CN111824397A - Flight control-landing gear control-terrain recognition multi-system fusion control system - Google Patents

Abstract

The invention belongs to the field of control systems, and relates to a flight control-undercarriage control-terrain recognition multi-system fusion control system. The system comprises: the system comprises a flight control system, a bionic leg type undercarriage control system, a terrain recognition system, a data transmission system and a ground system. The invention is based on an unmanned vertical take-off and landing aircraft platform, a bionic leg type undercarriage system is installed, a multi-system fusion control system based on sensor feedback information fusion is formed, and the multi-system fusion control system is used for improving the terrain self-adaptive capacity of the unmanned vertical take-off and landing aircraft during landing.

Description

Flight control-landing gear control-terrain recognition multi-system fusion control system

technical field

The invention belongs to the field of control systems, and relates to a multi-system fusion control system of flight control-landing gear control-terrain recognition.

Background technique

Unmanned vertical take-off and landing vehicles have attracted much attention in the military field due to their small size, flexibility, and no casualties. Hovering, low-speed flight, low-altitude ultra-low-altitude flight, in-situ steering, flying in any direction, etc. Therefore, when the take-off and landing site is limited, the flight space is small, and the low-altitude and low-speed tasks are required, it has broad application prospects as an ideal aircraft. In recent years, due to the increasingly complex modern tasks that need to be performed, the intelligent level of unmanned vertical take-off and landing vehicles has become more and more demanding. The angle requirements are relatively high, and it may be difficult to take off and land normally when encountering some special terrain such as rocky beaches and areas with large slopes.

At this stage, due to the development of computer technology and the needs of performing tasks, domestic and foreign scientific research institutions have successively carried out research on the autonomous landing technology of unmanned helicopters and multi-rotor UAVs. For example, the PLA University of Science and Technology and the State Key Laboratory of Helicopter Rotor Dynamics of Nanjing University of Aeronautics and Astronautics have designed a visual image feature processing algorithm to realize the landing control of unmanned helicopters and have achieved good results; The method of installing two cameras with parallel optical axes is used for autonomous landing on warships and on land; Zhejiang University's "Yuquan Wings" unmanned helicopter has realized the first domestic hovering technology in the 2005 Air Robot Competition, and has successively realized autonomous routes. The flight function, in 2010, realized the autonomous landing function based on the airborne sensor. However, due to the complexity of vision and image processing technology and the shortcomings of traditional fixed landing gear only as support and passive energy absorption, the real-time, stability and robustness of unmanned vertical take-off and landing aircraft landing in harsh environments There are problems, and even the difficulty in fulfilling the functional requirements.

SUMMARY OF THE INVENTION

Purpose of the invention: Because the traditional unmanned vertical take-off and landing aircraft has shortcomings in the design of autonomous take-off and landing, the present invention discloses a multi-system fusion control system of flight control-landing gear control-terrain recognition, which is used to control the unmanned vertical take-off and landing. The process of autonomous take-off and landing of an aircraft. The present invention is based on an unmanned vertical take-off and landing aircraft platform, and a bionic leg landing gear system is installed to form a multi-system fusion control system based on sensor feedback information fusion, which is used to improve the terrain adaptive capability of the unmanned vertical take-off and landing aircraft during landing. . The system can effectively improve the self-adaptive ability of the unmanned vertical take-off and landing vehicle during landing, and reduce the dependence on the flight control system's own capabilities and the landing environment. The system can realize the adjustment and control of the attitude and parking position of the unmanned vertical take-off and landing aircraft during the landing process, and change the passive control of the landing process to the active control, forming an effective fusion control system. The system can realize the autonomous fuselage balance adjustment ability of the unmanned vertical take-off and landing aircraft during the landing process, which can effectively improve the safety of landing. The system can effectively improve the landing adaptability of unmanned vertical take-off and landing aircraft, break through many limitations of the application environment, and conform to the development trend of multi-purpose, intelligent and complex environment adaptability of aircraft.

Technical solutions:

The invention provides a multi-system fusion control system of flight control-landing gear control-terrain recognition, comprising: a flight control system, a bionic leg landing gear control system and a terrain recognition system;

Among them, the flight control system determines the realization and control performance of the autonomous flight process of the unmanned vertical take-off and landing vehicle; the bionic legged landing gear control system is the control system of the bionic legged landing gear, which determines the attitude and movement process of the landing gear. Cooperate with the flight control system to complete the self-adaptive take-off and landing of the unmanned vertical take-off and landing aircraft; the terrain recognition system is to achieve a large-scale selection of landing terrain before the unmanned vertical take-off and landing aircraft lands, accurately identify the bottom surface, and determine the landing area. The landing range realizes real-time online tracking of the landing site, and informs the flight control system and the bionic legged landing gear control system of terrain information, so as to realize the real-time navigation of the flight control system and the attitude pre-swing and adjustment of the bionic legged landing gear;

The data information of the flight control system, the leg landing gear control system and the terrain recognition system are all communicated with the ground system through the data transmission system to realize the ground monitoring of the data and status information.

