Composite aircraft detection system for industrial flaw detection

Abstract

The invention discloses a composite aircraft detection system for industrial flaw detection, the composite aircraft detection system comprises a composite aircraft and a ground station, the composite aircraft is used for receiving the control of the ground station, carrying out close-range reconnaissance by flying against an industrial building, and wirelessly sending a collected building image to the ground station; wherein the ground station is used for wirelessly transmitting a control signal to the composite aircraft, receiving a building image from the composite aircraft, and analyzing the building image to obtain a damage result of the industrial building. Aiming at the problems of intensive labor, lack of a multi-scene effective detection scheme and the like in industrial flaw detection, the invention organically combines a composite aircraft technology, an image processing technology and various detection requirements, realizes high universality and high intelligence of industrial detection, and has a wide application prospect.

Description

Technical Field

[0001] This application relates to the field of aircraft control technology, specifically to a composite aircraft detection system for industrial flaw detection. Background Technology

[0002] Currently, the inspection and flaw detection of industrial structures such as bridges, tunnels, and dams mainly rely on manual inspections or large-scale machinery-assisted operations. This not only requires a large investment of manpower and resources but also results in low inspection efficiency and high time costs. Although intelligent inspection equipment such as pier structure scanning and inspection robots and power grid maintenance drones have emerged, they generally suffer from the following technical shortcomings:

[0003] 1. Limited applicability: Existing equipment is usually designed for specific types of industrial buildings, making it difficult to adapt to the diverse testing needs of bridges, tunnels, dams, glass curtain walls, etc.

[0004] 2. Insufficient functional expandability: Most devices adopt a fixed function design and lack modular configuration capabilities, making it impossible to quickly adjust the detection modules according to actual detection tasks;

[0005] 3. Limited adhesion and mobility: For complex building surfaces such as vertical walls and horizontal ceilings, existing equipment has difficulty achieving stable adhesion and flexible movement, affecting the detection coverage.

[0006] In summary, existing technologies suffer from low levels of intelligence, generalization, and automation. Summary of the Invention

[0007] To address the aforementioned technical bottlenecks, this application provides a composite aircraft detection system for industrial flaw detection.

[0008] This application provides a composite aircraft detection system for industrial flaw detection, comprising a composite aircraft and a ground station; the composite aircraft is used to receive control from the ground station, fly to an industrial building for close-range reconnaissance, and wirelessly transmit the acquired building images to the ground station; the ground station is used to wirelessly transmit control signals to the composite aircraft and receive the building images from the composite aircraft, and obtain the damage results of the industrial building through analysis of the building images.

[0009] Furthermore, the composite aircraft includes a wireless receiving module, a flight control module, a sensing and acquisition module, a camera module, an image transmitting module, a motion module, and a power module. The wireless receiving module receives control signals from the ground station and converts these signals into flight control signals acceptable to the flight control module. The sensing and acquisition module senses the composite aircraft's flight attitude through sensors and transmits the acquired attitude information to the flight control module. The flight control module generates flight control commands based on the flight control signals and the attitude information, and generates detection control commands upon approaching an industrial building. The motion module is signal-connected to the flight control module and uses... The flight attitude of the composite aircraft is adjusted based on the flight control commands; the flight attitude of the composite aircraft includes cruise flight, overpass flight, and climbing flight; the camera module is signal-connected to the flight control module and is used to capture images of the walls and roof of the industrial building based on the detection control commands to obtain building images; the image transmission module is used to acquire the building images from the camera module and transmit the building images to the ground station wirelessly; the power module is used to provide the DC power required for the operation of the wireless receiving module, the flight control module, the sensing and acquisition module, the camera module, the image transmission module, and the motion module.

[0010] Furthermore, when the flight control module generates flight control commands based on the flight control signals and the attitude information, the specific steps include: the flight control module activates a corresponding flight mode based on the flight control signals, and generates the flight control commands in combination with the attitude information under the corresponding flight mode; when the flight mode is configured as cruise mode, the generated flight control commands enable the composite aircraft to have free flight capability and to fly to or away from industrial buildings; when the flight mode is configured as climbing mode, the generated flight control commands enable the composite aircraft to attach to the wall of the industrial building and travel along the wall while attached to the wall; when the flight mode is configured as eaves mode, the generated flight control commands enable the composite aircraft to attach to the top surface of the industrial building and travel along the top surface while attached to the top surface.

[0011] Furthermore, the motion module includes four sets of ducted fans and four sets of friction wheels driven by geared motors; the four sets of ducted fans are used to provide lift for vertical take-off and landing and power for flight, as well as to adjust the flight attitude through differential steering; the four sets of friction wheels driven by geared motors are used to provide power for traveling along the wall and power for traveling along the top surface.

[0012] Furthermore, the ground station includes a wireless transmission module, a dual-channel receiving module, an image acquisition module, and an image processing module; the wireless transmission module acquires the control command and wirelessly transmits the control command to the composite aircraft; it generates the control command during user interaction; the dual-channel receiving module receives the building image wirelessly transmitted by the composite aircraft; the image acquisition module performs preliminary processing on the building image, including image sampling and format conversion, and transmits the pre-processed building image to the image processing module; the image processing module performs algorithmic analysis on the building image to obtain damage results, including damage location, damage type, and / or damage degree; the damage results are displayed to the user via a display.

