Building detection system and method based on unmanned aerial vehicle

Abstract

The invention discloses a building detection system and method based on an unmanned aerial vehicle. The device is characterized by comprising an adsorption module, an electromagnetic knocking module and a surface treatment module, wherein the adsorption module is used for adsorbing the unmanned aerial vehicle to a to-be-detected surface of a target building; the surface treatment module is used for cleaning surface pollutants if it is detected that the surface pollutants exist on the to-be-detected surface; and the electromagnetic knocking module is used for performing electromagnetic knocking on the to-be-detected surface and receiving a sound wave signal generated by the electromagnetic knocking, so that non-contact detection limitation can be realized, the unmanned aerial vehicle can be stably attached to the surface of the structure, misjudgment caused by surface pollution can be prevented, damage and defects in the structure can be detected, and the detection accuracy can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of drone building detection, and particularly to a drone-based building detection system and method. Background Art

[0002] Drone building detection is an intelligent detection method that combines drones, drone sensors, and artificial intelligence. Drones can quickly cover the structural surfaces of complex building structures, replacing manual climbing or aerial work platforms. Moreover, drones can perform intelligent automatic inspections based on artificial intelligence, improving the efficiency and reliability of building detection. Currently, drones often collect data on building structures through a combination of high-precision sensors (infrared, cameras, millimeter-wave radars, and lidar), and then detect and identify hazards in the building structure. However, the drone sensors in the prior art lack close-range detection means. Mainstream systems such as "Lexus 2" use a 61-megapixel full-frame imaging system, which can achieve crack recognition at the 0.1-mm level, but need to maintain a safe distance of more than 5 meters and cannot obtain microscopic details such as steel structure welds and bolt detachment. The Yapai drone detection system relies on manual operation for close observation and has more than 50% detection blind spots in complex parts such as bridge bellies and stay cables. Additionally, it lacks physical detection means, and for visual / infrared imaging, the detection rate of deep diseases such as concrete hollowing and steel bar corrosion is less than 40%. Moreover, the environmental adaptability of drones is insufficient. The imaging quality of Sony cameras drops by 60% in rainy and foggy weather, and existing drones lack a surface cleaning function, resulting in a misdetection rate of more than 30% due to dust accumulation. The Yapai system has a route deviation of ±1.2 meters in a 7-level wind environment and cannot meet the requirements of high-wind-pressure scenarios such as cross-sea bridges. Summary of the Invention

[0003] The present invention provides a drone-based building detection system and method to solve the technical problem that drones in the prior art cannot perform close-range physical detection on building structures.

[0004] According to one aspect of the present invention, there is provided a drone-based building detection system applied to a drone, including: an adsorption module, an electromagnetic tapping module, and a surface treatment module; wherein,

[0005] The adsorption module is used to adsorb the drone to the surface to be detected of the target building;

[0006] The surface treatment module is used to clean the surface contaminants if the surface contaminants are detected on the surface to be detected;

[0007] The electromagnetic tapping module is used to perform electromagnetic tapping on the surface to be detected and receive the acoustic wave signals generated by the electromagnetic tapping.

[0008] According to another aspect of the present invention, there is provided a drone-based building detection method, which is applied to a drone and includes:

[0009] Adsorb the drone to the surface to be detected of the target building through an adsorption module;

[0010] If the surface treatment module detects surface contaminants on the surface to be detected, clean the surface contaminants;

[0011] Electromagnetically strike the surface to be detected through the electromagnetic striking module and receive the acoustic wave signal generated by the electromagnetic strike.

[0012] The technical solution of the embodiment of the present invention adsorbs the drone to the surface to be detected of the target building through the adsorption module, which can adsorb the drone to the building surface, realize the close-range detection of the building by the drone, and adapt to extreme working environments; if the surface treatment module detects surface contaminants on the surface to be detected, clean the surface contaminants, which can effectively remove surface contaminants, reduce misjudgment caused by surface pollution, and can also clean the building surface and extend the service life of the building structure; strike the surface to be detected through the electromagnetic striking module and receive the acoustic wave signal generated by the electromagnetic strike. Through the contact detection of the electromagnetic strike, accurate identification of microscopic diseases can be realized, the detection accuracy can be improved, and the technical problem that the drone in the prior art cannot perform close-range physical detection on the building structure can be solved.

