Distributed electric propulsion aircraft lateral course model prediction control method and system

Abstract

The invention provides a distributed electric propulsion aircraft lateral course model prediction control method and system, and the method comprises the steps: completing the six-degree-of-freedom linear state space modeling of a distributed electric propulsion aircraft, obtaining the aerodynamic parameters of the aircraft, and constructing a flight dynamics model containing multiple propellers; under the flight dynamics model of the multiple propellers, according to a flight task instruction and state variables of the current position and speed of the airplane, calculating an expected flight path, and converting the expected flight path into an attitude angle; the edge computing platform computes an optimal control instruction based on the flight state variable and the attitude angle, and the target attitude is realized by integrating the propeller, the control surface and the flight state constraint; and the edge computing platform realizes the optimal control of the lateral heading attitude of the aircraft according to the optimal control instruction. The precision, stability and fault tolerance of flight control are remarkably improved, and an efficient and reliable solution is provided for practical application of model prediction control in a distributed electric propulsion aircraft.

Description

Technical Field

[0001] The present invention relates to the field of aircraft control, and particularly to a lateral-directional model predictive control method and system for distributed electric propulsion aircraft. Background Art

[0002] Through the collaborative work of multiple electric thrusters, the distributed electric propulsion system can achieve efficient and precise attitude control, enabling the aircraft to adapt to complex flight environments and diverse mission requirements, and having broad application prospects. Traditional aircraft attitude control mainly relies on manipulating aerodynamic control surfaces such as elevators, rudders, and ailerons to adjust the flight attitude. However, this control method is relatively single, and in the face of complex flight environments or system failures (such as control surface failures or airframe damage), there is a risk of decreased control performance or even loss of control. In addition to relying on control surface control, distributed electric propulsion aircraft can also achieve attitude control through thrust adjustment of electric thrusters. However, how to coordinate the joint control of control surfaces and thrusters in a dynamic flight environment remains an urgent technical problem to be solved.

[0003] To solve this problem, model predictive control is introduced into the attitude control of distributed electric propulsion aircraft. As an optimal control strategy, model predictive control can calculate the optimal control command in real time according to the aircraft dynamics model within a predetermined prediction time domain, and fully consider the equality and inequality constraints of the system to ensure the feasibility and robustness of the control strategy. Compared with traditional PID control methods, model predictive control shows significant advantages in dealing with the nonlinear, coupled, and time-varying characteristics of aircraft, especially in complex flight tasks or emergency failure scenarios, and can quickly adjust the control strategy to respond to changes.

[0004] However, existing research has not fully combined the overall dynamics model of the aircraft, making it difficult to achieve efficient coordination between control surfaces and thrusters. In addition, due to the implementation of model predictive control algorithms relying on high-performance computing platforms and real-time data feedback, there are still some gaps in how to apply them in the attitude control of distributed electric propulsion aircraft. Summary of the Invention

[0005] The purpose of the present invention is to provide a lateral-directional model predictive control method and system for distributed electric propulsion aircraft to solve the above problems.

[0006] To achieve the above purpose, the present invention adopts the following technical solutions: In the first aspect, the present invention provides a lateral-directional model predictive control method for distributed electric propulsion aircraft, including: For a distributed electric propulsion aircraft, complete six-degree-of-freedom linear state space modeling and obtain aircraft aerodynamic parameters, and construct a flight dynamics model including multiple thrusters; Under the flight dynamics model of multiple thrusters, according to the flight mission instructions and the state variables of the aircraft's current position and speed, the desired flight trajectory is calculated and converted into attitude angles. Based on the flight state variables and attitude angles, the edge computing platform calculates the optimal control instructions and achieves the target attitude by integrating the thrusters, control surfaces, and flight state constraints. The edge computing platform realizes the optimal control of the aircraft's lateral and directional attitudes according to the optimal control instructions.

[0007] Furthermore, for the distributed electric propulsion aircraft, a six-degree-of-freedom linear state space model is established and the aircraft's aerodynamic parameters are obtained to construct a flight dynamics model including multiple thrusters, including: Based on the distributed electric propulsion aircraft, a model is established and the aircraft's aerodynamic parameters are obtained using computational fluid dynamics or wind tunnel experiments. First, a model of the controlled distributed electric propulsion aircraft is established, and the aircraft model is meshed and aerodynamically solved using computational fluid dynamics software; or using wind tunnel experiments, the aerodynamic derivatives are obtained to establish the dynamics model.

[0008] Furthermore, a six-degree-of-freedom linear state space model of the aircraft including multiple thrusters is established:

[0009] Among them, the state variables , is the flight speed, is the sideslip angle, is the angle of attack, is the roll rate, is the pitch rate, is the roll rate, is the roll angle, is the pitch angle, is the yaw angle, and the control variables , is the elevator deflection angle, is the aileron deflection angle, is the vertical tail deflection angle, , is the thrust of the thruster; The form of the discrete state space model is

[0010] Among them, is the dimensional measurable state variable, is the dimensional control input, is the dimensional system output.

