Active safety detection system for glass curtain wall service status

Abstract

The invention provides an active safety detection system for the service status of a glass curtain wall, comprising a detection device, wherein the detection device comprises a camera, a laser stroboscopic lighting box, a laser controller, an electronic control box, an industrial computer, a piezoelectric ceramic, a first lens, a second lens, a dichroic prism, a first reflector, a second reflector and a third lens; the first reflector forms a certain small angle with an optical axis; the electronic control box outputs a sinusoidal signal modulation signal to generate an ultrasonic wave to excite a wall, so that a glass curtain wall to be tested installed on the wall forms vibration; the laser controller is used to control the laser stroboscopic lighting box to generate a stroboscopic light with the same frequency as the vibration formed on the glass curtain wall to be tested to illuminate the glass curtain wall to be tested; light reflected by the glass curtain wall to be tested passes through a 4f system formed by the first lens and the second lens, and is incident on the dichroic prism and then is divided into two first and second sub-beams; the first and second sub-beams are respectively reflected by the first and second reflectors and converged through the third lens to interfere.

Description

Technical Field

[0001] The present invention relates to the field of detection technology, and in particular to an active safety detection system for the service status of a glass curtain wall based on intelligent vision and big data. Background Art

[0002] Glass curtain wall refers to the building's external protective structure or decorative structure that has a certain displacement capacity relative to the main structure and does not share the effects on the main structure. The wall has two types of glass: single-layer and double-layer. Many high-rise buildings currently use glass curtain walls.

[0003] However, these glass curtain walls have many potential safety hazards, and currently a technology that can perform safety inspections on glass curtain walls is needed. Summary of the invention

[0004] A brief overview of the present invention is provided below in order to provide a basic understanding of certain aspects of the present invention. It should be understood that this overview is not an exhaustive overview of the present invention. It is not intended to identify key or important parts of the present invention, nor is it intended to limit the scope of the present invention. Its purpose is merely to present certain concepts in a simplified form as a prelude to a more detailed description discussed later.

[0005] The present invention provides an active safety detection system for the service status of a glass curtain wall, so as to at least solve the above-mentioned problems in the prior art.

[0006] The present invention provides an active safety detection system for the service status of a glass curtain wall, the active safety detection system comprises a detection device, the detection device comprises a camera, a laser stroboscopic lighting box, a laser controller, an electronic control box, an industrial computer, a piezoelectric ceramic, a first lens, a second lens, a dichroic prism, a first reflector, a second reflector and a third lens; the electronic control box is used to output a sinusoidal signal modulation signal to generate an ultrasonic wave to excite a wall, so that a glass curtain wall to be tested installed on the wall forms vibration; the laser controller is used to control the laser stroboscopic lighting box to generate a stroboscopic light with the same frequency as the vibration formed on the glass curtain wall to be tested to illuminate the glass curtain wall to be tested; the first lens and the second lens form a 4f system, so that the light reflected by the glass curtain wall to be tested passes through the 4f system and is incident on the dichroic prism and is then divided into a first sub-beam and a second sub-beam; the first sub-beam After the first sub-beam is reflected by the beam splitting surface of the beam splitting prism, it is incident on the first reflector along a direction forming a preset angle with the mirror normal direction of the first reflector; after the second sub-beam is transmitted through the beam splitting surface of the beam splitting prism, it is incident on the second reflector along the mirror normal direction of the second reflector; wherein the piezoelectric ceramic is used to connect the second reflector to control the second reflector to generate a preset positive displacement or negative displacement in the direction along the optical axis; after the first sub-beam reflected back from the first reflector is transmitted through the beam splitting prism, it is output to the camera after passing through the third lens; after the second sub-beam reflected back from the second reflector is reflected through the beam splitting prism, it is output to the camera after passing through the third lens, and converges with the first sub-beam output to the camera and interferes, and the corresponding interference image is received by the camera, so as to send the interference image to the industrial computer through the camera.

[0007] Preferably, the preset angle is an angle greater than 0 and less than 30 degrees; the preset angle is 2 degrees, 5 degrees or 10 degrees.