Further, the flight control system includes: a flight control computer, a sensor mechanism and an executive mechanism; the flight control computer is designed to realize the flight control algorithm of the unmanned vertical take-off and landing aircraft, processes the real-time status information to be transmitted by the sensor mechanism, and generates itself. The transmission control command is transmitted to the actuator; the sensor mechanism collects the altitude information, attitude information, acceleration information and other flight data of the unmanned vertical take-off and landing vehicle, and transmits it to the flight control computer for calculation; the function of the actuator is to realize the flight The instructions of the control computer are executed to complete the autonomous flight of the unmanned vertical take-off and landing aircraft.

Further, the flight control system needs an onboard power supply to supply power for the flight control-landing gear control-terrain recognition multi-system fusion control system.

Further, the bionic leg landing gear control system includes: a controller, a motor driver, a modular joint unit and a load sensor;

Among them, the controller realizes the control algorithm design of the bionic leg landing gear, completes the position and attitude calculation and attitude control, and processes the state information fed back by the motor driver and the load sensor; the motor driver receives the signal from the controller and completes the solution of the motor control signal. The closed-loop control of the modular joint unit completes the action of the bionic leg; the modular joint unit is used as the driving joint of the bionic leg; the load sensor is a sensor installed at the foot end, which is used to measure the force signal feedback after the bionic leg landing gear touches the ground. This is transmitted to the controller for post-landing adaptive control.

Further, the terrain recognition system includes: lidar, visual camera, inertial measurement unit IMU inertial navigation and high-performance onboard processor;

The high-performance onboard processor is used to: perform joint spatiotemporal calibration of lidar and visual camera, realize fusion of laser point cloud and image pixel information, and perform real-time online low-altitude UAV based on the fusion information and IMU inertial navigation information. Terrain modeling, according to the modeled 3D terrain data, accurately identify the terrain information, automatically identify the landing point according to the terrain information and realize the real-time online tracking of the landing point, calculate the attitude of the UAV relative to the landing point in real time and feed it to the flight controller system.

Further, the flight control system and the bionic legged landing gear control system conduct communication and data exchange, the terrain recognition system and the bionic legged landing gear control system conduct communication and data exchange, and there is no direct communication between the flight control system and the terrain recognition system.

Further, the flight control computer includes a DSP and a programmable gate array FPGA; the sensor mechanism includes another inertial measurement unit IMU, a GPS navigation system, a compass, and an altimeter; and the execution mechanism includes an engine and a steering gear.

Further, the modular joint unit includes a brake, a servo motor and an encoder.

Advantages of the present invention:

A flight control-landing gear control-terrain recognition multi-system fusion control system is proposed, combined with the bionic leg landing gear system, to realize the autonomous landing of a real unmanned vertical take-off and landing vehicle. In the field of military and civilian use, the invention can greatly improve the intelligence level of unmanned vertical take-off and landing aircraft, improve its recovery rate and utilization rate, break the landing scene limitations and endurance limitations of unmanned vertical take-off and landing aircraft, and help To improve aircraft and landing safety. In addition to small unmanned vertical take-off and landing aircraft, applying this technology to large-scale aircraft can realize autonomy and intelligence in many aspects of material transportation, disaster relief, and industrial production.

Description of drawings

Taking the mechanical leg mechanism with 4 degrees of freedom and in the extended state as an example, the connection relationship is explained.

Fig. 1 is the component relation schematic diagram of the present invention;

Figure 2 is a schematic diagram of the composition of the flight control system;

Figure 3 is a schematic diagram of the composition of the landing gear control system;

Figure 4 is a schematic diagram of the composition of the terrain recognition system;

Figure 5 is a schematic diagram of communication and work collaboration.

In the figure: 1—flight control system, 2—bionic leg landing gear control system, 3—terrain recognition system, 4—data transmission system, 5—ground system; 11—flight control computer, 111—DSP, 112—programmable gate Array FPGA, 12—sensor mechanism, 121—inertial measurement unit IMU, 122—navigation system, 123—compass, 124—altimeter, 13—actuator, 131—engine, 132—steering gear, 14—onboard power supply; 21— Controller, 22—motor driver, 23—modular joint unit, 231—brake, 232—servo motor, 233—encoder, 24—load sensor; 31—lidar, 32—vision camera, 33—inertial measurement unit ( Inertial measurement unit, IMU) inertial navigation, 34 - high-performance onboard processor.

