Outer wall hollowing detection robot and positioning detection method

Abstract

The invention discloses an outer wall hollowing detection robot which comprises a robot body unit, a guide unit, a traction unit, a detection unit and a positioning unit. The guide unit is configured to be of a rigid guide frame structure capable of being fixedly installed on the outer wall of a building. The robot main body unit is configured to be movably mounted on the guide unit; the traction unit is configured to drive the robot main body unit to ascend and descend along the guide unit through manual operation or motor driving; the detection unit is arranged on the robot main body unit, and sound wave signals of an outer wall are stimulated and collected through knocking; the positioning unit is configured to determine the real-time position of the robot main body unit on the guiding unit and associate the position information with the sound wave signal; the technical problems that in the prior art, high-altitude detection operation is poor in safety, detection data are not accurate due to machine body shaking, the hollowing defect position cannot be accurately positioned, and detection equipment is difficult to deploy are solved.

Description

Technical Field

[0001] This invention relates to the field of non-destructive testing technology for building quality, and in particular to a robot for detecting hollow areas in exterior walls and a positioning and detection method therefor. Background Technology

[0002] Hollow areas can easily form between the exterior wall cladding and the main structure due to construction, material, or environmental factors. These defects pose a risk of cladding detachment, seriously threatening public safety. Currently, the main detection methods have significant limitations:

[0003] Manual tapping inspection: inefficient, subjective results, and extremely high risk of working at height.

[0004] Unmanned aerial vehicle (UAV)-based detection: highly susceptible to wind interference, poor positioning accuracy, and difficult to operate stably and continuously.

[0005] Existing track-based robots have complex and bulky track systems, are inconvenient to install, and are prone to shaking, resulting in inaccurate detection data.

[0006] Therefore, there is an urgent need in this field for an external wall hollow detection solution that can take into account the safety of high-altitude operations, the stability of the detection process, the accuracy and reliability of data, and the convenience and efficiency of deployment. Summary of the Invention

[0007] The purpose of this invention is to provide a robot for detecting hollow areas in exterior walls, so as to solve the technical problems in the prior art, such as poor safety of high-altitude inspection operations, inaccurate inspection data due to machine shaking, inability to accurately locate the location of hollow defects, and difficulty in deploying inspection equipment.

[0008] To achieve the above objectives, the present invention adopts the following technical solution:

[0009] A robot for detecting hollow exterior walls includes a main robot unit, a guiding unit, a traction unit, a detection unit, and a positioning unit.

[0010] The guide unit is configured as a rigid guide frame structure that can be fixedly installed on the exterior wall of the building;

[0011] The robot body unit is configured to be movably mounted on the guide unit;

[0012] The traction unit is configured to drive the robot body unit to move up and down along the guide unit via manual operation or motor drive;

[0013] The detection unit is mounted on the main body of the robot and collects the sound wave signals from the outer wall by tapping.

[0014] The positioning unit is configured to determine the real-time position of the robot body unit on the guide unit and associate the position information with the acoustic signal.

[0015] Furthermore, the guide unit includes two parallel vertical guide rails, which are segmented and connected in a detachable manner.

[0016] The vertical guide rail of the guide unit has a T-shaped cross-section; the track inlay groove of the robot body unit is a T-shaped groove adapted to the T-shaped guide rail cross-section;

[0017] The T-shaped groove has rubber sealing edges on both sides of the opening facing the outside of the guide rail. When the robot body unit is installed on the guide rail, the rubber sealing edges can generate elastic deformation to achieve tight fit and assist in guidance.

[0018] Furthermore, the traction unit includes a traction rope mounted on the robot's main body unit. The fixed end of the traction rope is fixedly connected to the robot's main body unit, and the free end of the traction rope extends to the building's roof. Lifting and lowering are achieved by manual operation or by a traction motor.

[0019] Furthermore, the detection unit includes a rotating arm drive control device, a rotating arm, a striking rod, a striking head, and an acoustic sensor; the striking rod is made of an elastic material, with its middle part fixedly installed at the front end of the rotating arm; the striking head is made of a rigid material and is disposed at the end of the striking rod; the drive control device drives the rotating arm and the striking rod to rotate, causing the striking head to strike the exterior wall, and the acoustic sensor is used to collect the generated sound wave signal.