[0013] Furthermore, the dual-channel receiving module has two RX5808 modules, both of which are used to perform the receiving and demodulation of the analog video signal corresponding to the building image through the automatic frequency sweep function, and to output a digital video signal in USB bus format to the image acquisition module.

[0014] Furthermore, the image processing module includes a model configuration submodule and an algorithm core submodule. The model configuration submodule is used to flexibly select the architecture and weight parameters of the network model according to the corresponding building type in the building image, so as to enhance the network model's ability to adapt to the detection scene. The model configuration submodule is also used to train a lightweight YOLO-S-based model according to the selected network model's architecture and weight parameters. The algorithm core submodule is used to extract image features from the building image using the lightweight YOLO-S-based model, accurately determine the damage area based on the image features, and further analyze to obtain the damage location, damage type, and / or damage degree.

[0015] Furthermore, the model configuration submodule and the algorithm core submodule are integrated into a pre-defined industrial inspection software. This industrial inspection software provides a weight selection window, a video source selection window, a network download window, and a processing result display window via a display. The weight selection window presents options to the user for flexibly selecting the weight parameters corresponding to the network module. The video source selection window displays the acquisition channels for the building images, including real-time video from the composite aircraft and historical video from local storage. The network download window provides services for downloading the network model architecture, weight parameters, and updating the industrial inspection software. The processing result display window displays the building images and information about the corresponding damage location, damage type, and / or damage extent.

[0016] Furthermore, the ground station also includes a remote controller, which serves as a carrier for users to perform human-computer interaction operations and has operation areas corresponding to setting flight missions, switching flight modes, and controlling flight attitude.

[0017] Furthermore, the remote controller includes a mode control module, which is used to perform passive flight mode switching and automatic flight mode switching. When the mode control module is configured for passive flight mode switching, the user can switch between cruise mode, climb mode, and eaves mode through button operation or touch operation in the corresponding operation area. When the mode control module is configured for automatic flight mode switching, it can automatically switch from cruise mode to climb mode or eaves mode by autonomously judging the flight attitude of the composite aircraft, and automatically return to cruise mode and enter the passive flight mode switching operation state after detecting factors that hinder the transition, so that the user can intervene and control the composite aircraft in a timely manner through manual operation.

[0018] The beneficial effects of this invention are that, in response to the massive demand for industrial facility and building inspection in my country, it provides a composite aircraft detection system. Through the information interaction and cooperation between the composite aircraft and the ground station, it can realize the acquisition of building images and damage analysis of industrial buildings. Furthermore, by utilizing the organic combination of multi-flight mode switching and modular design, it effectively solves the technical bottlenecks of insufficient versatility and low automation of existing detection technologies. It can replace manual labor in completing inspection operations in high-risk environments, eliminating the risk of personnel injury or death, and providing an intelligent and universal solution for industrial facility inspection. It has significant technological advancements and industrial application value. Attached Figure Description

[0019] Figure 1 This is a general module diagram of a composite aircraft system for industrial flaw detection according to this application;

[0020] Figure 2 This is a schematic diagram of the receiver PCB board and image processing module in one embodiment of this application;

[0021] Figure 3 This is a schematic diagram of the control steps of the composite aircraft system in the flaw detection method of one embodiment of this application;

[0022] Figure 4 This is a screenshot of the interface of industrial testing software in one embodiment of this application;

[0023] Figure 5 This is a schematic diagram of the flying eaves mode switching process in a flaw detection method in one embodiment of this application;

[0024] Figure 6 This is a schematic diagram of the climbing mode switching process in a flaw detection method according to one embodiment of this application. Detailed Implementation

[0025] The present application will now be described in further detail with reference to the accompanying drawings.

[0026] A composite aircraft detection system for industrial flaw detection, such as Figure 1 As shown, the system comprises two parts: a composite aircraft and a ground station. The composite aircraft is responsible for close-range reconnaissance, image acquisition, and signal transmission of industrial buildings; the ground station is responsible for controlling the composite aircraft, receiving images, and processing them. Specifically, the composite aircraft is controlled by the ground station to fly to the industrial building for close-range reconnaissance and to wirelessly transmit the acquired building images to the ground station; the ground station is used to wirelessly transmit control signals to the composite aircraft and to receive building images from the composite aircraft, and to obtain damage results for the industrial building through analysis of the building images.

[0027] It should be noted that the composite aircraft detection system has multiple flight modes, including cruise mode, climbing mode, and eaves mode, facilitating close-range attachment to the interior and exterior surfaces of buildings for close-range, high-definition, real-time image acquisition. In cruise mode, the composite aircraft can fly freely in space under the user's control; in climbing mode, the composite aircraft can attach to the vertical interior and exterior walls of buildings such as tunnel walls, factory walls, dam surfaces, and glass buildings, allowing for free movement and image acquisition; in eaves mode, the composite aircraft can attach to the horizontal interior and exterior walls of buildings such as ceilings and viaducts, allowing for free movement and image acquisition.

[0028] In this embodiment, as Figure 1 As shown, the hardware of the composite aircraft includes a flight control module 107, a wireless receiving module 105, a sensing and acquisition module 106, a motion module 108, a camera module 110, a power supply module 111, and an image transmitting module 109.