[0011] Further, under the flight dynamics model of the multi-thruster, according to the flight mission instructions and the state variables of the aircraft's current position and speed, calculate the expected flight trajectory and convert it into attitude angles, including: According to the flight mission instructions input by the ground control system and the state variables such as the aircraft's current position and speed, calculate the expected flight trajectory and convert it into the corresponding attitude angles; preset the flight mission waypoints or input the flight waypoints in real time by the ground control system and transmit them to the flight controller. Combine the sensors such as GPS, IMU and airspeed indicator on the aircraft to measure the current position and attitude of the aircraft, calculate the expected roll angle and yaw angle according to the current position and the target trajectory, and calculate the expected thruster thrust and pitch angle according to the set altitude and speed.

[0012] Further, the edge computing platform calculates the optimal control instructions based on the flight state variables and attitude angles, and realizes the target attitude by integrating the thrusters, control surfaces and flight state constraints, including: (1) Construct the objective function

[0013] Among them, is the predicted system output value at time k + i at the current time k, is the reference input of the target values of the roll angle and yaw angle, p is the prediction time domain, m is the control time domain, and Q and R are weighting matrices; the larger the weighting factor in Q, the smaller the error between the system output corresponding to this factor and the reference quantity is desired; the larger the weighting factor in R, the smaller the control action corresponding to this factor is desired; (2) Add the control constraints of the thrusters and control surfaces and the flight state constraints to the model predictive control:

[0014] Among them,

[0015]

[0016] Among them,

[0017] (3) Based on the above control objectives and constraints, build a model predictive control algorithm model based on the ROS system framework, and then generate the ROS node code of the model predictive control through the ROS code generation tool, and deploy it to the edge computing platform for compilation and operation; (4) Run the model predictive control algorithm on the edge computer of the edge computing platform, obtain the sensor measurement data from the flight controller through the real-time communication module, and calculate the flight attitude state feedback; combine the expected attitude angles obtained in real time, calculate the optimal control instructions online, and generate the control signals for the electric thrusters and control surfaces.

[0018] Further, the edge computing platform realizes the optimal control of the aircraft's lateral and directional attitude according to the optimal control instruction, including: The edge computing platform transmits the optimal control instruction to the flight controller for parsing, and then transfers it to the actuator to drive multiple thrusters and control surfaces to work together. Modify and compile the control output channels, communication protocols, etc. of the flight control firmware according to the requirements of the controlled distributed electric propulsion aircraft, and flash the compiled flight control firmware to the flight controller; Subsequently, real-time communication between the flight controller and the edge computing platform is realized through MAVROS, and the optimal control instruction calculated by the model predictive controller is sent to the thrusters and servos through the flight controller, so as to realize the attitude control of the aircraft.

[0019] Further, when the control instruction output by the model predictive controller is lower than the set safety threshold, the control system will automatically switch to the PID controller.

[0020] In a second aspect, the present invention provides a lateral and directional model predictive control system for a distributed electric propulsion aircraft, including: A model construction module for performing six-degree-of-freedom linear state space modeling on the distributed electric propulsion aircraft, obtaining the aircraft's aerodynamic parameters, and constructing a flight dynamics model including multiple thrusters; An expected flight trajectory acquisition module for calculating the expected flight trajectory according to the flight mission instruction and the state variables of the aircraft's current position and speed under the flight dynamics model of multiple thrusters, and converting it into attitude angles; An attitude control module for the edge computing platform to calculate the optimal control instruction based on the flight state variables and attitude angles, and realize the target attitude by integrating the thrusters, control surfaces and flight state constraints; the edge computing platform realizes the optimal control of the aircraft's lateral and directional attitude according to the optimal control instruction.

[0021] In a third aspect, the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the lateral and directional model predictive control method for the distributed electric propulsion aircraft are realized.

[0022] In a fourth aspect, the present invention provides a computer-readable storage medium storing a computer program, and when the computer program is executed by a processor, the steps of the lateral and directional model predictive control method for the distributed electric propulsion aircraft are realized.