[0013] It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of the present invention, nor is it used to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. Description of the Drawings

[0014] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0015] Figure 1 It is a structural diagram of a drone-based building detection system provided by an embodiment of the present invention;

[0016] Figure 2 It is a flowchart of a drone-based building detection method provided by an embodiment of the present invention. Detailed Embodiments

[0017] To enable those skilled in the art to better understand the solution of the present invention, the following will clearly and completely describe the technical solution in the embodiments of the present invention in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0018] It should be noted that the terms "first", "second", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily need to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of the present invention described here can be implemented in an order other than those illustrated or described here. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0019] Figure 1 This embodiment of the present invention provides a structural diagram of a building detection system based on an unmanned aerial vehicle (UAV). This embodiment is applicable to the situation where the UAV adsorbs to a building and then detects the building, such as Figure 1 As shown, the system includes: an adsorption module 110, an electromagnetic tapping module 120, and a surface treatment module 130; among them,

[0020] The adsorption module 110 is used to adsorb the UAV to the surface to be detected of the target building;

[0021] The surface treatment module 120 is used to clean the surface contaminants if the surface contaminants are detected on the surface to be detected;

[0022] The electromagnetic tapping module 130 is used to perform electromagnetic tapping on the surface to be detected and receive the acoustic wave signals generated by the electromagnetic tapping.

[0023] Among them, the target building can be the building scanned by the UAV; for example, the target building can be a bridge, a high-rise residential building, a commercial building, etc.

[0024] Among them, the surface to be detected can be the building structure surface of the target building. For example, when the target building is a bridge, the surface to be detected can be the bridge steel structure, bridge steel cables, and steel members with coatings; when the target buildings are high-rise residential buildings and commercial buildings, the surface to be detected can be concrete, glass curtain walls, and concrete with metal inserts.

[0025] Among them, the surface contaminants can be the objects for pollution detection existing on the surface to be detected; for example, the surface dust can be the surface dust of particles with a particle size ≥ 50 μm and / or the contaminated objects with a reflectance difference ≥ 15%.

[0026] Among them, the acoustic wave signal can be the acoustic wave signal generated by electromagnetic knocking on the surface to be detected of the target building. It should be noted that the acoustic wave signal is composed of a reflected wave and a direct wave. The reflected wave is used for detecting concrete hollowing; the direct wave is used for analyzing steel component cracks.

[0027] Specifically, when the drone detects the target building, the adsorption module 110 of the drone adsorbs the drone to the surface to be detected of the target building. The surface treatment module 120 detects the surface contaminants on the surface to be detected. If surface contaminants are detected on the surface to be detected, the surface contaminants are cleaned; after cleaning the surface contaminants on the surface to be detected, the electromagnetic knocking module 130 performs electromagnetic knocking on the surface to be detected and receives the acoustic wave signal generated by the electromagnetic knocking.

[0028] Optionally, in another optional embodiment of the present invention, the adsorption module includes a multimodal sensor, a permanent magnet array unit, and a vacuum adsorption unit; among them,

[0029] The multimodal sensor is used to identify the surface to be detected in real time and determine the surface material information corresponding to the surface to be detected;

[0030] The permanent magnet array unit is used to adsorb the drone to the ferromagnetic surface if the surface material information is a ferromagnetic surface;

[0031] The vacuum adsorption unit is used to adsorb the drone to the non-ferromagnetic surface if the surface material information is a non-ferromagnetic surface.

[0032] Optionally, the permanent magnet array unit can be a Halbach permanent magnet array. The Halbach permanent magnet array is embedded in the groove of the composite material at the bottom of the drone, and the surface is flush with the drone housing to avoid protrusion affecting airtightness. The Halbach permanent magnet array generates a gradient magnetic field through pulse width modulation. For example, the gradient magnetic field generated by the permanent magnet array unit is a 0.5 - 1.2 T gradient magnetic field.

[0033] Optionally, the vacuum adsorption unit can be a negative pressure compensation system. The negative pressure compensation system is configured with a double-vortex vacuum pump and a silica gel sealing skirt to form a dynamic negative pressure cavity; the double-vortex vacuum pumps are symmetrically installed in the central area at the bottom of the drone, inside the Halbach permanent magnet array, directly connected to the air cavity of the silica gel sealing skirt, and the pump axis coincides with the center of gravity of the drone to ensure uniform distribution of the adsorption load and avoid imbalance of the flight attitude.

[0034] Optionally, the magnetic adsorption force generated by the Halbach permanent magnet array is ≥ 200 N / m 2 , and the vacuum adsorption generated by the negative pressure compensation system is ≥ -80 kPa.

[0035] Optionally, the Halbach permanent magnet array and the double-vortex vacuum pump are powered by the main power system of the unmanned aerial vehicle.