Detailed ways

The present invention provides a flight control-landing gear control-terrain recognition multi-system fusion control system, as shown in FIG. 1 , including: a flight control system 1, a bionic leg landing gear control system 2, a terrain recognition system 3, and a data transmission system 4 and ground systems 5.

As shown in Fig. 2, the flight control system 1 determines the realization and control performance of the autonomous flight process of the unmanned vertical take-off and landing aircraft. The flight control system 1 is mainly composed of flight control computer 11 , sensor mechanism 12 and actuator 13 and other parts. The flight control computer 11 includes a DSP 111 and a programmable gate array FPGA 112 , and its main functions are to realize the flight control algorithm design of the unmanned vertical take-off and landing aircraft, process the real-time status information transmitted by the sensor mechanism 12 , and transmit control commands to the executive mechanism 13 . The sensor mechanism 12 includes an inertial measurement unit IMU121 (including a 3D gyroscope and a 3D accelerometer), a GPS navigation system 122, a compass 123, and an altimeter 124. The main function is to collect the altitude information, attitude information, acceleration information, etc. of the unmanned vertical take-off and landing aircraft. flight data, and send it to the flight control computer 11 for calculation. The actuator 13 includes an engine 131 and a steering gear 132, and its main function is to execute the instructions of the flight control computer 11 and complete the autonomous flight of the unmanned vertical take-off and landing aircraft. The flight control system 1 requires an onboard power supply 14 to supply power to the flight control computer 11, the sensor mechanism 12, and the like.

As shown in Figure 3, the bionic legged landing gear control system 2 is the control system of the bionic legged landing gear, which determines the attitude and movement process of the landing gear, and cooperates with the flight control system 1 to complete the self-adaptive takeoff and landing of the unmanned vertical take-off and landing aircraft. . The bionic leg landing gear control system 2 is mainly composed of a controller 21 , a motor driver 22 , a modular joint unit 23 and a load sensor 24 . The controller 21 is a single-board industrial computer, and its main function is to realize the control algorithm design of the bionic leg landing gear, complete the position and attitude calculation and attitude control, and process the state information fed back by the motor driver 22 and the load sensor 24 . The function of the motor driver 22 is to receive the signal from the controller 21, complete the calculation of the motor control signal, and control the modular joint unit 23 in a closed loop to complete the action of the bionic leg. The modular joint unit 23 is used as the driving joint of the bionic leg, and is composed of a brake 231 , a servo motor 232 and an encoder 233 . The load sensor 24 is a sensor installed at the foot end, and is used to measure the force signal feedback after the bionic leg landing gear touches the ground and transmit it to the controller 21 to complete the adaptive control after landing.

As shown in Figure 4, the function of the terrain recognition system 3 is to select a wide range of landing terrain, accurately identify jungles, roads, rivers, buildings, flats, etc. Real-time online tracking of the landing site in the range of landing, and inform flight control system 1 and bionic legged landing gear control system 2 of terrain information to realize real-time navigation of flight control system 1 and attitude pre-swing and adjustment of bionic legged landing gear , and then provide the basis for the realization of autonomous landing of unmanned vertical take-off and landing aircraft. The terrain recognition system 3 includes a lidar 31 , a vision camera 32 , an IMU inertial navigation 33 and a high-performance onboard processor 34 . The working process is as follows: joint spatio-temporal calibration of lidar 31 and visual camera 32 to achieve fusion of laser point cloud and image pixel information; then, based on image frame information and point cloud information, study the large-scale method using both geometric and texture features. The real-time reconstruction technology of large-scale 3D scenes realizes the real-time online low-altitude terrain modeling of UAVs. Next, according to the established real-time 3D terrain data, study how to use deep learning methods to combine the image features and geometric features of objects to achieve robust The semantic segmentation of the drone can accurately identify the terrain information; finally, the method based on machine learning automatically identifies the landing point and realizes the real-time online tracking of the landing point, and calculates the attitude of the UAV relative to the landing point in real time and feeds it to the flight control system 1 .

As shown in Figure 5, during the working process, the flight control system 1 communicates and exchanges data with the bionic legged landing gear control system 2, and the terrain recognition system 3 also communicates and exchanges data with the bionic legged landing gear control system 2. There is no direct communication between the control system 1 and the terrain recognition system 3. Finally, the data information of the flight control system 1, the leg landing gear control system 2 and the terrain recognition system 3 are all communicated with the ground system through the data transmission system 4 to realize ground monitoring of data and status information. In the debugging stage, the data of the flight control computer 11 of the flight control system 1, the controller 21 of the bionic leg landing gear control system 2 and the high-performance onboard processor 34 of the terrain recognition system 3 are shared through EtherCAT. The higher integration level realizes all its data calculation and control functions through a high-performance on-board computer.