[0020] The striking heads are arranged in pairs at both ends of the elastic striking rod; during operation, the rotating arm drive control device drives the rotating arm and the striking rod to rotate together, so that the rigid striking heads at both ends alternately hammer the surface of the exterior wall.

[0021] The rotating arm drive control device includes a rotating arm rotary motor and a speed controller; the speed controller is used to adjust the motor speed to control the tapping frequency, and is connected to the positioning unit to coordinate the tapping frequency with the robot's moving speed.

[0022] Furthermore, the positioning unit includes an optical sensor mounted on the robot's main body and multiple transverse identification tracks mounted between two parallel vertical guide rails. The multiple transverse identification tracks are evenly spaced and have position markers for optical sensor identification.

[0023] Furthermore, a robot for detecting hollow exterior walls also includes a detachable data storage device disposed on the main body of the robot, for storing the acoustic signal and its associated location information.

[0024] Furthermore, a robot for detecting hollow exterior walls also includes an adsorption mechanism; the adsorption mechanism adopts a negative pressure adsorption device or an electromagnetic adsorption device, the negative pressure adsorption device includes a negative pressure fan disposed on the main body unit of the robot, used to generate negative pressure to make the robot adhere to the wall surface; the electromagnetic adsorption device includes an electromagnet disposed on the main body unit of the robot, used to adhere and fix it to any position of the vertical guide rail.

[0025] Furthermore, the robot's main body unit is equipped with a removable battery for powering electrical components.

[0026] A method for locating and detecting hollow areas in exterior walls includes the following steps:

[0027] Step 1) Assemble the guide unit on the roof of the building. Lower the assembled guide unit from top to bottom along the exterior wall of the building under the action of gravity, and fix the top and bottom of the guide unit to the corresponding positions on the building.

[0028] Step 2) Install the traction unit onto the robot body unit, and then install the robot body unit onto the guide unit;

[0029] Step 3) Control the traction unit to move the robot body unit along the guide unit. During the movement of the robot body unit, control the excitation mechanism of the detection unit to perform periodic mechanical excitation on the outer wall surface.

[0030] Step 4) Perform the following operations simultaneously:

[0031] 4.1) Acquire the acoustic response signal generated by the mechanical excitation using an acoustic sensor;

[0032] 4.2) The positioning unit obtains the real-time position information of the robot body unit on the guiding unit by reading the absolute position reference mark pre-set on the guiding unit;

[0033] Step 5) Bind the acoustic response signal with the corresponding real-time location information, that is, align and store the acoustic response signal and location information with a unified timestamp to generate a spatiotemporal related dataset;

[0034] Step 6) Based on the spatiotemporal correlation dataset, analyze and locate the hollow defects of the exterior wall.

[0035] Furthermore, the frequency and force of the alternating taps are controlled by adjusting the rotation speed of the detection unit; and the triggering frequency of the mechanical excitation is dynamically adjusted according to real-time position information or the moving speed of the robot's main body unit.

[0036] Furthermore, step 6) specifically involves the following steps:

[0037] 6.1) Signal Acquisition and Analog-to-Electronic Signal Conversion: The hammering frequency and detection time parameters are set through the control interface of the detection equipment. The acoustic wave acquisition device captures the acoustic wave signal generated by the hammering of the external wall and converts it into a continuous-time domain electrical signal. The electrical signal satisfies the following relationship:

[0038]

[0039] In the formula: This represents an electrical signal, where A represents the amplitude, f represents the frequency, and Φ represents the phase.

[0040] 6.2) Sampling, filtering, and quantization processing: According to the preset sampling frequency... The continuous electrical signal is sampled to determine the sampling period. And satisfy the sampling frequency ≥2 ,in The highest frequency of the signal; frequencies higher than the Nyquist frequency are filtered out using a low-pass filter. The signal components are analyzed; the sampled discrete signal values ​​are mapped to finite-precision digital quantities for quantization, thus completing the conversion from analog to digital signals.

[0041] 6.3) Signal Refinement Processing and Feature Parameter Processing: The quantized discrete digital signal is processed by frame segmentation, dividing the continuous data into several short time frames, each containing a preset number of sampling points;

[0042] Windowing is applied to each frame of the signal; the window function is... Using Hanming windows, the following relational expression is satisfied:

[0043]

[0044] in, This represents the discrete-time signal obtained after analog-to-digital conversion and processing, where N represents the frame length.