[0008] Preferably, the detection device further comprises a sound controller and a transmitter; the electronic control box is connected to the sound controller, and the electronic control box controls the ultrasonic waves generated by the sound controller and the transmitter to excite the wall.

[0009] Preferably, the detection device further comprises an infrared heat sensing device for sensing the glass curtain wall to be detected to obtain heat sensing data.

[0010] Preferably, the industrial computer is used to perform the preprocessing on the obtained interference image.

[0011] Preferably, the preprocessing includes filtering and noise reduction.

[0012] Preferably, the industrial computer is also used to identify and analyze the preprocessed interference image using a pre-built glass curtain wall detection data recognition algorithm model.

[0013] Preferably, the glass curtain wall detection data recognition algorithm model is obtained by obtaining a large amount of training data in advance and training the model through a neural network.

[0014] Preferably, the active safety detection system comprises a plurality of detection devices, and the plurality of detection devices are used to be arranged at a plurality of detection points to obtain detection data of the glass curtain wall corresponding to the detection points, and the detection data at least comprises a corresponding interference image.

[0015] Preferably, the training data has corresponding labels, which include the type of loss and the level of loss; the types of loss are divided into cracks, pits, scratches, breakages and stains; each type of loss is divided into level 0, level 1, level 2 and level 3, level 0 means no loss, and levels 1-3 mean that the loss is getting more and more serious.

[0016] The active safety detection system for the service status of a glass curtain wall of the present invention can measure the largest possible area in the detection of the service status of a building glass curtain wall, which is helpful to find abnormal vibration distribution (usually corresponding to defects) in a large area, and also improves the scanning speed and greatly improves the work efficiency. The time required for measurement is shorter than that of the traditional time averaging method.

[0017] The active safety detection system for the service status of the glass curtain wall of the embodiment of the present invention integrates a 4f system, has a simple structure, a stable optical path, and a large field of view, and is more suitable for the active safety detection system for the service status of the glass curtain wall of a building.

[0018] These and other advantages of the present invention will become more apparent through the following detailed description of the best embodiments of the present invention in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The present invention may be better understood by referring to the following description in conjunction with the accompanying drawings, wherein the same or similar reference numerals are used throughout the drawings to represent the same or similar components. The accompanying drawings, together with the following detailed description, are included in and form a part of this specification and are used to further illustrate the preferred embodiments of the present invention and to explain the principles and advantages of the present invention. Among them:

[0020] Figure 1 is a schematic diagram showing an exemplary structure of an active safety detection system for a glass curtain wall service state of the present invention;

[0021] Figure 2 is a schematic diagram showing another exemplary structure of an active safety detection system for a glass curtain wall service state of the present invention;

[0022] Figure 3is a schematic diagram showing an example structure of a preferred embodiment of an active safety detection system for a glass curtain wall service state of the present invention;

[0023] Figure 4 is a schematic diagram of laser stroboscopic lighting;

[0024] Figure 5 It is the image of the object being measured;

[0025] Figure 6 is the obtained interference pattern;

[0026] Figure 7 It is a real-time subtraction fringe pattern;

[0027] Figure 8 is a schematic diagram showing a 4f optical system consisting of a first lens and a second lens;

[0028] Fig. 9 It is a schematic diagram showing the field angle principle of the active safety detection system for the service status of the glass curtain wall according to the present invention.

[0029] Those skilled in the art will appreciate that the elements in the drawings are shown only for simplicity and clarity and are not necessarily drawn to scale. For example, the size of some elements in the drawings may be enlarged relative to other elements to help improve the understanding of the embodiments of the present invention. DETAILED DESCRIPTION

[0030] Exemplary embodiments of the present invention will be described below in conjunction with the accompanying drawings. For the sake of clarity and conciseness, not all features of the actual implementation are described in the specification. However, it should be understood that many implementation-specific decisions must be made in the process of developing any such actual implementation in order to achieve the developer's specific goals, such as meeting those constraints related to the system and business, and these constraints may vary from implementation to implementation. In addition, it should be understood that although the development work may be very complex and time-consuming, it is only a routine task for those skilled in the art who benefit from the content of the present invention.