[0029] The wireless receiving module 105 receives control signals from the ground station and converts these signals into flight control signals that the flight control module 107 can accept. The sensing and acquisition module 106 senses the flight attitude of the composite aircraft through sensors and transmits the acquired attitude information to the flight control module 107. The flight control module 107 generates flight control commands based on the flight control signals and attitude information, and generates detection control commands when the aircraft approaches an industrial building. The motion module 108 is signal-connected to the flight control module 107 and is used to adjust the flight attitude of the composite aircraft based on the flight control commands. The flight attitudes of the combined aircraft include cruise flight, overpass flight, and climbing flight; the camera module 110 is signal-connected to the flight control module 107 and is used to capture images of the walls and roof of the industrial building based on detection and control commands to obtain building images; the image transmission module 109 is used to acquire building images from the camera module and transmit the building images to the ground station wirelessly; the power module 111 is used to provide the DC power required for operation of the wireless receiving module 105, the flight control module 107, the sensing and acquisition module 106, the camera module 110, the image transmission module 109, and the motion module 108.

[0030] The core component of the hybrid aircraft is the flight control module 107. The flight control module 107 processes the signal inputs from the sensing and acquisition module 106 and the wireless receiving module 105, is responsible for calculating the hybrid aircraft's flight attitude and position, and responds to control signals issued by the user. For example, the flight control module 107 can use an STM32H7 chip as its main control chip.

[0031] Specifically, the flight control module 107 can combine the position and distance parameters (i.e., the acquired attitude information) provided by the sensing and acquisition module 106, use the extended Kalman filter algorithm to calculate the precise position and attitude parameters of the composite aircraft, and use the ADRC active disturbance rejection algorithm as the control algorithm to control the motion module 108, achieving fast and precise flight attitude control. The flight control module 107 is also responsible for outputting signals to control the motion module 108 and the camera module 110.

[0032] In one specific embodiment, when the flight control module 107 generates flight control commands based on flight control signals and attitude information, the specific operation process includes the following: The flight control module 107 activates the corresponding flight mode based on the flight control signals, and generates flight control commands in combination with attitude information under the corresponding flight mode. When the flight mode is configured as cruise mode, the generated flight control commands enable the composite aircraft to have free flight capability and to fly to or away from industrial buildings; when the flight mode is configured as climbing mode, the generated flight control commands enable the composite aircraft to attach to the wall of the industrial building and travel along the wall while attached to the wall; when the flight mode is configured as eaves mode, the generated flight control commands enable the composite aircraft to attach to the roof of the industrial building and travel along the roof while attached to the roof.

[0033] The wireless receiver module 105 is fixed to the tail of the hybrid aircraft and receives radio signals at a frequency of 2.4 GHz, which are control signals from the ground station. After processing, the signals are transmitted to the flight control module 107 to control the hybrid aircraft. For example, the wireless receiver module 105 can use an R9DS receiver to demodulate the PPM signal transmitted by the remote controller.

[0034] The sensing and acquisition module 106 of the composite aircraft is signal-connected to the flight control module 107. The sensing and acquisition module 106 consists of a series of sensors that provide a basis for attitude calculation and position determination of the composite aircraft. These sensors may include a six-axis attitude sensor, a positioning system, a barometer sensor, and a TOF lidar module.

[0035] A six-axis attitude sensor is used to acquire three-axis acceleration and angle; the output parameters can be expressed as (a x ,a y ,a z ,pitch,yaw,roll), where pitch, yaw, and roll are the pitch, yaw, and roll angles of the compound aircraft, respectively.

[0036] The positioning system is used to acquire location information. The composite aircraft is equipped with two positioning systems: one is a geomagnetic positioning module welded to the plate, and the other is an external GPS compass module. The flight control module 107 reads data from these two positioning systems and uses an extended Kalman filter algorithm to continuously correct the data from the on-board geomagnetic positioning module using data from the external GPS compass module, thereby obtaining accurate positioning information. The geomagnetic positioning module relies on the Earth's magnetism to operate. Given the differences in geomagnetic characteristics at different geographical locations, it needs to be calibrated before each use to ensure positioning accuracy.

[0037] When the barometer sensor does not receive a GPS positioning signal or the GPS signal strength is below a certain threshold, the flight control module 107 will automatically change the position calculation logic of the composite aircraft and determine the altitude of the composite aircraft by reading the data from the barometer sensor.

[0038] The TOF lidar module is mounted on top of the hybrid aircraft and uses an 850nm infrared light source to achieve high-precision distance measurement. When the hybrid aircraft switches from cruise mode to climbing mode or overhang mode, it needs to measure the distance from itself to the surface of the industrial building to assist the aircraft in adjusting its attitude.

[0039] The camera module 110 includes a 1080P high-definition camera module and its dedicated hardware driving circuit. The camera module 110 is fixedly mounted to the belly of the composite aircraft via a shock-absorbing bracket, and is used to perform high-definition image acquisition tasks in an attached state. The camera module 110 is electrically connected to the image transmission module 109 via a flexible data transmission line, enabling real-time encoding and wireless transmission of the acquired images.

[0040] The camera module 110 has significant technical advantages: (1) It has strong multimodal adaptability and supports modular replacement design. It can be quickly replaced with functional modules such as infrared thermal imaging cameras and multispectral cameras according to the needs of the inspection task, meeting the diverse needs of industrial flaw detection, security monitoring, environmental monitoring and other scenarios. (2) It has good structural compatibility and adopts standardized interface protocol. The replacement process does not require modification of the overall structure of the aircraft, which significantly improves the efficiency of operation. (3) It has the characteristics of attachment acquisition. With the aircraft climbing power system, it can achieve high-precision defect positioning imaging within the attachment distance range of 0.5-10cm, effectively solving the technical problem of limited line of sight of traditional inspection equipment.