[0045] The signal after windowing is: To reduce spectrum leakage caused by signal truncation;

[0046] Each frame of the windowed signal is processed by fractional Fourier transform (FRFT). By adjusting the transform order parameter between 0 and 1, the gradual characteristics of the signal from the time domain to the frequency domain are obtained to assist in the main frequency calibration.

[0047] The frequency domain representation is obtained by calculating the discrete Fourier transform of each frame of the signal:

[0048] k=0,1,2…,N-1;

[0049] Combined with frequency resolution: Determine the main frequency of the signal ;

[0050] The power value is obtained by squaring the frequency domain representation, and the amplitude spectrum is calculated to extract the amplitude data of each frame signal.

[0051] Wavelet transform is used to denoise the signal. By selecting the optimal threshold β, wavelet coefficients with amplitudes higher than β are retained, while wavelet coefficients with amplitudes lower than β are set to zero, thereby achieving signal-to-noise separation. The threshold satisfies the following relationship:

[0052]

[0053] in, For adjustment coefficients and 0 ≤ ≤ 1, Represents wavelet coefficients;

[0054] Finally, the discrete amplitude sequence and the signal's dominant frequency are extracted from the processed signal. Data on the start-end time interval of each frame signal and the signal amplitude attenuation trend;

[0055] 6.4) Multidimensional determination of hollow sound: Based on the feature parameters extracted in step 6.3), cross-analysis is performed in combination with the single-pulse acoustic echo signal model;

[0056] The model expression for the pulse acoustic echo signal is:

[0057]

[0058] in, The amplitude parameter represents the signal, corresponding to the peak value of the discrete amplitude sequence; This represents the attenuation coefficient, obtained by fitting the amplitude attenuation trend. Represents a time variable. This represents the delay parameter, corresponding to the sound wave propagation time; This indicates the signal frequency, which is consistent with the extracted main frequency.

[0059] Anomaly detection is performed on the amplitude of the detection points:

[0060]

[0061]

[0062] In the formula, Indicates the average amplitude. This represents the critical value of the amplitude, and n is the number of detection points; when both of the above formulas meet the conditions, it is determined that the amplitude is abnormal.

[0063] Anomaly detection is performed on the slope of the detection points:

[0064]

[0065]

[0066]

[0067] In the formula: and These represent the sound wave propagation times at the i-th and i-1-th detection points, respectively. and These represent the depths of point i and point i-1, respectively.

[0068] When the slope changes beyond the preset range, it is determined to be an abnormal slope.

[0069] When a certain area simultaneously meets the criteria of amplitude anomaly and slope anomaly, the area is determined to be a hollow area of ​​the wall.

[0070] Compared with the prior art, the present invention has the following significant advantages:

[0071] 1) High rigidity and stability: The unique "ladder-shaped" guide unit transforms the traditional flexible guide system into a rigid support frame, fundamentally overcoming the problem of body shaking caused by the robot's own weight and periodic impact reaction force, and providing a physical basis for acoustic sensors to collect high signal-to-noise ratio and repeatable acoustic wave data.

[0072] 2) Multiple safety guarantees: Through the triple safety guarantee mechanism consisting of "negative pressure adsorption + magnetic assistance + T-shaped groove", the robot can still be firmly attached to the wall when encountering gusts of wind or unexpected changes in traction force, which greatly reduces the risk of falling from height.

[0073] 3) Objectivity and traceability of test results: Through the synchronous operation of the positioning unit and the detection unit, precise binding of acoustic data and location information is achieved. This frees the test results from dependence on subjective human auditory experience, transforming them into objective and quantifiable data pairs. Any abnormal acoustic signal can be traced back to the specific location on the wall, providing precise coordinates for subsequent verification and maintenance.

[0074] 4) Portability and versatility: The segmented design of the guide unit and the modular and detachable design of the robot's core components enable the entire system to be broken down into parts, easily transported to the site, and quickly assembled, greatly improving the equipment's versatility and engineering application efficiency. Attached Figure Description

[0075] Figure 1This is a schematic diagram of the robot mounting structure of the present invention;

[0076] Figure 2 This is a schematic diagram of the bottom structure of the robot of the present invention;

[0077] Figure 3 This is a front view of the robot of the present invention;

[0078] Figure 4 This is a side view of the robot of the present invention;

[0079] Figure 5 This is a top view of the guide unit of the present invention;

[0080] Figure 6 This is a side view of the guide unit of the present invention;

[0081] Figure 7 This is a schematic diagram of the vertical guide rail structure of the present invention;

[0082] In the diagram: A. Horizontal identification track; B. Vertical guide rail; C. Pin connector; D. Traction rope; E. Negative pressure fan; F. Removable data storage device; G. Optical sensor; H. Acoustic sensor; I. ​​Rotating arm; J. Rubber striking rod; K. Striking head; L. Removable battery; M. Electromagnet; N. Rubber sealing edge; O. Track inlay groove; P. Rotating arm drive control device. Detailed Implementation

[0083] The present invention will be further described below with reference to the accompanying drawings.