[0031] It is also necessary to explain here that, in order to avoid obscuring the present invention due to unnecessary details, only the device structure and / or processing steps closely related to the scheme according to the present invention are shown in the accompanying drawings, while other details that are not closely related to the present invention are omitted.

[0032] The invention provides an active safety detection system for the service status of a glass curtain wall.

[0033] Combine the following Figure 1 An example of an active safety detection system for the service status of the above-mentioned glass curtain wall is described below.

[0034] In view of the current status of safety detection of the service status of existing glass curtain walls, the present invention proposes an active safety detection system for the service status of glass curtain walls. The interference image obtained by the system can be subsequently analyzed to obtain effective information on the service status of the glass curtain wall, such as whether it is safe, whether it is damaged, etc., to achieve safety detection.

[0035] like Figure 1 As shown, the active safety detection system for the service status of the glass curtain wall includes a detection device.

[0036] The detection device is a device for obtaining an interference image of the service status of a glass curtain wall.

[0037] The structure of the detection device can be, for example, Figure 1 As shown, it includes a camera 1, a laser stroboscopic lighting box 2, a laser controller 3, an electronic control box 4, an industrial computer 5, a piezoelectric ceramic 10, a first lens 11, a second lens 12, a beam splitter prism 13, a first reflector 14, a second reflector 15 and a third lens 16.

[0038] The electronic control box 4 is used to output a sinusoidal signal modulation signal to generate an ultrasonic wave to excite the wall 9, so that the glass curtain wall 8 to be tested installed on the wall 9 forms vibration.

[0039] The laser controller 3 is used to control the laser stroboscopic lighting box 2 to generate a stroboscopic light of the same frequency as the vibration formed on the glass curtain wall 8 to be tested to illuminate the glass curtain wall 8 to be tested; the first lens 11 and the second lens 12 form a 4f system, so that the light reflected by the glass curtain wall 8 to be tested passes through the 4f system and is incident on the beam splitter prism 13 to be divided into a first sub-beam and a second sub-beam.

[0040] After being reflected by the beam splitting surface of the beam splitting prism 13 , the first sub-beam is incident on the first reflector 14 along a direction forming a preset angle with the normal direction of the mirror surface of the first reflector 14 .

[0041] After being transmitted through the beam splitting surface of the beam splitting prism 13 , the second sub-beam is incident on the second reflector 15 along the normal direction of the mirror surface of the second reflector 15 .

[0042] The piezoelectric ceramic 10 is used to connect the second reflector 15 to control the second reflector 15 to generate a preset positive displacement or negative displacement in the direction along the optical axis.

[0043] The first sub-beam reflected from the first reflector 14 is transmitted through the beam splitter prism 13 and then through the third lens 16 and output to the camera 1 .

[0044] The second sub-beam reflected from the second reflector 15 is reflected by the beam splitter prism 13, and then output to the camera 1 after passing through the third lens 16, and converges with and interferes with the first sub-beam output to the camera 1, and the corresponding interference image is received by the camera 1, so as to send the interference image to the industrial computer 5 through the camera 1.

[0045] The preset angle may be an angle greater than 0 and less than 30 degrees.

[0046] For example, the preset angle is 2 degrees, 5 degrees, or 10 degrees.

[0047] In an example, the detection device may further include a sound controller 6 and a transmitter 7 .

[0048] like Figure 2 As shown, the electronic control box 4 is connected to the sound controller 6 , and the electronic control box 4 controls the ultrasonic waves generated by the sound controller 6 and the transmitter 7 to excite the wall 9 .

[0049] In this way, the industrial computer 5 can preprocess the obtained interference image (such as denoising, filtering, etc., or other preprocessing may also be included), and then use the pre-built glass curtain wall detection data recognition algorithm model to recognize and analyze the preprocessed interference image.

[0050] The glass curtain wall detection data recognition algorithm model can be obtained, for example, by obtaining a large amount of training data in advance and training the model through a neural network.

[0051] In one example, the detection device may further include an infrared heat sensing device, which may be used to sense the glass curtain wall 8 to be tested and obtain sensing data (such as infrared imaging, radiation and spectrum measurement, etc.).