[0041] The power module 111 is responsible for providing the necessary electrical energy to the composite aircraft. For example, the power module 111 consists of a lithium polymer battery pack, a 12V voltage regulator module, three 5V voltage regulator modules, and a power management unit. The 12V DC / DC voltage regulator module powers the motion module 108, and the three 5V DC / DC voltage regulator modules power the flight control module 107, the camera module 110, and the sensing and acquisition module 106, respectively.

[0042] The motion module 108 provides power for the movement of the composite aircraft, including four ducted fans and four friction wheels driven by geared motors. The four ducted fans provide lift for vertical takeoff and landing and propulsion for flight, as well as adjusting flight attitude through differential steering. The four friction wheels driven by geared motors provide propulsion for movement along walls and rooftops. Specifically, the motion module 108 includes a cruise propulsion system consisting of four ducted fans and a climbing propulsion system consisting of four geared motors and their driven friction wheels. The user can independently control each friction wheel, enabling the composite aircraft to move freely in four directions (forward, backward, left, and right) while in a attached state through differential steering. The geared motors increase torque, enhancing the composite aircraft's ability to pass over industrial building surfaces and rooftops. By default, and in cruise mode, the friction wheels are locked; switching to climbing or overhanging mode unlocks them. The remote controller's two channels control the composite aircraft's movement relative to the user in the forward, backward, left, and right directions while in an attached state.

[0043] In this embodiment, as Figure 1 As shown, the ground station includes a wireless transmission module 104, a dual-channel receiving module 101, an image acquisition module 102, and an image processing module 103. The dual-channel receiving module 101, the image acquisition module 102, and the image processing module 103 together realize the reception, demodulation, format conversion, and image processing of the analog video signals transmitted by the composite aircraft.

[0044] The wireless transmission module 104 acquires control commands and wirelessly transmits them to the composite aircraft; it also generates control commands during user-computer interaction. The dual-channel receiving module 101 receives building images wirelessly transmitted by the composite aircraft; the image acquisition module 102 performs preliminary processing on the building images, including image sampling and format conversion, and transmits the pre-processed images to the image processing module 103; the image processing module 103 performs algorithmic analysis on the building images to obtain damage results, including damage location, damage type, and / or damage degree; these damage results are then displayed to the user via a monitor.

[0045] In one specific embodiment, the dual-channel receiving module 101 has two RX5808 modules, both used to perform the task of receiving and demodulating the analog video signal corresponding to the building image through automatic frequency sweeping, and to output a digital video signal in USB bus format to the image acquisition module 102. For example Figure 2The dual-channel receiving module 101 and the image acquisition module 102 are integrated on a single receiver PCB board. The PCB board includes two RX5808 modules, namely RX5808 module 21 and RX5808 module 22, for receiving and demodulating analog video signals. The PCB board may also include an MCU chip 24 for controlling the receiver PCB board, performing operations such as frequency sweeping, signal selection, and screen refresh; a switch and buzzer for human-machine interaction, allowing users to manually select frequencies, sweep frequencies, view spectrum strength, and manually select output channels, while the buzzer emits a prompt tone. The PCB board may also include an OLED screen 25, which communicates with the MCU chip 24 using the IIC protocol to display information related to the receiver PCB's operation. The PCB board may also include an analog electronic switch 26, whose on / off state is controlled by the MCU chip 24, defaulting to selecting the stronger signal of the two receiving channels. The PCB board may also include an image acquisition card chip 27, responsible for sampling and format conversion of the demodulated analog video signal, converting it to a USB format signal. The PCB board may also include an EEPROM chip 28, which is responsible for storing the firmware required by the image acquisition card chip 27.

[0046] like Figure 2 As shown, RX5808 modules 21 and 22 are connected to MCU chip 24, and the switch and buzzer 23 and screen 25 are also connected to MCU chip 24. MCU chip 24 is responsible for controlling analog electronic switch 26 to select channels. Analog electronic switch 26 outputs the selected channel from its two input signals to image acquisition card chip 27. EEPROM chip 28 of image acquisition module 102 is connected to image acquisition card chip 27 and is used to store the firmware required for the operation of image acquisition card chip 27. After sampling and format conversion of the input analog signal, image acquisition card chip 27 inputs the signal to image processing module 103. Image processing algorithm 29 in image processing module 103 performs real-time image processing and outputs the final detection result. This image processing algorithm 29 can be regarded as a functional module with specific data processing and calculation functions.

[0047] The screen 25 can be a 1.3-inch OLED screen with a resolution of 128×64, communicating with the MCU chip 24 via the IIC protocol. The MCU chip 24 uses four GPIO ports to connect to the OLED screen, namely VCC, GND, SCL, and SDA. Under the control of the MCU chip 24, the screen 25 can refresh in real time and display information such as the signal strength of the two channels, the selected channel, and the current receiving frequency.