[0036] Optionally, the multi-modal sensor can identify the surface material of the surface to be detected in real time, and obtain the surface material information of the surface to be detected. The surface material information includes ferromagnetic surfaces, non-ferromagnetic surfaces, and complex surfaces. Exemplarily, ferromagnetic surfaces can be the surfaces of steel structures and bridge cables; non-ferromagnetic surfaces can be the surfaces of concrete and glass curtain walls; complex surfaces can be the surfaces of concrete with metal inserts and steel members with coatings.

[0037] Specifically, after the unmanned aerial vehicle flies to the surface to be detected of the target building, it scans the surface to be detected of the target building through the multi-modal sensor of the unmanned aerial vehicle, identifies the surface to be detected in real time, and identifies the surface material information corresponding to the surface to be detected; if the multi-modal sensor scans that the surface to be detected is a ferromagnetic surface, after the unmanned aerial vehicle approaches the surface to be detected, it activates the permanent magnet array unit to generate a magnetic adsorption force, and adsorbs the unmanned aerial vehicle to the ferromagnetic surface; if the multi-modal sensor scans that the surface to be detected is a non-ferromagnetic surface, after the unmanned aerial vehicle approaches the surface to be detected, it activates the vacuum adsorption unit to adsorb the unmanned aerial vehicle to the non-ferromagnetic surface.

[0038] Optionally, when the surface to be detected is a ferromagnetic surface, the magnetic adsorption force is mainly provided by the Halbach permanent magnet array, the vacuum adsorption unit is not activated, and a gradient magnetic field is generated through pulse width modulation to 0.8 - 1.2 T, providing a magnetic adsorption force of ≥ 200 N / m 2 to adsorb the unmanned aerial vehicle to the ferromagnetic surface; when the surface to be detected is a non-ferromagnetic surface, the adsorption of the unmanned aerial vehicle completely depends on the double-vortex vacuum pump, the Halbach permanent magnet array is turned off to save energy consumption, only the basic magnetic field of the permanent magnet is maintained, the negative pressure is increased to -80 kPa through the double-vortex vacuum pump, and a stable airtight adsorption is formed through the silicone sealing skirt to vacuum-adsorb the unmanned aerial vehicle to the non-ferromagnetic surface.

[0039] Optionally, in another optional embodiment of the present invention, the unmanned aerial vehicle further includes an edge computing unit; the edge computing unit is used to, if the surface material information is a complex surface, calculate the output force ratio of the permanent magnet array unit and the vacuum adsorption unit according to a preset dynamic load balancing algorithm, and determine the first output force ratio of the permanent magnet array unit and the second output force ratio of the vacuum adsorption unit;

[0040] The permanent magnet array unit is used to magnetically adsorb between the drone and the complex surface according to the first output ratio;

[0041] The vacuum adsorption unit is used to perform vacuum adsorption between the drone and the complex surface according to the second output ratio.

[0042] Among them, the edge computing unit can be the NVIDIA Jetson AGX Xavier set on the drone.

[0043] Among them, the preset dynamic load balancing algorithm can be an algorithm preset for dynamically adjusting the output ratios of the permanent magnet array unit and the vacuum adsorption unit. Exemplarily, the preset dynamic load balancing algorithm can be an algorithm based on dynamic weights, an algorithm based on machine learning, and / or an algorithm based on a neural network model.

[0044] Among them, the first output ratio can be the output ratio of the permanent magnet array unit during the adsorption of the drone; the second output ratio can be the output ratio of the vacuum adsorption unit during the adsorption of the drone.

[0045] Specifically, after the drone flies to the surface to be detected of the target building, the multi-modal sensor of the drone scans the surface to be detected of the target building, identifies the surface to be detected in real time, and identifies the surface material information corresponding to the surface to be detected; if the multi-modal sensor scans that the surface to be detected is a complex surface, the edge computing unit calculates the output ratios of the permanent magnet array unit and the vacuum adsorption unit according to the preset dynamic load balancing algorithm when the surface material information is a complex surface, and determines the first output ratio of the permanent magnet array unit and the second output ratio of the vacuum adsorption unit; the permanent magnet array unit performs magnetic adsorption between the drone and the complex surface according to the first output ratio; the vacuum adsorption unit performs vacuum adsorption between the drone and the complex surface according to the second output ratio. Exemplarily, when the complex surface is a steel beam coated with fireproof paint, the magnetic adsorption of the permanent magnet array unit needs to penetrate the coating, resulting in magnetic attenuation in the magnetic adsorption, and the magnetic attenuation caused by the coating thickness is compensated by the vacuum adsorption unit.

[0046] Optionally, in another optional embodiment of the present invention, the multi-modal sensor is further used to generate a vacuum adsorption assistance instruction if it is recognized that the ferromagnetic surface meets the preset vacuum adsorption unit assistance condition;

[0047] The vacuum adsorption unit is further used to perform cyclic vacuum adsorption between the drone and the ferromagnetic surface according to the vacuum adsorption assistance instruction.