[0023] Compared with the prior art, the present invention has the following technical effects: The present invention breaks the traditional aerodynamic control surface for aircraft attitude control. The distributed electric propulsion technology can enable the aircraft attitude control through the collaborative work of multiple electric thrusters and control surfaces. Even in the case of failure of control surfaces, one or more electric thruster failures and other fault scenarios, it can still achieve the aircraft attitude control, effectively improving the dynamic response, handling performance and system fault tolerance of the distributed electric propulsion aircraft; The present invention realizes the optimal control allocation between the control surfaces and electric thrusters of the distributed electric propulsion aircraft based on model predictive control. For a preset flight mission, the position control module calculates the future desired attitude angle in real time according to the flight trajectory, fully considering the constraints of aircraft state variables, control variables, etc. Then, the model predictive controller can calculate the optimal control command according to the desired attitude angle and the current state feedback by the sensor; The present invention successfully applies the model predictive control to an actual distributed electric propulsion aircraft. To apply the model predictive control to the actual system, the communication between the flight controller and the edge computing platform is realized through MAVROS, and the optimal control command calculated by the model predictive controller in real time is transmitted to the flight controller to adjust the output of the servo and electric thrusters to achieve precise flight attitude control. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a flow block diagram for the deployment and implementation of the model predictive control algorithm of the present invention.

[0025] Figure 2 It is a flow chart of the model predictive control algorithm of the present invention.

[0026] Figure 3 It is a flow chart of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0027] The present invention will be described in detail below with reference to the drawings and embodiments: Embodiment 1, please refer to Figure 3 , the present invention provides a lateral and directional model predictive control method for a distributed electric propulsion aircraft, including: For the distributed electric propulsion aircraft, complete the six-degree-of-freedom linear state space modeling and obtain the aircraft aerodynamic parameters, and construct a flight dynamics model including multiple thrusters; Under the flight dynamics model of multiple thrusters, calculate the desired flight trajectory according to the flight mission command and the state variables of the current position and speed of the aircraft, and convert it into the attitude angle; The edge computing platform calculates the optimal control command based on the flight state variables and the attitude angle, and realizes the target attitude by integrating the constraints of the thrusters, control surfaces and flight state; The edge computing platform realizes the optimal control of the lateral and directional attitude of the aircraft according to the optimal control command.

[0028] Embodiment 2. The present invention provides a lateral-directional model predictive control method for a distributed electric propulsion aircraft, which specifically includes: Step 1. This system consists of an edge computing platform, a flight controller, an actuator, and a ground control system. Among them, the edge computing platform is responsible for real-time flight data processing, online optimization calculation of the model predictive controller, and instruction transmission; the flight controller is used to collect flight state data, parse and forward control instructions; the actuator includes multiple electric thrusters and control surfaces, and is used to execute control instructions; the ground control system is used to input flight mission instructions and communicate with the flight controller, and the overall coordination completes the lateral-directional control of the distributed electric propulsion aircraft. Step 2. Based on the studied distributed electric propulsion aircraft, complete the modeling and obtain the aircraft aerodynamic parameters by using computational fluid dynamics or wind tunnel experiments, and construct a six-degree-of-freedom linear state space dynamic model including multiple thrusters. Step 3. The position control module calculates the desired flight trajectory according to the flight mission instructions input by the ground control system and the state variables such as the current position and speed of the aircraft, and converts it into the corresponding attitude angles. Step 4. The edge computing platform obtains the flight state feedback and the desired attitude angles from the flight controller in real time. The lateral-directional control calculates the optimal control instructions for the differential value of the thrusters, the aileron deflection angle, and the vertical tail deflection angle by the model predictive controller, and realizes the target attitude by comprehensively considering the constraints of the thrusters, control surfaces, and flight states. Step 5. Design a fault protection measure. When the control instructions output by the model predictive controller are lower than the set safety threshold, the system will switch to the working mode of the PID controller to ensure the continuity of attitude control and the safety of the aircraft. Step 6. The edge computing platform transmits the optimal control instructions to the flight controller for parsing, and then transfers them to the actuator to drive multiple electric thrusters and control surfaces to work together to achieve the optimal control of the lateral-directional attitude of the aircraft.

[0029] Pre-model the designed distributed electric propulsion aircraft; Use computational fluid dynamics software to perform mesh division and aerodynamic solution on the aircraft model, or manufacture a distributed electric propulsion prototype for wind tunnel experiments to obtain high-precision aerodynamic parameters and establish a dynamic model; In view of the complexity of the dynamic model, combined with the small perturbation linearization theory, while maintaining the model accuracy, simplify the computational complexity to form a six-degree-of-freedom linearized state space model of the aircraft including multiple thrusters suitable for real-time control.