[0052] In another example, the system may include multiple Figure 1 The detection device shown (optionally, each as Figure 1 The device shown can also be provided with one or more infrared thermal sensing devices. In this way, the above-mentioned devices are provided at multiple preset locations in the city. For example, the Internet of Things can be used to realize data acquisition and communication between multiple preset locations, thereby obtaining relevant data of the glass curtain walls at the above-mentioned multiple preset locations (corresponding interference images and sensor data, etc.). In this way, the obtained training data has corresponding labels, such as the type of loss and the level of loss, etc. For example, the type of loss can be divided into cracks, pits, scratches, breakages, stains, etc., and the level can be divided according to the severity, such as each type of loss can be divided into level 0, level 1, level 2 and level 3, level 0 means no loss, and levels 1-3 mean that the loss is getting more and more serious.

[0053] In this way, big data can be collected and processed in advance through technologies such as sensor networks, the Internet of Things, data preprocessing and big data management.

[0054] Next, a preferred embodiment of the present invention is described.

[0055] When obtaining training data, we can obtain training data under complex interference conditions such as haze, variable lighting, and occlusion, and obtain a glass curtain wall detection data recognition algorithm model under complex environments. Through deep learning algorithms, we can realize the detection and risk assessment of the service status of glass curtain walls. All glass curtain walls of buildings are regularly inspected and data analysis is performed to achieve big data security of all building curtain walls in the city, and realize active safety detection and prediction of the service status of building glass curtain walls.

[0056] Since the glass curtain wall of the building is very large, the field of view and the energy of laser stroboscopic lighting need to be considered when measuring large-area vibration. The existing measurement device has a relatively small field of view due to its structure and size, which also limits its working distance. Usually, the maximum field of view does not exceed 28°.

[0057] In addition to the field of view, large-area vibration measurement also requires attention to the lighting energy issue. On the one hand, because the use of stroboscopic lighting reduces the exposure of the digital camera, the laser needs to provide sufficient lighting energy; on the other hand, because the area of ​​the target to be measured is large, the laser is also required to provide sufficient lighting energy. These two factors require large-area measurement to require high-power laser lighting, and at the same time, it is necessary to ensure that the laser has good temporal coherence and spatial coherence.

[0058] In this way, quantitative measurement of the spatial gradient of vibration amplitude can be achieved, or non-destructive testing can be performed according to the vibration distribution. This method has full-field, non-contact, rapid, strong anti-interference ability, stable structure, low laser coherence requirements, and can be applied to non-destructive testing of glass curtain walls.

[0059] To solve this problem, we adopt Figure 3 The example structure shown measures defects in the glass curtain wall of a building, combined with Figure 3 To illustrate the structure and working principle of the system. Figure 3 In the dotted box, the detection device is included, wherein 30 includes Figure 1 or Figure 2 The camera 1, piezoelectric ceramic 10, first lens 11, second lens 12, beam splitter prism 13, first reflector 14, second reflector 15 and third lens 16 shown serve as a detection end.

[0060] It consists of five parts: a detection end 30, a laser stroboscopic lighting box 2, a laser controller 3, an electronic control box 4, and an industrial computer 5 (the industrial computer 5 is provided with a corresponding software module). During measurement, the electronic control box outputs a sinusoidal signal modulation signal to generate an ultrasonic wave to excite the wall 9, causing the glass curtain wall 8 to vibrate. Because the sound intensity changes in a sinusoidal wave, each point on the glass curtain wall also vibrates according to the sine law. If the surface conditions of the glass curtain wall are uniform, the distribution of the vibration amplitude of the glass curtain wall changes uniformly; if the glass curtain wall has defects such as cracks and damage, the vibration amplitude of the wall position corresponding to the defect is obviously inconsistent with the surrounding area. Through the adjustment of the synchronous signal generator in the electronic control box, the laser stroboscopic lighting box generates a stroboscopic lighting wall with the same frequency as the vibration, and then the detection end 30 records the obtained interference pattern in the computer (industrial computer 5). Through the stroboscopic lighting method, the detection end 30 can obtain the spatial gradient of the vibration amplitude distribution of the glass curtain wall, and the distribution of the vibration amplitude can be obtained after integration. By observing and analyzing the shape of the speckle fringes related to the amplitude distribution, the location and size of the glass curtain wall defects can be obtained.