[0048] Among them, the auxiliary conditions of the vacuum adsorption unit can be surface roughness, the presence of a coating or rust on the ferromagnetic surface. It should be noted that when there is surface roughness, the presence of a coating or rust on the ferromagnetic surface, the magnetic adsorption of the permanent magnet array unit will be unstable, and the vacuum adsorption unit needs to operate intermittently to cope with the magnetic adsorption instability caused by local unevenness.

[0049] Among them, the vacuum adsorption auxiliary instruction can be a control instruction for intermittently turning on the vacuum adsorption unit.

[0050] Specifically, when the multi-modal sensor identifies that the surface material information of the surface to be detected is a ferromagnetic surface, if it also identifies that there is surface roughness, the presence of a coating or rust on the ferromagnetic surface, it is considered that the ferromagnetic surface meets the preset auxiliary conditions of the vacuum adsorption unit, generates a vacuum adsorption auxiliary instruction, and the vacuum adsorption unit is intermittently turned on according to the vacuum adsorption auxiliary instruction to perform cyclic vacuum adsorption between the drone and the ferromagnetic surface to cope with the magnetic adsorption instability caused by local unevenness.

[0051] Optionally, in another alternative embodiment of the present invention, the surface treatment module includes a pollutant detection sensor and a Venturi air sweeping unit, among which,

[0052] The pollutant detection sensor is used to detect the surface to be detected in real time and determine the surface detection result of the surface to be detected;

[0053] The edge computing unit is further configured to generate a pollutant cleaning instruction if there are surface pollutants in the surface detection result;

[0054] The Venturi air sweeping unit is used to sweep the surface pollutants according to the pollutant cleaning instruction.

[0055] Among them, the pollutant detection sensor can be a camera installed on the drone and an infrared sensor for detecting reflectivity differences.

[0056] Among them, the surface detection result can be the detection result of the pollutant detection sensor for detecting pollutants on the surface to be tested. It should be noted that if the camera captures surface dust on the surface to be tested, the surface detection result can be the presence of surface pollutants, and the surface pollutants are surface dust; if the infrared sensor detects pollutants with reflectivity differences, the surface detection result can be the presence of surface pollutants, and the surface pollutants are pollutants; if the camera captures surface dust on the surface to be tested and the infrared sensor detects pollutants with reflectivity differences, the surface detection result can be the presence of surface pollutants, and the surface pollutants are pollutants and surface dust.

[0057] Among them, the pollutant cleaning instruction can be an instruction to start the Venturi air sweeping unit.

[0058] Optionally, the Venturi air-sweeping unit can be integrated at the front end of the bottom of the UAV, located between the vacuum adsorption unit, the permanent magnet array unit and the electromagnetic knocking module. The air outlet of the Venturi air-sweeping unit faces the bottom detection surface of the UAV, forming a 15° angle with the surface, ensuring that the transient air flow of 30 m / s covers the projection area of the UAV (with a diameter of about 0.5 m); the gas storage tanks of the Venturi air-sweeping unit are symmetrically arranged in the middle of the UAV fuselage and are connected to the Venturi nozzle of the Venturi air-sweeping unit through high-pressure pipelines (with a pressure resistance of ≥1.2 MPa). The air flow path is aligned with the edge of the silicone sealing skirt, and the local negative pressure (about -10 kPa) generated by the Bernoulli effect is used to enhance the airtightness and assist the vacuum pump in adsorption.

[0059] Optionally, the Venturi air-sweeping unit adopts modular integration. The nozzle assembly of the Venturi air-sweeping unit: The Venturi tube is 3D printed with titanium alloy, and the inner wall is polished (Ra ≤ 0.4 μm) to reduce air flow friction loss; the fixing bracket of the Venturi air-sweeping unit uses a carbon fiber material bracket to fix the nozzle in the UAV; the capacity of the high-pressure gas storage tank of the Venturi air-sweeping unit is 2 L, with an aluminum inner liner wound with carbon fiber, a pressure of 0.8 MPa, and is connected to the Venturi tube through a quick-release joint; the Venturi air-sweeping unit is controlled by a solenoid valve: The air flow pulse frequency (1 - 10 Hz) is adjusted through a high-frequency solenoid valve (response time ≤ 5 ms) to match the cleaning requirements. The air path interface of the Venturi air-sweeping unit is sealed with an O-ring to prevent high-pressure leakage; a detachable dust cover is installed outside the nozzle to avoid foreign object blockage during flight. In the preset card slot of the bottom shell of the aircraft, vibration is isolated through a shock-absorbing gasket (silicone material).