[0030] Among them, the state variables , is the flight speed, is the sideslip angle, is the angle of attack, is the roll angular velocity, is the pitch angular velocity, is the roll angular velocity, is the roll angle, is the pitch angle, is the yaw angle, control variable , is the elevator deflection angle, is the aileron deflection angle, is the vertical tail rudder deflection angle, , is the thrust of the thruster; The basic form of the discrete state - space model is:

[0031] Among them, is dimensional measurable state variables, is dimensional control input, is dimensional system output; According to the flight mission instructions input by the ground control system, the flight controller fuses the data of GPS, IMU, airspeed meter and other sensors to obtain the position information and attitude parameters of the aircraft in real - time; Then, through the path - following controller of the position control module, the desired roll angle and yaw angle are calculated based on the current position and the target trajectory, and the desired thruster thrust and pitch angle are calculated by the total energy controller according to the target altitude and speed; When designing the model predictive control, the optimization goal is to improve the attitude angle tracking accuracy by minimizing the control cost function and optimize the system energy consumption at the same time, which is reflected in the objective function as:

[0032] Among them, is the predicted system output value at time k for time k + i,

[0033] is the reference input of the target values of the roll angle and yaw angle, p is the prediction horizon, m is the control horizon, Q and R are the weighting matrices; the larger the weighting factor in Q, the smaller the error between the system output corresponding to this factor and the reference quantity is expected; the larger the weighting factor in R, the smaller the corresponding control action is expected; When designing the model predictive control, the control constraints of the thruster and control surfaces and the flight state constraints are added:

[0034] Among them,

[0035]

[0036] Among them,

[0037] Based on the above control objectives and condition constraints, communication with the flight control system can be achieved based on the ROS system framework, using MATLAB / Simulink to build a model predictive control algorithm in MATLAB / Simulink, generating ROS node code for model predictive control through the ROS code generation tool, and deploying it to the edge computing platform for compilation and operation, so as to realize the execution environment of the real-time control algorithm to the embedded; Running the model predictive control algorithm on the edge computing platform, obtaining sensor measurement data from the flight controller in real time through MAVROS, including key information such as flight attitude, speed, and position, and based on the current flight state and the desired attitude angle transmitted by the position control module, calculating the optimal lateral and longitudinal attitude control commands by the model predictive controller; Since the implementation of the model predictive control algorithm depends on a high-performance computing platform and real-time data feedback, a perfect fault protection measure needs to be designed in the actual control system; For example, when the control command output by the model predictive controller is lower than the set safety threshold, the control system will automatically switch to the PID controller working mode, that is, when the edge computing platform causes delay or interruption due to excessive computing load or hardware failure, the control system can still maintain the basic attitude control ability to ensure the safety and flight stability of the aircraft.; In addition, the control outputs of the distributed electric propulsion aircraft are independent of each other. It is necessary to redefine and modify the flight control firmware, integrate custom functions into the flight controller by adding new control output channels and recompiling and flashing the firmware to support the coordinated control of multiple electric thrusters and control surfaces; Finally, the edge computing platform transmits the optimal control command to the flight controller for parsing, and then passes it to the actuator to drive multiple electric thrusters and control surfaces to work together to achieve the optimal control of the aircraft's lateral and longitudinal attitude.

[0038] Embodiment 3, the present invention provides a method for lateral model predictive control of a distributed electric propulsion aircraft, including: Such as Figure 1 As shown, the implementation of the method and system for lateral model predictive control of a distributed electric propulsion aircraft based on an edge computing platform includes the following modules: a ground control system, a flight controller, an edge computer platform, an actuator, a sensor system, and an example distributed electric propulsion aircraft: The main function of the ground control system is to receive and process flight mission requirements, including but not limited to setting flight mission waypoints, planning mission flight paths, and realizing real-time switching and control of flight modes. In addition, the ground control system can also achieve direct intervention through remote control operation. At the same time, the ground control system and the flight controller can communicate wirelessly through the MAVLink communication protocol, sending waypoint instructions, control mode switching signals, etc., to realize the docking of mission planning and the flight controller.

[0039] The sensor system is the core data acquisition module in the whole control process, including GPS module, IMU module, airspeed indicator and other auxiliary sensors. This system can collect flight data (such as position, speed, attitude, etc.) in real time, and use state estimation algorithms (such as Kalman filter) to fuse multi-sensor data, output high-precision position and attitude information, and provide reliable input data for the subsequent control module.

[0040] The GPS navigation and positioning module supports multi-signal reception of four satellite navigation systems: Beidou, Galileo, Glonass and GPS, and has high-precision horizontal positioning. The IMU module uses triaxial accelerometers and gyroscopes with high dynamic range, which can accurately capture high-frequency vibrations and attitude changes during flight, and meet the dynamic control requirements in complex flight environments. The airspeed indicator is used to measure the oncoming flow speed of the aircraft, and combines with the static pressure sensor to reflect the aerodynamic state of the aircraft in real time. The flight controller realizes a low-latency and high-concurrency task scheduling mechanism through an optimized real-time operating system (RTOS), ensuring the real-time performance and reliability of flight control, and efficiently exchanging data with the sensor system and the control algorithm model through the uORB (publish / subscribe) communication mechanism, supporting the execution of complex flight tasks.