[0084] Reference Figures 1 to 7 A robot for detecting hollow exterior walls includes a main robot unit, a guiding unit, a traction unit, a detection unit, and a positioning unit.

[0085] The guiding unit forms the core motion framework of the robot, consisting of two parallel vertical guide rails B. This guiding unit adopts a segmented modular design, with each guide rail unit having a male-female mating pin connector C at its end. Locking with these pins allows for quick and detachable connection of multiple guide rail units in the vertical direction. Between the two parallel vertical guide rails, multiple lateral recognition tracks A (optical sensor recognition tracks) are fixedly connected at precise predetermined intervals. The lateral recognition tracks A and the vertical guide rails B together form a stable, rigid trapezoidal structure. During installation, this trapezoidal guiding unit is unfolded on the building roof and lowered to the outdoor ground level under its own weight. The top and bottom ends of the guiding unit are quickly anchored by plugging into or embedding into pre-drilled fixing holes in the building structure, or other fastening methods can be used to ensure the stability of the guiding unit.

[0086] The robot body unit is movably mounted on the guide unit via a track inlay slot O at its bottom. In a preferred embodiment, the vertical guide rail B of the guide unit has an inverted T-shaped cross-section and is made of an alloy material that can be magnetically attracted. Correspondingly, the guide unit of the robot body unit is a T-shaped slot precisely fitted to the cross-section of the T-shaped guide rail, with permanent magnets or electromagnets embedded at key contact points within the slot as magnetic attraction materials. This T-shaped slot structure also has lip-shaped rubber seals N on both sides of the opening facing the guide rail. When the robot body unit is mounted on the guide rail, the rubber seals N undergo elastic deformation, achieving both a tight fit with the guide rail to isolate dust and providing auxiliary frictional guidance.

[0087] The traction unit employs a simple and reliable human-driven solution, comprising a traction rope D mounted on the upper part of the robot's main body, with its free end extending upwards to the building's roof. Operators can directly drive the robot's main body up and down along the guide unit by manually releasing and retracting the traction rope within a safe area on the roof.

[0088] The detection unit, which performs the core detection function, is located on the main robot unit and includes a rotating arm drive control device P, a rotating arm I, a rubber striking rod J, striking heads K, an acoustic sensor H, and a detachable data storage device F. The rotating arm I (alloy rotating arm) is driven by the rotating arm drive control device P and performs precise rotational movements. The middle part of the rubber striking rod J is rigidly fixed to the front end of the rotating arm I via a flange, forming a synchronously rotating assembly. Striking heads K (steel striking heads) are arranged in pairs at both ends of the rubber striking rod J. During operation, the rotating arm drive control device P drives this rigid assembly to rotate at a constant speed, causing the steel striking heads at both ends to cyclically and alternately strike the surface of the building's exterior wall, thereby generating sound waves. The acoustic sensor H is fixed near the striking heads K to collect the sound wave signals generated by the impact.

[0089] The rotating arm drive control device P further includes a speed controller. This speed controller receives control commands, adjusts and stabilizes the rotational speed of the rotating arm's motor, thereby precisely controlling the frequency and force of the steel hammer striking the exterior wall, ensuring the consistency of the excitation signal.

[0090] The positioning unit, used to accurately map the position of the hollow area, consists of multiple lateral identification tracks A fixedly connected to the guide unit and optical sensors G mounted on the robot's main body unit. When the robot's main body unit rises and falls, the optical sensors G calculate and determine the robot's absolute vertical coordinates in real time by detecting their relative position to each lateral optical sensor identification track. This high-precision position information is synchronously correlated with the acoustic data collected by the detection unit.