[0048] The switch and buzzer 23 can be connected to the GPIO port of the MCU chip 24. Users can use this switch to perform operations such as automatic frequency sweeping and manual frequency selection; the buzzer will emit a corresponding prompt tone during operation. Simultaneously, the operation results will be displayed in real time on the screen 25 while the user is performing the operation.

[0049] Both RX5808 modules 21 and RX5808 modules 22 described above have the capability to receive video signals in the 5.8GHz band. The RX5808 receiver module has 15 pins, with the following functions: Pin 13 connects to the antenna for signal reception; Pin 7 is the analog audio signal output pin; Pin 8 is the analog video signal output pin; Pins 1, 2, and 3 use the SPI protocol for communication, allowing the receiving frequency to be written to the RX5808 receiver module externally; Pin 6 is the RSSI pin, and its level reflects the signal strength of the receiving frequency.

[0050] In one specific embodiment, Figure 2 The processor MCU chip 24 consists of an STM32F411 chip. This chip uses two GPIO ports to read the RSSI pin levels of RX5808 modules 21 and 22 to determine the signal strength, and selects the channel at the selection pin connected to the analog electronic switch 26 based on the determination result; at the same time, it uses three GPIO ports to connect to pins 1, 2 and 3 of the two RX5808 receiver modules to write the receiving frequency point via the SPI protocol.

[0051] In one specific embodiment, the receiver automatically initiates a frequency sweep program upon power-up. The 5.8GHz band is divided into six sub-bands: A, B, C, D, E, and F. Each sub-band contains eight frequency points, totaling 48 frequency points. The frequency sweep process automatically begins when the user connects the receiver PCB to the host computer using a USB cable. The system filters out the band with the highest energy from these six bands and further selects the frequency with the highest energy from that band as the frequency for receiving the signal. After the frequency sweep is completed, the MCU chip 24 compares the energy of the signals received by the two RX5808 receiver modules and then automatically selects the signal with the higher energy from the analog electronic switch 26 to input to the image acquisition card chip 27. Alternatively, the user can manually select the receiving frequency, manually select the analog electronic switch 26, and manually perform the frequency sweep operation to meet the needs of different scenarios.

[0052] In one specific embodiment, Figure 2The analog electronic switch 26 in the chip is composed of a CD4066 analog electronic switch chip. The CD4066 analog electronic switch chip has 14 pins, supporting four channels of analog signal selection. The specific pin functions are as follows: eight input / output pins are used for analog signal input or output; four control pins are used to control the opening or closing of the corresponding channel; one VDD power supply pin provides power to the chip; and one VSS ground pin is used for grounding. Specifically, pins 1 and 2 are the input / output pins for channel A signals, regardless of input / output direction; pins 3 and 4 are the input / output pins for channel B signals, also regardless of input / output direction; pin 5 is the control pin for channel B signals, and pin 13 is the control pin for channel A signals. The control logic of the CD4066 analog electronic switch chip is as follows: when the control pin input is high, the corresponding channel is turned on; when the input is low, the channel is turned off.

[0053] On the receiver PCB, pins 1 and 2 of the CD4066 analog electronic switch chip are connected to pins 8 of RX5808 module 21 and RX5808 module 22, respectively, to receive analog video signals output from the two RX5808 receiver modules in analog CVBS format. Pin 13 of the CD4066 analog electronic switch chip is connected to a GPIO port of MCU chip 24 for controlling the selection of the A-channel signal. When the RSSI pin output level of RX5808 module 21 is greater than the RSSI pin output level of RX5808 module 22, MCU chip 24 outputs a high level on the GPIO port connected to pin 13, selecting the analog video signal of RX5808 module 21 as output; when the RSSI pin output level of RX5808 module 21 is less than the RSSI pin output level of RX5808 module 22, MCU chip 24 outputs a low level on the GPIO port connected to pin 13, selecting the analog video signal of RX5808 module 22 as output. In addition, the RX5808 module also has the function of receiving analog audio signals. The analog audio signals are also input to the CD4066 analog electronic switch chip, and its gating logic is the same as that of analog video signals.

[0054] Under the control of MCU chip 24, CD4066 analog electronic switch chip selects the signal with higher energy from the two RX5808 receiving modules for output. This signal is then converted from analog to digital and processed in subsequent image processing.

[0055] like Figure 2As shown, the output signal of the analog electronic switch 26 is input to the image acquisition card chip 27. In one embodiment, the image acquisition card chip 27 consists of an MS2107 chip, and an EEPROM chip 28 is connected to the image acquisition card chip 27. The EEPROM chip 28 consists of an 8-bit AT24C16C chip with a speed of 16Kbit. The firmware required for the operation of the MS2107 chip is stored in the EEPROM chip 28 and needs to be programmed before use. The image acquisition card chip 27 performs ADC sampling and format conversion on the analog video signal, converting it into a digital video signal in USB bus format before inputting it into the computer for processing. The image acquisition card chip 27 is also deployed on the receiver PCB board, and the processed signal is input to the image processing module 103 via a USB connection cable.

[0056] In one specific embodiment, the image processing module 103 uses image processing algorithm 29 to implement image processing functions. Image processing algorithm 29 is fine-tuned based on the YOLO-S lightweight model. This algorithm has the ability to identify targets such as cracks in images in real time, and can accurately measure the length, width, and angle of the identified targets. For example, the image processing module 103 includes a model configuration submodule and an algorithm core submodule (…). Figure 2 (Illustrated diagram in the image); The model configuration submodule is used to flexibly select the architecture and weight parameters of the network model according to the corresponding building type in the building image, so as to enhance the network model's ability to adapt to the detection scene; The model configuration submodule is also used to train a lightweight YOLO-S model based on the selected network model architecture and weight parameters; The algorithm core submodule is used to extract image features from the building image using the lightweight YOLO-S model, accurately determine the damage area based on the image features, and further analyze to obtain the damage location, damage type and / or damage degree.