[0060] Optionally, the Venturi air-sweeping unit defaults to the automatic mode. When the surface detection unit detects surface pollutants, the edge computing unit generates a pollutant cleaning instruction to start the Venturi air-sweeping unit to clean the surface pollutants.

[0061] Specifically, the pollutant detection sensor of the UAV detects the surface to be detected in real time, determines the surface detection result of the surface to be detected. If there are surface pollutants in the surface detection result, the edge computing unit generates a pollutant cleaning instruction, and the Venturi air-sweeping unit cleans the surface pollutants according to the pollutant cleaning instruction.

[0062] Optionally, in another alternative embodiment of the present invention,

[0063] The UAV further includes a remote communication unit and a remote control unit; wherein,

[0064] The remote communication unit is used to send the surface detection result to the UAV ground station; the remote control unit is used to receive the pollutant cleaning instruction sent by the UAV ground station.

[0065] Optionally, the telecommunication unit enables the UAV to communicate with the ground station. The ground station is the core control center of the UAV and communicates bidirectionally with the UAV's telecommunication unit in the form of a data link. It can receive the surface detection results sent by the UAV through the telecommunication unit. The staff at the ground station can select emergency cleaning or fixed-point deep cleaning through the ground station remote control to generate a pollutant cleaning instruction, and send the pollutant cleaning instruction to the UAV's telecommunication unit, thereby controlling the Venturi air sweeping unit to clean the surface pollutants.

[0066] Optionally, when the Venturi air sweeping unit cleans the surface pollutants, it can be divided into a pre-cleaning stage and a continuous cleaning stage. In the pre-cleaning stage, 3-5 pulses of air flow (each pulse lasting 0.2 seconds) are emitted to remove the loose dust in the detection area. The low-frequency pulse (1Hz) is started as needed to maintain the surface cleanliness and prevent secondary pollution. Exemplarily, in the pre-cleaning stage, the removal rate is ≥90%.

[0067] Optionally, when the Venturi air sweeping unit cleans the surface pollutants, it can dynamically adjust the relevant parameters of the cleaning. The relevant parameters can include the air flow intensity and the spraying angle. Exemplarily, the air flow intensity dynamically adjusts the output pressure of the air storage tank (0.5-0.8MPa) according to the surface roughness (Ra 0.8-6.3μm); the spraying angle adjusts the nozzle pitch angle (±10°) through a micro servo to adapt to the curved surface cleaning (the radius of curvature ≥0.3m).

[0068] Optionally, the Venturi air sweeping unit can clean before the UAV adsorbs or during the UAV's adsorption. Cleaning before the UAV adsorbs is the core node of the UAV detection, mainly used to remove the surface dust, assist the UAV to obtain pollution-free detection data; it can also enhance the adsorption, use the Bernoulli effect to form a local negative pressure at the edge of the silicone sealing skirt, and assist the vacuum pump to quickly establish a stable adsorption. Among them, the local negative pressure is about -10kPa, and the adsorption establishment time can be shortened by 50%.

[0069] During the UAV's adsorption, it is an auxiliary node of the UAV detection. It can prevent the dust from covering the surface to be detected during the operation through intermittent pulse air flow. When there is a slight air leakage due to building vibration or wind load, the air flow is instantaneously enhanced to restore the airtightness of the negative pressure chamber.

[0070] Optionally, in another optional embodiment of the present invention, the electromagnetic knocking module includes a detection and positioning unit, an edge computing unit, an electromagnetic knocking unit, and a piezoelectric sensor array; wherein,

[0071] The detection and positioning unit is used to scan the surface to be detected, determine the surface visual information and the three-dimensional point cloud information, and determine the area to be knocked according to the surface visual information and the three-dimensional point cloud information;

[0072] The edge computing unit is configured to perform a tapping calculation on the area to be tapped according to a preset tapping calculation algorithm to determine optimal tapping parameters;

[0073] The electromagnetic tapping unit is configured to perform electromagnetic tapping on the area to be tapped according to the optimal tapping parameters;

[0074] The piezoelectric sensor array is configured to receive the acoustic wave signals generated by the electromagnetic tapping.

[0075] Wherein, the detection and positioning unit consists of a binocular vision camera and a millimeter-wave radar of the unmanned aerial vehicle. The binocular vision camera and the millimeter-wave radar can automatically scan the surface to be detected, obtain the binocular vision point cloud of the surface to be detected through the binocular vision camera, that is, the surface vision information; detect the surface point cloud of the surface to be detected through the millimeter-wave radar, that is, the three-dimensional point cloud information.