[0091] To address the complex environment at high altitudes and enhance system safety, this invention also includes an adsorption mechanism. This mechanism employs a negative pressure adsorption device, which includes at least one negative pressure fan E. The negative pressure fan E is embedded in the front of the robot's main body unit opposite the building's exterior wall. Upon activation, the fan generates continuous negative pressure, firmly securing the robot's main body unit and guide unit against the wall. Furthermore, an electromagnetic adsorption device can be used for assistance. This device includes an electromagnet M mounted on the robot's main body unit, used for attaching and fixing to any position on the vertical guide rail.

[0092] The robot's main body is equipped with a rechargeable and detachable battery pack, which provides centralized power to the detection unit, positioning unit, and negative pressure adsorption mechanism.

[0093] The detachable data storage device is located inside the protective compartment of the robot's main unit. It is used to store the acoustic waveform data collected by the acoustic sensors and their synchronously associated absolute position information. After the detection task is completed, the storage device can be removed from the robot's main unit and connected to a computer for offline data processing, analysis, and report generation.

[0094] The detection accuracy of this invention is based on reliable acoustic signal acquisition and scientific comparative analysis. Its core lies in using a pre-constructed database of acoustic characteristics of external wall hollowness as a comparison benchmark to achieve objective identification of hollowness.

[0095] The database is built using the following steps:

[0096] 1) Sample collection: In the laboratory or typical engineering site, for normal walls and hollow walls with known conditions (confirmed by drilling method, etc.), use standard excitation device to conduct tests and simultaneously collect their dynamic response signals (including but not limited to sound wave and vibration signals).

[0097] 2) Feature Extraction and Storage: The collected dynamic response signals are processed and analyzed to extract and quantify one or more feature parameters that can characterize the wall condition. These feature parameters constitute the basis for distinguishing between wall hollowness and normal conditions.

[0098] 3) Database content: Finally, the typical signal characteristics of walls in different states are associated with their corresponding explicit state labels to form the feature database.

[0099] In the actual testing process, the following procedure is followed:

[0100] 1) Real-time feature extraction: The real-time acoustic signal collected by the detection unit will undergo the same analysis process to extract its key acoustic features.

[0101] 2) Feature comparison and discrimination: The features extracted in real time are compared with the typical feature parameter ranges or waveform templates of different states stored in the feature database.

[0102] 3) Generate diagnostic results: Based on the comparison results and according to preset criteria (e.g., if the feature value of the real-time signal falls within the "hollow" feature range, it is determined to be a hollow), the status assessment of the current detection point is output. This status information is stored together with the precise location information, and finally a visualized hollow defect distribution map is generated.

[0103] Example:

[0104] The on-site deployment and workflow of this invention are as follows:

[0105] (I) System Deployment Phase:

[0106] The operators first assemble the guide unit on the building's roof. The segmented alloy guide rails are then connected segment by segment using pin connectors to form a complete unit of the required length. This segmented, detachable design breaks down the long guide rails into short, easily transportable rods, perfectly solving the transportation and access problems of ultra-high-rise building inspection equipment, which is often impassable by elevators or stairs. The assembled ladder-shaped guide unit is then lowered from the roof, relying on its own weight to descend vertically to the ground. After adjusting its position, the top of the guide rail is fixed to the roof's embedded parts using pins, and the bottom is inserted into a pre-set positioning base on the ground, completing the rigid installation of the entire system.

[0107] (II) Robot Deployment and Testing Phase:

[0108] Align the T-shaped track slot at the bottom of the robot's main body unit with the T-shaped guide rail of the guide unit and engage it. After connecting the traction rope, the operator starts the negative pressure fan and detection system on the roof.

[0109] To further optimize the safety of the traction unit, guide pulleys made of polymer materials can be installed at the edge of the building roof. The traction rope changes direction through these pulleys; this effectively reduces friction between the rope and the roof edge, prevents wear, and extends the rope's lifespan. Simultaneously, the free end of the traction rope should be wound onto a mechanical or electric winch with a self-locking function. Operators control the winch to raise and lower the traction rope, saving effort and ensuring reliable suspension at any position, greatly improving operational safety and controllability.

[0110] The operator controls the winch to guide the robot down the guide rail at a suitable speed. During the descent, the speed controller controls the rotating arm's motor to operate at a constant speed, driving the striking mechanism to repeatedly strike the wall. Acoustic sensors simultaneously collect sound signals, and optical sensors calculate the robot's precise height coordinates in real time. The coordinates and sound wave data are simultaneously stored in a detachable data storage device on the robot body.