[0057] In one specific embodiment, the model configuration submodule and algorithm core submodule in the image processing module 103 are integrated into a preset industrial inspection software, such as... Figure 4 As shown, this industrial testing software provides a weight selection window, a video source selection window, a network download window, and a processing result display window via a monitor. Figure 4 The weight selection window displays options for users to flexibly choose the weight parameters corresponding to the network modules; the video source selection window displays the building image acquisition channels for users to choose from, including real-time video from the composite aircraft and historical video from local storage; the network download window provides services for downloading the network model architecture, weight parameters, and updates to the industrial inspection software; and the processing result display window displays building images and corresponding information on damage location, damage type, and / or damage extent.

[0058] Figure 4 This is an industrial inspection software that runs on Windows. The software is encapsulated using PYQT based on image processing algorithm 29 and can be deployed on PCs or embedded devices. Users can directly obtain inspection results through a graphical interface. The industrial inspection software has the following features: (1) Easy and flexible configuration. Users can flexibly choose the network architecture and weight configuration according to actual inspection needs to adapt to diverse industrial flaw detection scenarios and task objectives. (2) Support for software encapsulation and related functions. The industrial inspection software is developed based on the PyQt library and integrates functions such as real-time video reading, local video file import, and inspection result download. QApplication is used as the core control of the application, and QMainWindow is used as the main window framework to integrate functional components. QWidget is combined with the OpenCV library to achieve accurate capture and display of real-time video streams. QFileDialog component is used to achieve convenient import of local video files. QPushButton component is used in conjunction with QFileDialog to achieve saving of inspection results in the format and path specified by the user. (3) Support for operating environment and updates. The industrial inspection software has good compatibility and can run directly on the Windows operating system.

[0059] Figure 4 The industrial inspection software features a user-friendly graphical interface, including a processing result display window for intuitively showing the processing results of images or data; a weight selection window for users to flexibly select appropriate weights according to actual needs; a video source selection window that allows users to choose different video sources such as real-time video and local video files; and a network download window for software updates, model downloads, and weight downloads. This graphical interface design greatly lowers the barrier to entry for the software, allowing even users without professional technical knowledge to easily get started. Furthermore, the software supports network updates, enabling users to obtain the latest versions of the software, models, and weights in a timely manner through network downloads, ensuring that the software always maintains optimal performance and state-of-the-art inspection capabilities.

[0060] It should be noted that the composite aircraft detection system disclosed in this invention has some significant technical advantages: (1) Strong adaptability to multiple scenarios: The composite aircraft has cruise mode, eaves mode and climbing mode, and can achieve stable attachment and free movement on complex surfaces such as the inner and outer walls of vertical buildings and the top of horizontal buildings. It is suitable for various industrial building inspection scenarios such as bridges, tunnels, dams, and glass buildings. (2) Excellent modular function expansion capability: It adopts a modular design concept and supports quick replacement of detection modules (such as flaw detection sensors, high-definition cameras, etc.), which can flexibly adapt to different inspection task requirements. (3) High automation performance: It integrates visual navigation and autonomous control algorithms, and combines multi-sensor data fusion technology to realize automated functions such as detection path planning and obstacle avoidance, which significantly reduces the need for manual intervention. It can be understood that the composite aircraft detection system disclosed in this invention provides a universal and efficient solution for industrial building inspection through multi-mode switching, modular design and intelligent control, which effectively breaks through the application limitations of existing technologies.

[0061] Based on the composite aircraft detection system described above, this application also discloses a corresponding control method. This control method consists of a series of steps and algorithms, including user control steps for the composite aircraft system and attitude control algorithms for switching flight modes. The aforementioned flaw detection method, combined with the composite aircraft system, can achieve tasks such as attachment reconnaissance of industrial buildings, image transmission, and signal processing.

[0062] like Figure 3 As shown, a corresponding control method is provided for a composite aircraft detection system used for industrial flaw detection.

[0063] Step S31: Selection of Visual Model Type and Weights. Users must first select the type and weights of the large visual model based on different building materials and actual usage scenarios within the industrial inspection software. This operation aims to enable the system to adapt to diverse inspection needs and ensure the accuracy and reliability of the inspection results.

[0064] Step S32: Calibration of the composite aircraft and activation of the image transmission system. The composite aircraft's sensing and acquisition module 106 includes a magnetic compass. Given the varying geomagnetic characteristics of different geographical locations, it must be calibrated before each use of the composite aircraft. Simultaneously, the user must activate the composite aircraft's image transmission system to ensure that the onboard camera module 110 and image transmission module 109 are functioning correctly. After completing these operations, the user can receive real-time video signals from the composite aircraft at the ground station.

[0065] Step S33: The composite aircraft enters cruise mode. The user controls the composite aircraft to take off via remote control, putting it into cruise mode. In this mode, the user has full control over the composite aircraft and can freely control its flight attitude. When the composite aircraft approaches the location to be detected, the user can switch its flight mode via remote control.