The embodiments of the present invention mainly have two modes.

Specific Example 1: Autonomous Landing on Complex Ground

The unmanned vertical take-off and landing vehicle flies under the action of the flight control system 1. When it is about to land, it reaches a certain height range (such as 50m ~ 100m height), and when it reaches the height range, the flight control system 1 gives instructions to the bionic leg landing gear control system 2 , enter the standby state, and then instruct the terrain recognition system 3 to start the work. The lidar 31 and the visual camera 32 of the terrain recognition system 3 are activated, scan the terrain information, obtain the landing range through the high-performance onboard processor 34, and give a signal of whether the landing can be carried out, together with the coordinate information of the landing range, through the bionic legs The landing gear control system 2 transmits it to the flight control system 1. If it can land, the flight control system 1 controls the unmanned vertical take-off and landing aircraft to fly slowly in the X and Y directions (as shown in the coordinates in Figure 1) to find the determined suitable landing area. When landing, the flight control system 1 controls the unmanned vertical take-off and landing vehicle to fly in the X direction or the Y direction, find the landing point again, and repeat the above process.

In the case of being able to land, the flight control system 1 monitors the flight height in real time during the landing process of the unmanned vertical take-off and landing aircraft. When it reaches a height of 10m, the flight control system 1 controls the unmanned vertical take-off and landing aircraft to hover, and the bionic leg landing gear Turn it on to a right-angle state, and the soles of the feet are on the same plane. At this time, the lidar 31 and the visual camera 32 of the terrain recognition system 3 are activated again to accurately scan the terrain and geological information directly below and within a range of 3 meters to ensure that it is within the landing range. , the unmanned vertical take-off and landing vehicle continues to descend slowly, as far as possible to ensure that there is no displacement in the X direction or the Y direction.

During the slow descent, the terrain recognition system 3 scans and monitors the terrain information directly under the unmanned vertical take-off and landing vehicle in real time, and transmits the terrain coordinate information to the bionic leg landing gear control system 2 in real time. The coordinate information obtained from the solution completes the pre-swing of the bionic leg landing gear, and adjusts the pre-swing attitude in real time according to the terrain. After landing, that is, a strong signal feedback from the load sensor 24 at a certain foot end, the terrain recognition system 3 stops working and enters the feedback force control stage.

After landing, the posture of the bionic leg landing gear is adjusted by force control. At this stage, the flight control system 1 ensures that the fuselage of the unmanned vertical take-off and landing aircraft is level, and adapts to the ground through the leg landing gear until the legs and feet end. The force signals of the load sensors 24 are consistent, and the leg posture adjustment and landing are completed. After stopping, the system shuts down.

Specific embodiment 2: autonomous take-off

The unmanned vertical take-off and landing vehicle is parked on complex ground, the system is activated, and it is ready to take off. When the legged landing gear is lifted off the ground, that is, when it is detected that the measured forces of all the foot end load sensors 24 are 0, the leg posture is adjusted by the legged landing gear control system 2 to a right angle state so that all the foot ends are at the same level status to prevent the unexpected need to land here.

The flight control system 1 controls the aircraft to gradually rise and monitors the height information in real time. When the height reaches 10m, the flight control system 1 transmits a signal to the bionic leg landing gear control system 2 to control the bionic leg landing gear to retract. The leg landing gear control system 2 feeds back a retraction completion signal to the flight control system 1 , and the unmanned vertical take-off and landing vehicle flies away under the control of the flight control system 1 to perform the task.

The advantages of the present invention are as follows:

(1) The present invention is based on the flight control system, the bionic leg landing gear control system and the terrain recognition system, and realizes the integrated design and data sharing of the three parts. The system is an overall working mode, and its reliability is higher. The mode of distributed operation of each system can shorten the data transmission time and reduce the probability of errors in the process of working logic design;

(2) The present invention forms a new type of flight control system that is different from the original flight control system, can adapt to the collaborative work and integrated design of multiple systems, is suitable for more complete functions of the flight control system, and can match the work of bionic leg landing gear, The level of intelligence is further improved;

(3) The present invention can improve the intelligence level and operational capability of the unmanned vertical take-off and landing aircraft, improve the reliability and safety of autonomous take-off and landing, improve the intelligent technology level of the aircraft, break the limitation of application scenarios, and make the slope surface It is possible to take off and land autonomously on complex terrain such as , stepped ground, etc.