[0111] (III) Data Collection and Analysis Phase:

[0112] After completing the inspection of a vertical area, the operator retracts the robot, moves the guide unit to an adjacent area, and repeats the above process. Once all inspections are complete, the data storage device is removed from the robot. The spatiotemporal correlation dataset within the storage device provides the data foundation for implementing the hollow area localization and detection method. Based on this dataset, further feature comparison can be used to generate a hollow area defect distribution map with precise coordinates and an inspection report.

[0113] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A robot for detecting hollow areas in exterior walls, characterized in that, It includes the robot's main body unit, guiding unit, traction unit, detection unit, and positioning unit; The guide unit is configured as a rigid guide frame structure that can be fixedly installed on the exterior wall of the building; The robot body unit is configured to be movably mounted on the guide unit; The traction unit is configured to drive the robot body unit to move up and down along the guide unit via manual operation or motor drive; The detection unit is mounted on the main body of the robot and collects the sound wave signals from the outer wall by tapping. The positioning unit is configured to determine the real-time position of the robot body unit on the guide unit and associate the position information with the acoustic signal.

2. The robot for detecting hollow areas in exterior walls according to claim 1, characterized in that, The guide unit includes two parallel vertical guide rails, which are segmented and connected in a detachable manner. The vertical guide rail of the guide unit has a T-shaped cross section; the track inlay groove of the robot body unit is a T-shaped groove adapted to the T-shaped vertical guide rail cross section; The T-shaped groove has rubber sealing edges on both sides of the opening facing the outside of the guide rail. When the robot body unit is installed on the guide rail, the rubber sealing edges can generate elastic deformation to achieve tight fit and assist in guidance.

3. The robot for detecting hollow areas in exterior walls according to claim 1, characterized in that, The traction unit includes a traction rope mounted on the robot's main body unit. The fixed end of the traction rope is fixedly connected to the robot's main body unit, and the free end of the traction rope extends to the building's roof. Lifting and lowering are achieved by manual operation or by a traction motor.

4. The robot for detecting hollow areas in exterior walls according to claim 1, characterized in that, The detection unit includes a rotating arm drive control device, a rotating arm, a striking rod, a striking head, and an acoustic sensor; the striking rod is made of elastic material and its middle part is fixedly installed at the front end of the rotating arm; the striking head is made of rigid material and is disposed at the end of the striking rod; the drive control device drives the rotating arm and the striking rod to rotate, so that the striking head strikes the exterior wall, and the acoustic sensor is used to collect the generated sound wave signal.

5. The striking heads are arranged in pairs at both ends of the elastic striking rod; during operation, the rotating arm drive control device drives the rotating arm and the striking rod to rotate together, so that the rigid striking heads at both ends alternately hammer the surface of the exterior wall. The rotating arm drive control device includes a rotating arm rotary motor and a speed controller; the speed controller is used to adjust the motor speed to control the tapping frequency, and is connected to the positioning unit to coordinate the tapping frequency with the robot's moving speed.

6. The robot for detecting hollow areas in exterior walls according to claim 2, characterized in that, The positioning unit includes an optical sensor mounted on the robot's main body and multiple transverse identification tracks positioned between two parallel vertical guide rails. The multiple transverse identification tracks are equally spaced and have position markers for optical sensor identification.

7. The robot for detecting hollow areas in exterior walls according to claim 1, characterized in that, It also includes a detachable data storage device disposed on the main body unit of the robot for storing the acoustic signal and its associated location information.

8. The robot for detecting hollow areas in exterior walls according to claim 2, characterized in that, It also includes an adsorption mechanism; the adsorption mechanism adopts a negative pressure adsorption device and / or an electromagnetic adsorption device, the negative pressure adsorption device includes a negative pressure fan disposed on the robot body unit, used to generate negative pressure to make the robot stick to the wall; the electromagnetic adsorption device includes an electromagnet disposed on the robot body unit, used to attach and fix it to any position of the vertical guide rail.

9. The robot for detecting hollow areas in exterior walls according to claim 1, characterized in that, The robot's main body unit is equipped with a removable battery for powering electrical components.