[0061] Combine the following Figure 4 To describe the above-mentioned time averaging method and stroboscopic lighting method.

[0062] The stroboscopic illumination method uses laser stroboscopic illumination, whose frequency is consistent with the vibration frequency, so that the illumination is locked on a certain phase of the vibration. The stroboscopic illumination method obtains phase diagrams in two different states (for example, state one where the glass curtain wall is intact, and state two where the glass curtain wall is damaged or cracked), and then subtracts them to obtain the phase difference distribution corresponding to the displacement spatial gradient distribution of the measured target. For example, when the excitation signal is at the peak or trough, the stroboscopic illumination method obtains the corresponding phase diagrams, and subtracts them to obtain the displacement spatial gradient distribution corresponding to the maximum change in vibration amplitude. Therefore, the stroboscopic illumination method realizes the quantitative measurement of the full-field vibration amplitude spatial gradient distribution.

[0063] The schematic diagram of stroboscopic lighting is given by Figure 4 Given. When the exciting vibration is a simple harmonic vibration, the position of any point on the surface of the target being measured will vibrate in a simple harmonic vibration, and the relationship between the off-plane displacement of the target being measured and time can be described by the sine wave shown in the figure. The stroboscopic lighting always occurs at a specific phase of the vibration, such as the position of the peak shown in the figure, assuming that its initial phase is α at this time. If the pulse of the stroboscopic lighting is narrow, then for the detection end 30, the object being measured is equivalent to a stationary target, and the phase shifting technology can be used to obtain a phase diagram corresponding to the position of the vibration peak. Then, at the next moment, the stroboscopic lighting is controlled to move to another phase of the vibration, such as Figure 4As shown in the figure, after the phase β is moved, it reaches the position of the trough. Similarly, the phase diagram corresponding to the position of the vibration trough can be obtained. The key to stroboscopic lighting is that the stroboscopic light and the vibration have the same frequency, so that the lighting can be locked on a specific phase of the vibration. In theory, the pulse width of the stroboscopic light should be as narrow as possible, and the change in the vibration phase should be as small as possible. If the pulse width is relatively wide, the resulting interference pattern is actually the integral of the vibration over a short period of time. Because the change in vibration is nonlinear, a stroboscopic pulse width that is too large will affect the final phase diagram. Too narrow stroboscopic lighting will result in insufficient brightness of the laser lighting, so it is necessary to select a suitable stroboscopic duty cycle based on actual conditions.

[0064] See also Figure 1 or Figure 2 The light reflected by the object passes through the detection end 30 (equivalent to Figure 1 or Figure 2 The light beams are divided into two beams at a certain angle when the camera 1, piezoelectric ceramic 10, first lens 11, second lens 12, dichroic prism 13, first reflector 14, second reflector 15 and third lens 16 are used. Each point on the object forms two images on the image plane; or each point on the image plane corresponds to two points on the object.

[0065] The laser light emitted by the laser can be irradiated to the surface to be measured after passing through a beam expander or other beam expander. The object light forms an image of mutual interference after passing through the detection end 30, and then forms an image on the target surface of the digital camera. Figure 1-3 As shown, in the detection end 30, there is a small angle between the first reflector 14 and the optical axis, which can be set based on experience or determined by an experimental method. The second reflector 15 is perpendicular to the optical axis and connected to the piezoelectric ceramic to play a phase shift role.

[0066] After passing through the beam splitter prism, the object light is split into two beams, which are then reflected by the first reflector 14 and the second reflector 15. However, the light beam reflected by the first reflector 14 will be deflected by a certain angle, and the light beam reflected by the phase shift mirror will be parallel to the optical axis. Finally, the two reflected light beams are converged by the beam splitter prism again, and after passing through the imaging lens, two mutually displaced interference images are formed on the camera target surface.

[0067] Combine the following Figure 5-Figure 7 An example of measurement using the above-mentioned apparatus will be described.

[0068] Figure 5 shows the object under test, Figure 6 is the interference pattern obtained using the device, Figure 7 is based on Figure 6 The real-time subtraction fringe diagram can be obtained by analyzing Figure 7 The deformation of the object is analyzed by using the stripes.