[0041] The edge computing platform integrates CUDA parallel computing capabilities internally, which can accelerate the solution efficiency of the model predictive control algorithm to meet the high-performance requirements of real-time flight control. At the same time, the edge computing platform uses MAVROS as a communication bridge to interact with the flight controller in real time for flight attitude data, runs the model predictive control algorithm to generate optimal control instructions, and ensures precise control of the aircraft attitude and trajectory. It is the core computing unit of the system. The given example of the distributed electric propulsion aircraft adopts a multi-motor distributed layout. The electric thrusters are arranged transversely along the leading edge of the wing, with four electric thrusters configured on each of the left and right sides, providing significant thrust redundancy characteristics and flexible distributed control capabilities, and can effectively cope with the situation of thruster or control surface failures. When the electric thruster or rudder surface fails, the optimized control algorithm can compensate for the failed part by adjusting the thrust distribution of other electric thrusters to ensure the attitude control capabilities of roll and yaw. Especially in the case of rudder surface failure, the thrust difference between electric thrusters can replace the function of the rudder surface to maintain the attitude stability of the aircraft. The main structures of the aircraft include the fuselage, wings, electric propulsion system, tail, and sensor module; the flight controller and edge computing platform are integrated inside the fuselage, and the sensor module is distributed at key parts of the aircraft to measure and feedback flight state parameters in real time.

[0042] The following is the specific implementation process of the example: First, conduct dynamic modeling on the distributed electric propulsion aircraft; Specifically, it is necessary to establish a model of the controlled distributed electric propulsion aircraft, then use computational fluid dynamics software to mesh the aircraft model and perform aerodynamic solutions to obtain aerodynamic derivatives; manufacture a distributed electric propulsion prototype and use wind tunnel experiments to obtain high-precision aerodynamic parameters to establish a dynamic model. These aerodynamic derivatives are used to establish the aircraft dynamic model as the basis for controller design; Since the aircraft dynamic model is usually very complex and difficult to directly apply to real-time control, it is necessary to simplify the model; combined with the small perturbation linearization theory, the computational complexity can be simplified while maintaining the modeling accuracy to obtain a six-degree-of-freedom linearized state space model suitable for real-time control; this model takes into account the characteristics of multiple thrusters and can accurately describe the dynamic behavior of the distributed electric propulsion aircraft:

[0043] Among them, the state variables , is the flight speed, is the sideslip angle, is the angle of attack, is the roll angular velocity, is the pitch angular velocity, is the roll angular velocity, is the roll angle, is the pitch angle, is the yaw angle. The control variables , is the elevator deflection angle, is the aileron deflection angle, is the vertical tail deflection angle, , is the thrust of the thruster; The basic form of the discrete state space model is:

[0044] Among them, is The \(n\) -dimensional measurable state variable, is the \(m\) -dimensional control input, is the \(p\) -dimensional system output.

[0045] First, perform dynamic modeling on the distributed electric propulsion aircraft; Secondly, modify and compile the PX4 flight control firmware for the studied distributed electric propulsion aircraft: Since the control output of the distributed electric propulsion aircraft has a high degree of independence, it is necessary to modify the underlying layer of the flight control firmware, add additional control output channels and re - define their functions. After the modification is completed, compile the flight control firmware and flash the new code to the flight controller.

[0046] Then, set the mission waypoints according to the flight mission requirements and calculate the desired attitude angles and thrust values by the position control module: According to the set mission waypoints, the ground station of the ground control system gives them to the flight controller platform through the MAVLINK communication protocol. At this time, the path - following controller and the total - energy controller of the position control module combine the actual state information of the aircraft provided by the sensor system to calculate the desired attitude commands and thrust values of the aircraft; Among them, the core principle of the total - energy control is based on the total - energy management of the aircraft, and realizes altitude and speed control by adjusting the thruster thrust and pitch angle. The goal is to ensure that the aircraft maintains an appropriate energy balance during flight. The system calculates the required thruster thrust and pitch angle through the difference between the current flight state (such as altitude, speed, thrust, etc.) and the target state (such as target altitude, target speed); then, through PI controller adjustment, calculates the final longitudinal control output to ensure the flight stability of the aircraft and appropriate energy management; The path - following controller obtains the current position, speed and heading information of the aircraft in real - time, compares it with the predetermined target trajectory, calculates the lateral error (the vertical distance from the aircraft to the trajectory) and the heading error (the angle between the current heading of the aircraft and the trajectory direction) relative to the target trajectory, uses the L1 control algorithm to calculate the desired roll angle and yaw angle, and transmits them to the edge - computing platform through MAVLINK.