(4) The present invention can be applied not only to vertical take-off and landing aircraft, but also to intelligent unmanned vehicles, planetary detection landers and other fields, providing a better control system and having the advantage of scalable applications.

The multi-system fusion control system of flight control-landing gear control-terrain recognition system completes the process of data sharing, signal transmission and collaborative operation based on the flight control system, legged landing gear control system and terrain recognition system, effectively completing the multi-system The integrated design reduces the shortcomings of insufficient data sharing, high system failure rate and large design space in the original multi-system serial operation process, forming an unmanned vertical take-off and landing aircraft that can be applied to the installation of legged landing gear The dedicated multi-function flight control system has high practical value.

Claims (8)

1. a flight control-landing gear control-terrain recognition multi-system fusion control system, is characterized in that, comprises: flight control system, bionic leg landing gear control system, terrain recognition system, data transmission system and ground system; Among them, the flight control system determines the realization and control performance of the autonomous flight process of the unmanned vertical take-off and landing vehicle; the bionic legged landing gear control system is the control system of the bionic legged landing gear, which determines the attitude and movement process of the landing gear. Cooperate with the flight control system to complete the self-adaptive take-off and landing of the unmanned vertical take-off and landing aircraft; the terrain recognition system is to achieve a large-scale selection of landing terrain before the unmanned vertical take-off and landing aircraft lands, accurately identify the bottom surface, and determine the landing area. The landing range realizes real-time online tracking of the landing site, and informs the flight control system and the bionic legged landing gear control system of terrain information, so as to realize the real-time navigation of the flight control system and the attitude pre-swing and adjustment of the bionic legged landing gear; The data information of the flight control system, the leg landing gear control system and the terrain recognition system are all communicated with the ground system through the data transmission system to realize the ground monitoring of the data and status information. 2. system according to claim 1, is characterized in that, flight control system comprises: flight control computer, sensor mechanism and executive mechanism; The transmitted real-time status information and the transmission control commands generated by itself are transmitted to the actuator; the sensor mechanism collects the altitude information, attitude information, acceleration information and other flight data of the unmanned vertical take-off and landing vehicle, and transmits it to the flight controller The computer performs the calculation; the function of the actuator is to implement the instruction execution of the flight control computer and complete the autonomous flight of the unmanned vertical take-off and landing aircraft. 3 . The system according to claim 2 , wherein the flight control system requires an onboard power supply to supply power to the flight control-landing gear control-terrain recognition multi-system fusion control system. 4 . 4. The system according to claim 2, wherein the bionic leg landing gear control system comprises: a controller, a motor driver, a modular joint unit and a load sensor; Among them, the controller realizes the control algorithm design of the bionic leg landing gear, completes the position and attitude calculation and attitude control, and processes the state information fed back by the motor driver and the load sensor; the motor driver receives the signal from the controller and completes the solution of the motor control signal. The closed-loop control of the modular joint unit completes the action of the bionic leg; the modular joint unit is used as the driving joint of the bionic leg; the load sensor is a sensor installed at the foot end, which is used to measure the force signal feedback after the bionic leg landing gear touches the ground. This is transmitted to the controller for post-landing adaptive control. 5. The system according to claim 4, wherein the terrain recognition system comprises: lidar, visual camera, inertial measurement unit (IMU) inertial navigation and high-performance onboard processor; The high-performance onboard processor is used to: perform joint spatiotemporal calibration of lidar and visual camera, realize fusion of laser point cloud and image pixel information, and perform real-time online low-altitude UAV based on the fusion information and IMU inertial navigation information. Terrain modeling, according to the modeled 3D terrain data, accurately identify the terrain information, automatically identify the landing point according to the terrain information and realize the real-time online tracking of the landing point, calculate the attitude of the UAV relative to the landing point in real time and feed it to the flight controller system. 6. The system according to claim 1 is characterized in that, the flight control system and the bionic leg landing gear control system carry out communication and data exchange, the terrain recognition system and the bionic leg landing gear control system carry out communication and data exchange, and the flight control system and the bionic leg landing gear control system carry out communication and data exchange. There is no direct communication between the control system and the terrain recognition system. 7. system according to claim 6, is characterized in that, flight control computer comprises DSP and programmable gate array FPGA; Sensor mechanism comprises another inertial measurement unit IMU, GPS navigation system, compass, altimeter; Actuator comprises engine, steering gear. 8. The system of claim 5, wherein the modular joint unit includes a brake, a servo motor, and an encoder.

Images