[0076] Optionally, the unmanned aerial vehicle integrates the surface vision information and the three-dimensional point cloud information by adopting feature fusion or data fusion to generate a high-precision 3D point cloud map, and identifies the area to be tapped based on the high-precision 3D point cloud map. Exemplarily, the area to be tapped can be a weld seam and a bolt.

[0077] Wherein, the preset tapping calculation algorithm can be an algorithm preset for calculating the tapping parameters of the electromagnetic tapping unit. Exemplarily, the preset tapping calculation algorithm can be an improved ICP algorithm introducing the FPFH feature descriptor.

[0078] Wherein, the optimal tapping parameters can be the optimal pose and tapping parameters for the electromagnetic tapping unit to tap the area to be tapped.

[0079] Specifically, the detection and positioning unit of the unmanned aerial vehicle scans the surface to be detected, obtains the surface vision information and the three-dimensional point cloud information, constructs a high-precision 3D point cloud map based on the surface vision information and the three-dimensional point cloud information, identifies the area to be tapped based on the high-precision 3D point cloud map, the edge computing unit performs a tapping calculation on the area to be tapped according to the preset tapping calculation algorithm to determine the optimal tapping parameters; the electromagnetic tapping unit performs electromagnetic tapping on the area to be tapped according to the optimal tapping parameters, and the piezoelectric sensor array receives the acoustic wave signals generated by the electromagnetic tapping.

[0080] Optionally, in another optional embodiment of the present invention, the electromagnetic tapping unit includes a six-degree-of-freedom robotic arm and an electromagnetic tapping head; wherein,

[0081] The electromagnetic tapping head is configured to perform electromagnetic tapping on the area to be tapped according to the optimal tapping parameters, and detect the contact pressure in real time, and calculate the actual offset according to the contact pressure;

[0082] The edge computing unit is further configured to, if the actual offset is greater than a preset offset threshold, generate an adaptive adjustment instruction through a preset adaptive control algorithm, update the pose of the optimal knocking parameters according to the adaptive adjustment instruction, and determine the updated optimal knocking parameters;

[0083] The six-degree-of-freedom robotic arm is further configured to adjust its pose according to the updated optimal knocking parameters.

[0084] Optionally, the electromagnetic knocking unit consists of a six-degree-of-freedom robotic arm and an electromagnetic knocking head. The six-degree-of-freedom robotic arm realizes pose adjustment through a carbon fiber connecting rod and a harmonic reducer, and the electromagnetic knocking head generates shock waves of 10 - 100 Hz based on the Lorentz force principle.

[0085] Optionally, a 6-axis torque sensor is integrated in the electromagnetic knocking head to monitor the contact pressure.

[0086] Optionally, the preset adaptive control algorithm can be an algorithm for adaptively adjusting the optimal knocking parameters; the adaptive adjustment instruction can be an instruction for adaptively adjusting the optimal knocking parameters.

[0087] Among them, the contact pressure can be obtained by monitoring of the electromagnetic knocking head. The actual offset can be the offset value when the electromagnetic knocking head performs knocking; the preset offset threshold can be preset for identifying the offset value.

[0088] Optionally, the electromagnetic knocking head is connected to the six-degree-of-freedom robotic arm. The electromagnetic knocking head and the six-degree-of-freedom robotic arm perform electromagnetic knocking on the area to be knocked according to the optimal knocking parameters; when the edge computing unit of the unmanned aerial vehicle detects that the actual offset is greater than the preset offset threshold, it generates an adaptive adjustment instruction through the preset adaptive control algorithm, updates the pose of the optimal knocking parameters through the adaptive adjustment instruction, and determines the updated optimal knocking parameters; the six-degree-of-freedom robotic arm adjusts its pose according to the updated optimal knocking parameters, and the electromagnetic knocking head performs electromagnetic knocking on the area to be knocked according to the updated optimal knocking parameters. Exemplarily, the optimal knocking parameters can be a frequency of 10 - 100 Hz, a peak force of 10 - 50 N, and a duration of 1 - 100 ms.

[0089] Optionally, the electromagnetic knocking head is provided with an overcurrent protection module to limit the maximum current and prevent the coil of the electromagnetic knocking head from overheating; and a self-check program checks the pose of the six-degree-of-freedom robotic arm before each knocking to avoid accidental touch and collision.

[0090] Optionally, in another optional embodiment of the present invention, the piezoelectric sensor array includes a main sensor array and an auxiliary sensor array;

[0091] The main sensor array is configured to receive the reflected wave of the acoustic wave signal generated by the electromagnetic knocking;

[0092] The sensor auxiliary array is used to receive the direct wave of the acoustic wave signal generated by the electromagnetic knock.