[0069] See also Figure 6 , where the two images are displaced in the X direction, and S is the magnitude of the displacement. Thus, by using the above-mentioned device of the embodiment of the present invention to actively detect the service status of the glass curtain wall of a building, it is possible to obtain Figure 7 As shown in the butterfly spot, a person skilled in the art can determine the defect position and size of the glass curtain wall based on the butterfly spot.

[0070] In addition, in this preferred embodiment, the field of view of the system is expanded by using a 4f system, such as Figure 8 As shown, f1 and f2 in the figure are respectively the focal lengths of lens L1 (i.e., the first lens described above) and lens L2 (i.e., the second lens described above), for example, f1=f2. In the active safety detection system for the service status of the glass curtain wall of the embodiment of the present invention, the image plane of the 4f system imaging lens coincides with the input plane, and the detection plane coincides with the camera target plane. In this way, the active safety detection system for the service status of the glass curtain wall of the embodiment of the present invention can have a larger field of view, Fig. 9 A part of the active safety detection system for the service status of the above-mentioned glass curtain wall is schematically shown, and it can be seen that the field of view of the system is relatively large.

[0071] The active safety detection system for the service status of glass curtain walls based on the stroboscopic lighting method can measure the largest possible area in the detection of the service status of the glass curtain walls of buildings, which helps to discover abnormal vibration distribution over a large area (usually corresponding to defects). It also improves the scanning speed and greatly improves work efficiency. The time required for measurement is shorter than that of the traditional time averaging method.

[0072] The active safety detection system for the service status of the glass curtain wall of the embodiment of the present invention integrates a 4f system, has a simple structure, a stable optical path, and a large field of view, and is more suitable for the active safety detection system for the service status of the glass curtain wall of a building.

[0073] After the CCD camera obtains the image, it is saved in the computer, and then the image is used for the convolutional neural network to detect and grade the quality defects of the building's glass curtain wall. The deep neural network constructs a multi-layer network to represent the target in multiple layers, in order to represent the abstract semantic information of the data through multiple layers of high-level features and obtain better feature robustness.

[0074] In addition, the system can use a convolutional neural network (CNN) to implement model construction and detection, and has excellent performance in large-scale image processing. It includes a convolution layer and a pooling layer. For example, keras can be used as an advanced development API, which can convert known data into experimental results with minimal delay, and is implemented in an environment with a Tensorflow or caffe deep learning framework as the backend. The data used in the experiment is collected and made into a data set by the active safety detection system for the service status of the glass curtain wall of an embodiment of the present invention. The obtained pictures are read and separated according to training, verification, and test data, the neural network is trained, and the cost function is optimized using methods such as a cross entropy function to obtain the final active safety detection results of the service status of the glass curtain wall of the wall building.

[0075] In order to realize the active safety detection system of glass curtain wall service status based on intelligent vision and big data, the glass curtain wall wall instances that you want to detect can be collected and then quantified in a predetermined way. The wall data is obtained by reading the sample data of each city building, marking the defects, and then uploading it to the cloud data analysis center. The sample data is analyzed and learned using algorithms such as k-means, and the sample data is modeled and evaluated multiple times to determine the current defect status of the glass curtain wall. For example, this process can be performed over several weeks or months to collect examples of glass curtain wall walls, different lighting conditions, different times of the day, and different temperatures to create a diverse image database representing the wall. After obtaining and analyzing enough data, the system will perform secondary modeling on the glass curtain wall wall condition to inform the user whether the glass curtain wall wall has defects and the type of defects. By utilizing the active safety detection system for the service status of the glass curtain wall of the embodiment of the present invention, images of defects in the glass curtain wall can be collected on-site, and the defects in the glass curtain wall can be inferred without consuming a lot of manpower and material resources to collect wall information. This not only reduces the difficulty of project deployment, but also saves capital consumption. It has intelligent analysis and reminder functions and can effectively detect defects in the glass curtain wall.

[0076] In actual applications, the glass curtain wall defects of each city can be entered into the cloud data analysis center, and a database can be established for each city with a password set; after installing the wall diagnosis system interface and reading the glass curtain wall defect data according to the diagnosis system protocol, the data can be uploaded to the glass curtain wall defect database corresponding to each city in the cloud data analysis center.