10. A method for locating and detecting hollow areas in exterior walls, characterized in that, Includes the following steps: Step 1) Assemble the guide unit on the roof of the building. Lower the assembled guide unit from top to bottom along the exterior wall of the building under the action of gravity, and fix the top and bottom of the guide unit to the corresponding positions on the building. Step 2) Install the traction unit onto the robot body unit, and then install the robot body unit onto the guide unit; Step 3) Control the traction unit to move the robot body unit along the guide unit. During the movement of the robot body unit, control the excitation mechanism of the detection unit to perform periodic mechanical excitation on the outer wall surface. Step 4) Perform the following operations simultaneously: 4.1) Acquire the acoustic response signal generated by the mechanical excitation using an acoustic sensor; 4.2) The positioning unit obtains the real-time position information of the robot body unit on the guiding unit by reading the absolute position reference mark pre-set on the guiding unit; Step 5) Bind the acoustic response signal with the corresponding real-time location information, that is, align and store the acoustic response signal and location information with a unified timestamp to generate a spatiotemporal related dataset; Step 6) Based on the spatiotemporal correlation dataset, analyze and locate the hollow defects of the exterior wall.

11. The method for locating and detecting hollow areas in exterior walls according to claim 9, characterized in that, The specific steps of step 6) are as follows: 6.1) Analog-to-electrical signal conversion: Converting sound wave signals into continuous-time domain electrical signals, where the electrical signals satisfy the following relationship: In the formula: This represents an electrical signal, where A represents the amplitude, f represents the frequency, and Φ represents the phase. 6.2) Sampling, filtering, and quantization processing: According to the preset sampling frequency... The continuous electrical signal is sampled to determine the sampling period. And satisfy the sampling frequency ≥2 ,in The highest frequency of the signal; frequencies higher than the Nyquist frequency are filtered out using a low-pass filter. The signal components are analyzed; the sampled discrete signal values ​​are mapped to finite-precision digital quantities for quantization, thus completing the conversion from analog to digital signals. 6.3) Signal Refinement Processing and Feature Parameter Processing: The quantized discrete digital signal is processed by frame segmentation, dividing the continuous data into several short time frames, each containing a preset number of sampling points; Windowing is applied to each frame of the signal; the window function is... Using Hanming windows, the following relational expression is satisfied: in, This represents the discrete-time signal obtained after analog-to-digital conversion and processing, where N represents the frame length. The signal after windowing is: To reduce spectrum leakage caused by signal truncation; Each frame of the windowed signal is processed by fractional Fourier transform (FRFT). By adjusting the transform order parameter between 0 and 1, the gradual characteristics of the signal from the time domain to the frequency domain are obtained to assist in the main frequency calibration. The frequency domain representation is obtained by calculating the discrete Fourier transform of each frame of the signal: ,k=0,1,2…,N-1; Combined with frequency resolution: Determine the main frequency of the signal ; The power value is obtained by squaring the frequency domain representation, and the amplitude spectrum is calculated to extract the amplitude data of each frame signal. Wavelet transform is used to denoise the signal. By selecting the optimal threshold β, wavelet coefficients with amplitudes higher than β are retained, while wavelet coefficients with amplitudes lower than β are set to zero, thereby achieving signal-to-noise separation. The threshold satisfies the following relationship: in, For adjustment coefficients and 0 ≤ ≤ 1, Represents wavelet coefficients; Finally, the discrete amplitude sequence and the signal's dominant frequency are extracted from the processed signal. Data on the start-end time interval of each frame signal and the signal amplitude attenuation trend; 6.4) Multidimensional determination of hollow sound: Based on the feature parameters extracted in step 6.3), cross-analysis is performed in combination with the single-pulse acoustic echo signal model; The model expression for the pulse acoustic echo signal is: in, The amplitude parameter represents the signal, corresponding to the peak value of the discrete amplitude sequence; This represents the attenuation coefficient, obtained by fitting the amplitude attenuation trend. Represents a time variable. This represents the delay parameter, corresponding to the sound wave propagation time; This indicates the signal frequency, which is consistent with the extracted main frequency. Anomaly detection is performed on the amplitude of the detection points: In the formula, Indicates the average amplitude. This represents the critical value of the amplitude, and n is the number of detection points; when both of the above formulas meet the conditions, it is determined that the amplitude is abnormal. Anomaly detection is performed on the slope of the detection points: In the formula: and These represent the sound wave propagation times at the i-th and i-1-th detection points, respectively. and These represent the depths of point i and point i-1, respectively. When the slope changes beyond the preset range, it is determined to be an abnormal slope. When a certain area simultaneously meets the criteria of amplitude anomaly and slope anomaly, the area is determined to be a hollow area of ​​the wall.

Images