the deviation between the actual flight state of the UAV and the target flight state.