[0047] Finally, the edge - computing platform calculates, deploys and generates the optimal control commands based on the model predictive control algorithm: According to the design requirements of the model predictive control algorithm, first determine the optimization objective, that is, by minimizing the control cost function, improving the attitude - angle tracking accuracy, and at the same time optimizing the system energy consumption:

[0048] Among them, To predict the system output value at time k + i at the current time k,

[0049] is the reference input of the target values of the roll angle and yaw angle, p is the prediction horizon, m is the control horizon, Q and R are weighting matrices; the larger the weighting factor in Q, the smaller the error between the system output corresponding to this factor and the reference quantity is desired; the larger the weighting factor in R, the smaller the control action corresponding to this factor is desired; When designing the model predictive controller, various constraint conditions are introduced, including the thruster thrust range, the rudder deflection angle range, and the physical limitations of the aircraft flight state, which are expressed as linear inequality constraints, enabling the controller to generate optimal control commands on the premise of meeting safety conditions:

[0050] Among them,

[0051]

[0052] Among them,

[0053] In addition, the model predictive controller also combines path constraints and state constraints to ensure that the aircraft always stays within the safe flight envelope during the execution of the flight mission; Combining the above dynamic model and conditional constraints, based on the ROS system framework, MATLAB / Simulink can achieve communication with the flight control system; and build a model predictive control algorithm in MATLAB / Simulink. This algorithm solves the online optimization problem in each control cycle to ensure that the generated control command is the optimal solution under the current conditions; subsequently, through the ROS code generation tool, the ROS node code generated by the model predictive control algorithm is deployed to the edge computing platform and compiled; After compilation, the edge computing platform runs the model predictive control algorithm, receives in real time the flight attitude data and desired attitude angles parsed by the flight controller through MAVROS, and calculates online the optimal thruster differential value, aileron deflection angle, and vertical tail deflection angle commands; Finally, the calculated optimal control commands are published in the form of ROS topics and parsed by MAVROS into MAVLink messages and transmitted to the flight controller to drive the electric thrusters and servos to perform corresponding operations, realizing precise control of the aircraft attitude.

[0054] In addition, to ensure the safety and reliability of the control system, a fault protection mechanism is designed; The implementation of the fault protection mechanism relies on the fault-tolerant monitoring modules on the edge computing platform and the flight controller. This module can detect the effectiveness of the controller output in real time and quickly switch the control mode when necessary; When the control command output by the model predictive controller is lower than the set safety threshold, the system will automatically switch to the PID controller mode to cope with the possible calculation delay or interruption caused by excessive computing load or hardware failure of the edge computing platform, and avoid the loss of control of the aircraft due to controller failure.

[0055] An important feature of this system is its modular design, which is very convenient for subsequent function expansion. For example, by adding new sensor modules and updating control algorithms, it can support more types of flight tasks; especially the model predictive control algorithm built based on MATLAB / Simulink and deployed into the system, which greatly simplifies the algorithm development process and improves the development efficiency of flight control algorithms.

[0056] Figure 2 The flowchart of the model predictive control algorithm of the present invention is shown, which details the complete logic of path planning, modeling, prediction, and cooperative control, reflecting the full-process closed-loop design of the distributed electric propulsion aircraft control system: According to the preset flight mission requirements and environmental constraint conditions, input the target mission of the aircraft, and then generate a specific flight path through the path planning algorithm, including spatial position and waypoints, while meeting the aircraft performance limitations; Calculate the desired attitude angles (such as pitch angle, roll angle) and thrust requirements of the aircraft at each waypoint through the position control module, and adjust the output in real time according to the flight trajectory error to generate a reference attitude that meets the path requirements; Obtain the aerodynamic characteristic parameters of each thruster of the aircraft through computational fluid dynamics simulation or wind tunnel experiment, and build a dynamic model including thruster layout and mechanical characteristics based on the aerodynamic data; Perform small perturbation linearization on the non-linear flight dynamics equation under steady flight conditions to generate a linearized state space model reflecting the six-degree-of-freedom dynamic characteristics of the aircraft, the thrust action of multiple thrusters, and their coupling relationship; Set constraint conditions and design objective functions according to the actual limitations of the aircraft control system (such as maximum thruster thrust, rudder deflection angle range), including optimization objectives such as minimizing attitude angle error and minimizing energy consumption; Port the model predictive control algorithm built by MATLAB / Simulink to the edge computing platform to achieve high-performance embedded code deployment and real-time optimization calculation, and generate the optimal control command to output to the flight controller; The output command calculated by the model predictive control determines whether it meets the frequency requirements and physical constraint conditions, avoiding exceeding limitations such as thruster thrust or rudder deflection. If the constraint conditions are not met, it switches to a traditional PID controller to ensure the stability and safety of the aircraft; The flight controller transmits the optimal control command to each control surface and thruster, forming a multi-channel cooperative control strategy. Through the combined adjustment of the control surfaces and electric thrusters, it realizes the precise control of the target flight attitude and completes the flight mission; Based on the edge computing platform, the invention realizes the lateral and directional model predictive control method and system for distributed electric propulsion aircraft. From flight mission planning to actual control execution, it combines model predictive control, dynamics modeling, and embedded implementation. Its core is to utilize differential propulsion and control surface cooperative control to achieve efficient flight and attitude control through model predictive control, and at the same time sets up a fault protection mechanism.