[0066] Step S34: When the composite aircraft approaches the detection location, the user can switch the flight mode via remote control to enter the climbing / flying mode.

[0067] If you choose to enter the climbing mode, the composite aircraft will perform the attachment action on the vertical building surface; if you choose to enter the eaves mode, the composite aircraft will perform the attachment action on the horizontal building surface.

[0068] After the attachment operation is completed, the composite aircraft automatically unlocks its climbing propulsion system and then initiates industrial flaw detection. During the inspection process, the user can precisely control the composite aircraft to move freely on the surface to be inspected via remote control, meeting the detailed inspection needs of different inspection areas and ensuring the comprehensiveness and accuracy of the inspection work.

[0069] In climbing mode or flying mode, users can exit the detection mode at any time via remote control according to actual detection needs and put the composite aircraft back into cruise mode S33 to flexibly cope with different detection tasks and scene changes.

[0070] like Figure 5 As shown, in one specific embodiment, the composite aircraft detection system switches from cruise mode to eaves mode according to a specific set of execution logic.

[0071] Step S51: The user uses the remote control to switch the flight mode to the eaves mode, and the composite aircraft enters the eaves mode switching process.

[0072] In step S52, the compound aircraft will automatically push the throttle to 60% and maintain this position for 5 seconds. After 5 seconds, it will be determined whether the distance between the compound aircraft and the top is less than 1 meter. If so, proceed to step S53; otherwise, proceed to step S54.

[0073] In step S53, when the composite aircraft is less than 1 meter away from the top, the composite aircraft will continue to increase the throttle to 80% and maintain it for 3 seconds.

[0074] In step S54, when the distance between the composite aircraft and the top is greater than or equal to 1 meter, the composite aircraft reduces the throttle to 50%, switches the flight mode back to cruise mode, and the user regains full control of the composite aircraft.

[0075] After step S53, wait 3 seconds. If the distance between the composite aircraft and the top is less than 0.2 meters and the distance no longer changes, it is considered a successful overhang, and proceed to step S55. At this time, the composite aircraft will unlock the climbing power system. If it is considered an unsuccessful overhang, proceed to step S54. At this time, the throttle is reduced to 50%, the flight mode is switched back to cruise mode, and the user regains full control of the composite aircraft.

[0076] The above process enables the hybrid aircraft to switch from cruise mode to overflight mode. During the switching process, the hybrid aircraft will automatically complete the flight mode transition without user intervention. If the hybrid aircraft detects that the flight mode transition cannot be completed normally, it will promptly return to cruise mode and return control to the user to prevent the hybrid aircraft from going out of control.

[0077] like Figure 6 As shown, in one specific embodiment, the hybrid aircraft also follows a specific set of execution logic when switching from cruise mode to climb mode.

[0078] Step S61: The user uses the remote controller to switch the flight mode to the climbing mode, and the composite aircraft enters the climbing mode switching process.

[0079] In step S62, the composite aircraft controls its cruise propulsion system to roll forward at a pitch angle of 60°; simultaneously, it automatically increases the throttle to 60% for 3 seconds. After 3 seconds, if the aircraft is more than 1 meter away from the side wall, proceed to step S64; if the aircraft is less than 1 meter away from the side wall, proceed to step S63.

[0080] In step S63, when the distance between the compound aircraft and the side wall is less than 1 meter, the compound aircraft will maintain the pitch angle and push the throttle up to 100% for 2 seconds.

[0081] In step S64, when the distance between the compound aircraft and the side wall is greater than or equal to 1 meter, the compound aircraft automatically pushes the throttle to 50%, restores the pitch angle to 0°, switches the flight mode back to cruise mode, and the user regains full control of the compound aircraft.

[0082] After step S63, wait 2 seconds. If the distance between the composite aircraft and the side wall is less than 0.2 meters and the distance no longer changes, the climb is considered complete, and proceed to step S65. At this time, the composite aircraft will unlock the climbing power system. If the climb is not considered complete, proceed to step S64. At this time, the throttle is reduced to 50%, the pitch angle is restored to 0°, the flight mode is switched back to cruise mode, and the user regains full control of the composite aircraft.

[0083] The above process enables the hybrid aircraft to switch from cruise mode to climb mode. During the switching process, the hybrid aircraft will automatically complete the flight mode transition without user intervention. If the hybrid aircraft detects that the flight mode transition cannot be completed normally, it will promptly return to cruise mode and return control to the user to prevent the hybrid aircraft from going out of control.

[0084] The above description, in conjunction with specific embodiments, provides a further detailed explanation of this application and should not be construed as limiting the specific implementation of this application to these descriptions. For those skilled in the art, several simple deductions or substitutions can be made without departing from the inventive concept of this application.

Claims

1. A composite aircraft detection system for industrial flaw detection, characterized in that, Includes composite aircraft and ground stations; The composite aircraft is used to receive control from the ground station, fly to industrial buildings for close-range reconnaissance, and wirelessly transmit the collected building images to the ground station; The ground station is used to wirelessly transmit control signals to the composite aircraft and receive building images from the composite aircraft, and to obtain damage results of industrial buildings by analyzing the building images.