[0093] Optionally, the piezoelectric sensor array includes a main sensor array and an auxiliary sensor array. The main sensor array is embedded in the bottom detection surface of the drone, arranged in a ring around the electromagnetic knock head, with a diameter of 200 mm, and includes 64 piezoelectric ceramic sensors with a sensitivity of 50 mV / g for the piezoelectric ceramic sensors; the auxiliary sensor array is integrated at the end of the six-degree-of-freedom robotic arm, adjacent to the electromagnetic knock head, and the distance from the electromagnetic knock head is ≤10 mm, and is used for enhanced acquisition of near-field acoustic wave signals.

[0094] Optionally, the acoustic wave signal is stored in a solid-state drive with an anti-vibration design, and the retention period is ≥30 days.

[0095] Specifically, the main sensor array receives the reflected wave of the acoustic wave signal generated by the electromagnetic knock; the auxiliary sensor array receives the direct wave of the acoustic wave signal generated by the electromagnetic knock, and improves the diagnostic accuracy through the wave velocity difference.

[0096] Optionally, the edge computing unit of the drone can, during the flight of the drone, automatically perform path planning, automatically identify the surface to be detected, and automatically scan the surface material information of the surface to be detected to achieve automatic adsorption of the drone. In the case of surface contaminants on the surface to be detected, it automatically generates a contaminant cleaning instruction to clean the surface to be detected. The edge computing unit automatically determines the area to be knocked on the surface to be detected, generates optimal knocking parameters, controls the six-degree-of-freedom robotic arm and the electromagnetic knock head to perform electromagnetic knocking, and receives the acoustic wave signal generated by the electromagnetic knock, and automatically analyzes the acoustic wave signal to obtain the detection mechanism of the electromagnetic knock. The edge computing unit of the drone can realize full-automatic control of identifying the detection surface, automatic adsorption, automatic cleaning, and automatic detection, and effectively improve the detection efficiency.

[0097] The technical solution of the embodiment of the present invention adsorbs the drone to the surface to be detected of the target building through the adsorption module, can adsorb the drone to the building surface, realizes close-range detection of the building by the drone, and adapts to extreme working environments; if the surface treatment module detects surface contaminants on the surface to be detected, it cleans the surface contaminants, can effectively remove surface contaminants, reduce misjudgment caused by surface contamination, and can also clean the building surface and extend the service life of the building structure; the electromagnetic knock module performs electromagnetic knocking on the surface to be detected and receives the acoustic wave signal generated by the electromagnetic knock. Through contact detection of the electromagnetic knock, it can accurately identify microscopic diseases and improve the detection accuracy, solving the technical problem that drones in the prior art cannot perform close-range physical detection on building structures.

[0098] Figure 2 The flowchart of a building detection method based on a drone provided by an embodiment of the present invention is applicable to the situation of adsorbing and detecting the building structure on the outer surface of the building by the drone. This method can be executed by a building detection system based on a drone, and the building detection system based on a drone can be implemented in the form of hardware and / or software, and the building detection system based on a drone can be configured in the drone. As Figure 2 shown, the method includes:

[0099] S210. Adsorb the drone to the surface to be detected of the target building through the adsorption module.

[0100] S220. If the surface treatment module detects surface contaminants on the surface to be detected, clean the surface contaminants.

[0101] S230. Electromagnetically strike the surface to be detected through the electromagnetic striking module and receive the sound wave signal generated by the electromagnetic strike.

[0102] The technical solution of the embodiment of the present invention adsorbs the drone to the surface to be detected of the target building through the adsorption module, can adsorb the drone to the building surface, realizes the close-range detection of the building by the drone, and adapts to extreme working environments; if the surface treatment module detects surface contaminants on the surface to be detected, clean the surface contaminants, can effectively remove surface contaminants, reduce misjudgment caused by surface contamination, and can also clean the building surface and extend the service life of the building structure; strike the surface to be detected through the electromagnetic striking module and receive the sound wave signal generated by the electromagnetic strike, and through the contact detection of the electromagnetic strike, can realize the accurate identification of microscopic diseases, improve the detection accuracy, and solve the technical problem that the drone in the prior art cannot perform close-range physical detection on the building structure.

[0103] It should be understood that various forms of the processes shown above can be used, steps can be reordered, added or deleted. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and no limitation is made herein.