[0057] In another embodiment of the present invention, a lateral and directional model predictive control system for distributed electric propulsion aircraft is provided, which can be used to implement the above-mentioned 8. Lateral and directional model predictive control method for distributed electric propulsion aircraft. Specifically, the system includes: A model construction module, which is used to complete the six-degree-of-freedom linear state space modeling of the distributed electric propulsion aircraft and obtain the aircraft aerodynamic parameters, and construct a flight dynamics model including multiple thrusters; An expected flight trajectory acquisition module, which is used to calculate the expected flight trajectory based on the flight mission command and the state variables of the aircraft's current position and speed under the flight dynamics model of multiple thrusters, and convert it into attitude angles; An attitude control module, which is used for the edge computing platform to calculate the optimal control command based on the flight state variables and attitude angles, and achieve the target attitude by integrating the thrusters, control surfaces, and flight state constraints; the edge computing platform realizes the optimal control of the aircraft's lateral and directional attitude according to the optimal control command.

[0058] The division of modules in the embodiments of the present invention is illustrative, only a logical function division. In actual implementation, there may be other division methods. In addition, in each embodiment of the present invention, the functional modules can be integrated in one processor, or can exist separately physically, or two or more modules can be integrated in one module. The above-mentioned integrated modules can be implemented in the form of hardware or in the form of software function modules.

[0059] In another embodiment of the present invention, a computer device is provided. The computer device includes a processor and a memory. The memory is used to store a computer program, and the computer program includes program instructions. The processor is used to execute the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, and is suitable for implementing one or more instructions. Specifically, it is suitable for loading and executing one or more instructions in the computer storage medium to implement the corresponding method flow or corresponding function. The processor described in the embodiment of the present invention can be used for the operation of the lateral-directional model predictive control method of a distributed electric propulsion aircraft.

[0060] In another embodiment of the present invention, a storage medium is also provided, specifically a computer-readable storage medium (Memory). The computer-readable storage medium is a memory device in a computer device and is used to store programs and data. It can be understood that the computer-readable storage medium here can include both the built-in storage medium in the computer device and, of course, the extended storage medium supported by the computer device. The computer-readable storage medium provides a storage space that stores the operating system of the terminal. And in this storage space, one or more instructions suitable for being loaded and executed by the processor are also stored. These instructions can be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory. One or more instructions stored in the computer-readable storage medium can be loaded and executed by the processor to implement the corresponding steps of the lateral-directional model predictive control method of a distributed electric propulsion aircraft in the above embodiment.

[0061] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0062] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present invention. It should be understood that each flow and / or block in the flowchart and / or block diagram, as well as the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices produce means for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0063] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory produce a manufactured article including instruction means that implement the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0064] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0065] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: still can modify the specific implementation manners of the present invention or make equivalent replacements, and any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. A distributed electric propulsion aircraft lateral heading model predictive control method, characterized in that: include: For distributed electric propulsion aircraft, complete six-degree-of-freedom linear state space modeling and obtain aircraft aerodynamic parameters, and build a flight dynamics model including multiple thrusters; Under the multi-thruster flight dynamics model, the expected flight trajectory is calculated based on the flight mission instructions and the state variables of the aircraft's current position and speed, and converted into attitude angles; The edge computing platform calculates the optimal control instructions based on flight state variables and attitude angles, and integrates thrusters, control surfaces, and flight state constraints to achieve the target attitude. The edge computing platform achieves optimal control of the aircraft's lateral heading attitude based on the optimal control instructions.

2. The distributed electric propulsion aircraft lateral heading model predictive control method according to claim 1, characterized in that: The distributed electric propulsion aircraft is modeled in a six-degree-of-freedom linear state space and the aircraft aerodynamic parameters are obtained, and a flight dynamics model including multiple propellers is constructed, including: Based on the distributed electric propulsion aircraft, complete the modeling and use computational fluid dynamics or wind tunnel experiments to obtain the aircraft aerodynamic parameters, build a model of the controlled distributed electric propulsion aircraft in advance, use computational fluid dynamics software to mesh the aircraft model and perform aerodynamic solutions; or use wind tunnel experiments to obtain aerodynamic derivatives to establish a dynamic model.