[0077] A classification model is established for the glass curtain wall data obtained by the cloud data analysis center using deep learning algorithms, and the unbalanced data set is processed. Undersampling is used to balance the data set; avoid the phenomenon that the normal data ratio is too large and the abnormal data ratio is too small; a classification model is established for the data set to assist in analyzing the defects of the glass curtain wall; the data is labeled with two labels, normal and abnormal, and then divided into two parts in the same proportion: a data set with labels removed and a data set with labels retained; an anomaly detection mechanism based on algorithms such as k-means is used to cluster the unlabeled data set, and an anomaly detection mechanism based on the k-means algorithm is used to perform the same clustering analysis on the labeled data set, and the variance of the distance from other data points to the cluster center is compared to determine whether the point is in an abnormal state or a normal state. The unlabeled data set also needs to add various hybrid models in the later stage to calculate the probability of data being classified into different classes, so as to achieve fine classification of data points, obtain more accurate results, and avoid overfitting.

[0078] In addition, the active safety detection system for the service status of the glass curtain wall according to the embodiment of the present invention is used to detect defects of the glass curtain wall, collect its phase diagram, and perform operations such as filtering on the phase diagram. A prediction model is established for real-time data sets to provide technical support for early warning. The abnormal data points that appear on the wall are used to characterize the state of wall defects, and the autoencoder is used to detect abnormal information. A convolutional neural network is used to detect and grade the quality defects of the glass curtain wall. This is implemented in an environment where keras and other advanced development APIs are used and Tensorflow or caffe deep learning frameworks are used as the backend. The data is analyzed and the cost function is optimized using methods such as the cross entropy function to obtain a wall defect model. The model in the autoencoder has the functions of compressing and restoring data, and the compression and recovery are mainly for the feature data in the hidden layer. The extracted feature data will be passed through the sigmoid function to realize the discretization of the glass curtain wall defect data, thereby meeting the use conditions of the K-means algorithm. Each attribute weight is assigned through a neural network to overcome the disadvantage of equal weights of each attribute in the K-means algorithm. The defects of the autoencoder glass curtain wall can be re-modeled to give early warnings to defective glass curtain walls, and targeted repairs can be made later, greatly improving the level of personalized intelligent early warnings.

[0079] In addition, the data analyzed can be stored in the database of the corresponding city's glass curtain wall conditions through the cloud data analysis center, and the analysis results can be sent to the user's mobile terminal. When viewing, the password set in step 1 must be entered to ensure data security; the mobile terminal informs the user of the current glass curtain wall status based on the data, prompts dangerous glass curtain wall walls, and realizes real-time detection. Finally, establish measurement indicators; quantify the performance of the model, verify the correctness, stability and reliability of the model, and assist in subsequent improvements.

[0080] In practical applications, the above-mentioned system of the present invention can also comprehensively obtain the service status data of the glass curtain wall with multi-sensor information such as infrared thermal sensing technology, and perform preliminary collection and preliminary processing and screening of big data through technologies such as sensor networks, the Internet of Things, data preprocessing and big data management. Analyze complex interference conditions such as haze, variable illumination, and occlusion, and realize the detection and risk assessment of the service status of the glass curtain wall through deep learning algorithms. Regularly detect and analyze all glass curtain walls of buildings to achieve big data security of all building curtain walls in the city, and realize active safety detection and prediction of the service status of building glass curtain walls. It uses automated optimization algorithms to provide service warning and hidden danger mining methods and technologies for glass curtain walls based on security domains and intelligent models. For example, a prototype can be developed and tested on site in accordance with relevant standards and specifications, so as to improve the real-time and predictive level of online health diagnosis of urban building glass curtain walls.

[0081] Although the present invention has been described according to a limited number of embodiments, it will be apparent to those skilled in the art, with the benefit of the above description, that other embodiments may be envisioned within the scope of the invention thus described. In addition, it should be noted that the language used in this specification is selected primarily for readability and teaching purposes, rather than for explaining or defining the subject matter of the present invention. Therefore, many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the appended claims. The disclosure of the present invention is illustrative, not restrictive, with respect to the scope of the present invention, which is defined by the appended claims.