2. The composite aircraft detection system as described in claim 1, characterized in that, The composite aircraft includes a wireless receiving module, a flight control module, a sensing and acquisition module, a camera module, an image transmitting module, a motion module, and a power supply module; The wireless receiving module is used to receive control signals from the ground station and convert the control signals into flight control signals that the flight control module can accept. The sensing and acquisition module is used to sense the flight attitude of the composite aircraft itself through sensors, and to transmit the acquired attitude information to the flight control module. The flight control module is used to generate flight control commands based on the flight control signals and the attitude information, and to generate detection control commands when it arrives at an industrial building; The motion module is signal-connected to the flight control module and is used to adjust the flight attitude of the composite aircraft based on the flight control commands; the flight attitude of the composite aircraft includes cruise flight, overpass flight, and climbing flight; The camera module is signal-connected to the flight control module and is used to capture images of the walls and roof of the industrial building based on the detection and control commands, thereby obtaining images of the building. The image transmitting module is used to acquire the building image from the camera module and to transmit the building image to the ground station wirelessly. The power module provides the DC power required for the operation of the wireless receiving module, the flight control module, the sensing and acquisition module, the camera module, the image transmitting module, and the motion module.

3. The composite aircraft detection system as described in claim 2, characterized in that, When the flight control module generates flight control commands based on the flight control signals and the attitude information, it specifically includes: The flight control module activates the corresponding flight mode according to the flight control signal, and generates the flight control command in combination with the attitude information in the corresponding flight mode; When the flight mode is configured as cruise mode, the generated flight control commands enable the composite aircraft to have free flight capability and to fly to or away from industrial buildings. When the flight mode is configured as climbing mode, the generated flight control commands enable the composite aircraft to attach to the wall of an industrial building and travel along the wall while attached to it. When the flight mode is configured as the eaves mode, the generated flight control commands enable the composite aircraft to attach to the roof surface of the industrial building and travel along the roof surface while attached to it.

4. The composite aircraft detection system as described in claim 3, characterized in that, The motion module includes four sets of ducted fans and four sets of friction wheels driven by geared motors; The four sets of ducted fans are used to provide lift for vertical take-off and landing and power for flight, as well as to adjust flight attitude through differential steering. The four sets of friction wheels driven by geared motors are used to provide power for movement along the wall and along the top surface.

5. The composite aircraft detection system as described in claim 1, characterized in that, The ground station includes a wireless transmission module, a dual-channel receiving module, an image acquisition module, and an image processing module; The wireless transmission module has received the control command and wirelessly transmitted the control command to the composite aircraft; The control commands are generated during the user's human-computer interaction operation; The dual-channel receiving module is used to receive the building images wirelessly transmitted by the composite aircraft; The image acquisition module is used to perform preliminary processing of the building image, including image sampling and format conversion, and to transmit the pre-processed building image to the image processing module. The image processing module is used to perform algorithmic analysis on the building image to obtain damage results, including damage location, damage type, and / or damage degree; the damage results are then used to display the damage to the user via a monitor.

6. The composite aircraft detection system as described in claim 5, characterized in that, The dual-channel receiving module has two RX5808 modules, both of which are used to receive and demodulate the analog video signal corresponding to the building image through automatic frequency sweeping function, and to output a digital video signal in USB bus format to the image acquisition module.

7. The composite aircraft detection system as described in claim 5, characterized in that, The image processing module includes a model configuration submodule and an algorithm core submodule; The model configuration submodule is used to flexibly select the architecture and weight parameters of the network model according to the building type corresponding to the building image, so as to enhance the network model's ability to adapt to the detection scene; the model configuration submodule is also used to train a lightweight YOLO-S-based model according to the selected network model's architecture and weight parameters. The core submodule of the algorithm is used to extract image features from the building image using the YOLO-S lightweight model, accurately determine the damaged area based on the image features, and further analyze the damage location, damage type and / or damage degree.

8. The composite aircraft detection system as described in claim 7, characterized in that, The model configuration submodule and the algorithm core submodule are integrated into a preset industrial testing software. The industrial testing software is used to provide a weight selection window, a video source selection window, a network download window, and a processing result display window through a display. The weight selection window is used to display options to the user so that they can flexibly select the weight parameters corresponding to the network module. The video source selection window is used to display the acquisition channels of the building images to the user for selection. The acquisition channels of the building images include real-time video from the composite aircraft and historical video from local storage. The network download window is used to provide services for downloading the architecture and weight parameters of the network model, as well as for downloading updates to the industrial testing software. The processing result display window is used to display the building image and the corresponding information on the location, type and / or extent of damage.

9. The composite aircraft detection system as described in claim 5, characterized in that, The ground station also includes a remote controller, which serves as a carrier for human-computer interaction and has operation areas corresponding to setting flight missions, switching flight modes, and controlling flight attitude.

10. The composite aircraft detection system as described in claim 8, characterized in that, The remote controller includes a mode control module, which is used to perform passive switching of flight mode and automatic switching of flight mode. When the mode control module is configured for passive flight mode switching, the user can switch between cruise mode, climb mode and eaves mode by pressing buttons or touching the corresponding operation area. When the mode control module is configured for automatic flight mode switching, it automatically switches from cruise mode to climb mode or eaves mode by autonomously judging the flight attitude of the composite aircraft. It also automatically reverts to cruise mode and enters a passive flight mode switching state after detecting factors that hinder the transition, allowing the user to intervene and control the composite aircraft in a timely manner through manual operation.

Images