3. The distributed electric propulsion aircraft lateral heading model predictive control method according to claim 2, characterized in that: Build a six-degree-of-freedom linear state-space model of an aircraft with multiple thrusters: Among them, the state quantity , is the flight speed, is the sideslip angle, is the angle of attack, is the rolling angular velocity, is the pitch angular velocity, is the rolling angular velocity, is the roll angle, is the pitch angle, is the yaw angle, the control amount , is the elevator deflection angle, is the aileron deflection angle, is the vertical rudder deflection angle, , is the thrust of the propeller; The discrete state space model has the form in, for dimensional measurable state variables, for Dimensional control input, for Dimension system output.

4. The distributed electric propulsion aircraft lateral heading model predictive control method according to claim 1, characterized in that: Under the multi-thruster flight dynamics model, the expected flight trajectory is calculated according to the flight mission instructions and the state variables of the current position and speed of the aircraft, and converted into an attitude angle, including: Based on the flight mission instructions input by the ground control system and the aircraft's current position, speed and other state variables, the expected flight trajectory is calculated and converted into the corresponding attitude angle; the flight mission waypoints are set in advance or input in real time by the ground control system and transmitted to the flight controller, and the aircraft's current position and attitude are measured in combination with sensors such as the GPS, IMU and airspeed meter on the aircraft. The expected roll angle and yaw angle are calculated based on the current position and target trajectory, and the expected thruster thrust and pitch angle are calculated based on the set altitude and speed.

5. The distributed electric propulsion aircraft lateral heading model predictive control method according to claim 1, characterized in that: The edge computing platform calculates the optimal control instructions based on the flight state variables and attitude angles, and integrates the thrusters, control surfaces and flight state constraints to achieve the target attitude, including: (1) Constructing the objective function in, To predict the system output value at time k+i at the current time k, is the reference input of the target value of the roll angle and the yaw angle, p is the prediction time domain, m is the control time domain, Q and R are weighting matrices; the larger the weighting factor in Q, the smaller the error between the system output corresponding to the factor and the reference quantity is expected to be; the larger the weighting factor in R, the smaller the control action corresponding to the factor is expected to be; (2) Model predictive control adds control constraints for thrusters and control surfaces as well as flight state constraints: in, in, (3) Based on the above control objectives and constraints, a model predictive control algorithm model is built based on the ROS system framework. Then, the ROS node code of the model predictive control is generated through the ROS code generation tool, and deployed to the edge computing platform for compilation and operation; (4) Run the model predictive control algorithm on the edge computer of the edge computing platform, obtain sensor measurement data from the flight controller through the real-time communication module, and calculate the flight attitude state feedback; combine the expected attitude angle obtained in real time, calculate the optimal control instructions online, and generate control signals for the electric thrusters and control surfaces.

6. The distributed electric propulsion aircraft lateral heading model predictive control method according to claim 1, characterized in that: The edge computing platform realizes optimal control of the aircraft's lateral heading attitude according to the optimal control instruction, including: The edge computing platform transmits the optimal control instructions to the flight controller for analysis, and then passes them to the actuators to drive multiple thrusters and control surfaces to work together. The control output channels, communication protocols, etc. of the flight control firmware are modified and compiled according to the needs of the controlled distributed electric propulsion aircraft, and the compiled flight control firmware is flashed to the flight controller. Subsequently, real-time communication between the flight controller and the edge computing platform is achieved through MAVROS, and the optimal control instructions calculated by the model predictive controller are sent to the thrusters and servos through the flight controller, thereby realizing attitude control of the aircraft.

7. The distributed electric propulsion aircraft lateral heading model predictive control method according to claim 1, characterized in that: When the control command output by the model predictive controller is lower than the set safety threshold, the control system will automatically switch to the PID controller.

8. A distributed electric propulsion aircraft lateral heading model predictive control system, characterized in that: include: The model building module is used to complete the six-degree-of-freedom linear state space modeling and obtain the aircraft aerodynamic parameters for the distributed electric propulsion aircraft, and to build a flight dynamics model including multiple thrusters; The expected flight trajectory acquisition module is used in the multi-thruster flight dynamics model to calculate the expected flight trajectory according to the flight mission instructions and the state variables of the aircraft's current position and speed, and convert it into an attitude angle; The attitude control module is used for the edge computing platform to calculate the optimal control instructions based on the flight state variables and attitude angles, and to achieve the target attitude by integrating the thrusters, control surfaces and flight state constraints; the edge computing platform achieves optimal control of the aircraft's lateral heading attitude according to the optimal control instructions.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the distributed electric propulsion aircraft lateral heading model predictive control method as claimed in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the distributed electric propulsion aircraft lateral heading model predictive control method as claimed in any one of claims 1 to 7 are implemented.

Images