[0104] The above specific implementation manners do not constitute a limitation to the protection scope of the present invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A drone-based building inspection system, applied to drones, characterized in that: include: Adsorption module, electromagnetic knocking module and surface treatment module; among them, The adsorption module is used to adsorb the drone to the surface to be detected of the target building; The surface treatment module is used to clean the surface contaminants if it is detected that there are surface contaminants on the surface to be detected; The electromagnetic knocking module is used to perform electromagnetic knocking on the surface to be detected and receive the sound wave signal generated by the electromagnetic knocking.

2. The system according to claim 1, characterized in that The adsorption module includes a multimodal sensor, a permanent magnetic array unit and a vacuum adsorption unit; wherein, The multimodal sensor is used to identify the surface to be detected in real time and determine the surface material information corresponding to the surface to be detected; The permanent magnetic array unit is used to adsorb the drone to the ferromagnetic surface if the surface material information is a ferromagnetic surface; The vacuum adsorption unit is used to adsorb the drone to the non-ferromagnetic surface if the surface material information is a non-ferromagnetic surface.

3. The system according to claim 2, characterized in that The drone also includes an edge computing unit; The edge computing unit is used to calculate the output ratio of the permanent magnetic array unit and the vacuum adsorption unit according to a preset dynamic load balancing algorithm if the surface material information is a complex surface, and determine a first output ratio of the permanent magnetic array unit and a second output ratio of the vacuum adsorption unit; The permanent magnetic array unit is used to perform magnetic adsorption between the drone and the complex surface according to the first output ratio; The vacuum adsorption unit is used to perform vacuum adsorption between the drone and the complex surface according to the second output ratio.

4. The system according to claim 2, characterized in that Also includes: The multimodal sensor is further configured to generate a vacuum adsorption assist instruction if it is identified that the ferromagnetic surface satisfies a preset vacuum adsorption unit assist condition; The vacuum adsorption unit is further used to perform cyclic vacuum adsorption between the drone and the ferromagnetic surface according to the vacuum adsorption auxiliary instruction.

5. The system according to claim 3, characterized in that The surface treatment module includes a pollutant detection sensor and a venturi air sweep unit; wherein, The pollutant detection sensor is used to detect the surface to be detected in real time and determine the surface detection result of the surface to be detected; The edge computing unit is further configured to generate a pollutant cleaning instruction if there are surface pollutants in the surface detection result; The Venturi air sweeping unit is used to clean the surface pollutants according to the pollutant cleaning instruction.

6. The system according to claim 5, characterized in that The drone also includes a remote communication unit and a remote control unit; wherein, The remote communication unit is used to send the surface detection result to the UAV ground station; The remote control unit is used to receive the pollutant cleaning instruction sent by the UAV ground station.

7. The system according to claim 6, characterized in that The electromagnetic knocking module includes a detection and positioning unit, an edge computing unit, an electromagnetic knocking unit and a piezoelectric sensor array; wherein, The detection and positioning unit is used to scan the surface to be detected, determine surface visual information and three-dimensional point cloud information, and determine the area to be tapped according to the surface visual information and the three-dimensional point cloud information; The edge computing unit is used to perform posture calculation on the area to be tapped according to a preset posture calculation algorithm to determine optimal tapping parameters; The electromagnetic knocking unit is used to perform electromagnetic knocking on the area to be knocked according to the optimal knocking parameters; The piezoelectric sensor array is used to receive the sound wave signal generated by electromagnetic knocking.

8. The system according to claim 7, characterized in that The electromagnetic knocking unit includes a six-degree-of-freedom mechanical arm and an electromagnetic knocking head; wherein, The electromagnetic knocking head is used to perform electromagnetic knocking on the area to be knocked according to the optimal knocking parameters, detect the contact pressure in real time, and calculate the actual offset according to the contact pressure; The edge computing unit is further configured to generate an adaptive adjustment instruction through a preset adaptive control algorithm if the actual offset is greater than a preset offset threshold, update the optimal tapping parameter according to the adaptive adjustment instruction, and determine the updated optimal tapping parameter; The six-degree-of-freedom robotic arm is also used to adjust its position and posture according to the updated optimal knocking parameters.

9. The system according to claim 7, characterized in that The piezoelectric sensor array includes a main sensor array and an auxiliary sensor array; The sensor main array is used to receive the reflected wave of the sound wave signal generated by electromagnetic knocking; The sensor auxiliary array is used to receive the direct wave of the sound wave signal generated by electromagnetic knocking.

10. A drone-based building inspection method, applied to a drone, characterized in that: include: Adsorbing the drone to the surface to be detected of the target building through an adsorption module; If the surface treatment module detects that there are surface contaminants on the surface to be detected, the surface contaminants are cleaned; Electromagnetic knocking is performed on the surface to be detected through an electromagnetic knocking module, and a sound wave signal generated by the electromagnetic knocking is received.

Images