Claims

1. Active safety detection system for glass curtain wall service status, characterized by: The active safety detection system comprises a detection device, wherein the detection device comprises a camera (1), a laser stroboscopic lighting box (2), a laser controller (3), an electronic control box (4), an industrial computer (5), a piezoelectric ceramic (10), a first lens (11), a second lens (12), a beam splitter prism (13), a first reflector (14), a second reflector (15) and a third lens (16); The electronic control box (4) is used to output a sinusoidal signal modulation signal to generate an ultrasonic wave to excite the wall (9), so that the glass curtain wall (8) to be tested installed on the wall (9) vibrates; The laser controller (3) is used to control the laser stroboscopic lighting box (2) to generate stroboscopic light of the same frequency as the vibration formed on the glass curtain wall (8) to be tested, so as to illuminate the glass curtain wall (8) to be tested; the first lens (11) and the second lens (12) form a 4f system, so that the light reflected by the glass curtain wall (8) to be tested passes through the 4f system and is incident on the beam splitting prism (13) to be divided into a first sub-beam and a second sub-beam; After being reflected by the beam splitting surface of the beam splitting prism (13), the first sub-beam is incident on the first reflector (14) along a direction forming a preset angle with the normal direction of the mirror surface of the first reflector (14); After being transmitted through the beam splitting surface of the beam splitting prism (13), the second sub-beam is incident on the second reflector (15) along the normal direction of the mirror surface of the second reflector (15); The piezoelectric ceramic (10) is used to connect the second reflector (15) to control the second reflector (15) to generate a preset positive displacement or negative displacement in a direction along the optical axis; The first sub-beam reflected from the first reflector (14) is transmitted through the beam splitter (13) and then through the third lens (16) to be output to the camera (1); The second sub-beam reflected from the second reflector (15) is reflected by the beam splitter (13), and then output to the camera (1) after passing through the third lens (16), and converges with and interferes with the first sub-beam output to the camera (1), and the camera (1) receives a corresponding interference image, so that the interference image is sent to the industrial computer (5) through the camera (1); The industrial computer (5) is used to pre-process the obtained interference image.

2. The active safety detection system for the service status of a glass curtain wall according to claim 1, characterized in that: The preset angle is an angle greater than 0 and less than 30 degrees; the preset angle is 2 degrees, 5 degrees or 10 degrees.

3. The active safety detection system for the service status of the glass curtain wall according to claim 1 is characterized in that The detection device also includes a sound controller (6) and a transmitter (7); the electronic control box (4) is connected to the sound controller (6), and the electronic control box (4) controls the ultrasonic waves generated by the sound controller (6) and the transmitter (7) to excite the wall (9).

4. The active safety detection system for the service status of a glass curtain wall according to claim 1 is characterized in that The detection device also includes an infrared heat sensing device, which is used to sense the glass curtain wall to be detected and obtain heat sensing data.

5. The active safety detection system according to claim 1, characterized in that The preprocessing includes filtering and noise reduction.

6. The active safety detection system for the service status of a glass curtain wall according to claim 1, characterized in that: The industrial control computer (5) is also used to identify and analyze the preprocessed interference image using a pre-built glass curtain wall detection data identification algorithm model.

7. The active safety detection system for the service status of a glass curtain wall according to claim 6, characterized in that: The glass curtain wall detection data recognition algorithm model is obtained by obtaining a large amount of training data in advance and training the model through a neural network.

8. The active safety detection system for the service status of a glass curtain wall according to any one of claims 1 to 7, characterized in that: The active safety detection system includes a plurality of detection devices, which are used to be arranged at a plurality of detection points to obtain detection data of the glass curtain wall at the corresponding detection points, wherein the detection data at least includes a corresponding interference image.

9. The active safety detection system for the service status of a glass curtain wall according to claim 8, characterized in that: The training data has corresponding labels, and the labels include the type of loss and the level of loss; the type of loss is divided into cracks, pits, scratches, breakages and stains; Each loss is divided into level 0, level 1, level 2 and level 3. Level 0 means no loss, and levels 1-3 mean increasingly severe losses.

Images