Construction method and system of adjustable balance beam for prefabricated staircase lifting

Abstract

A construction method and system for lifting adjustable balance beams in prefabricated staircases are disclosed. This method employs machine vision-based artificial intelligence detection technology to extract central data from key frames in the lifting monitoring video of the adjustable balance beam. This focuses on the distribution of the beam's center of gravity, and further extracts the multi-scale correlation characteristics of the relative temporal changes in this center of gravity data. This information is then used to detect the lifting stability of the adjustable balance beam. This allows for accurate detection and judgment of the lifting stability, enabling the generation of an instability warning when instability is detected, thus ensuring construction safety.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of building construction, and more particularly, to a construction method and system of an adjustable balance beam for hoisting a fabricated staircase. BACKGROUND

[0002] In the process of on-site construction in the field of building, the staircases in buildings are mainly cast-in-place or steel structure. The construction method of cast-in-place staircases consumes a lot of resources such as formwork, scaffolding, energy and manpower, and the quality is uneven, the construction progress is slow, and the construction period is long. With the iteration of building construction methods, fabricated staircases are widely used in buildings. The existing fabricated staircases need to be split during installation, which raises many new technical problems.

[0003] For example, in the construction process of the adjustable balance beam for hoisting a fabricated staircase, the adjustable balance beam during hoisting is a source of danger, which may fall and cause damage and injury when it fluctuates.

[0004] Therefore, an optimized construction scheme of the adjustable balance beam for hoisting a fabricated staircase is expected. SUMMARY

[0005] To solve the above technical problems, the present application is proposed. The embodiments of the present application provide a construction method and system of an adjustable balance beam for hoisting a fabricated staircase, which adopts artificial intelligence detection technology based on machine vision to extract center data of each key frame in the hoisting monitoring video of the hoisted adjustable balance beam to focus on the center of gravity data distribution information of the hoisted adjustable balance beam, and extract the multi-scale correlation feature distribution information of the relative change characteristics of the center of gravity data in time sequence, so as to detect the hoisting stability of the hoisted adjustable balance beam. In this way, the hoisting stability of the hoisted adjustable balance beam can be accurately detected and judged, and an unstable warning prompt is generated when instability is detected, thereby ensuring the safety of construction.

[0006] According to one aspect of the present application, a construction method of an adjustable balance beam for hoisting a fabricated staircase is provided, which comprises:

[0007] obtaining a hoisting monitoring video of a hoisted adjustable balance beam;

[0008] extracting a plurality of hoisting monitoring key frames from the hoisting monitoring video;

[0009] extracting the center of gravity data of the hoisted adjustable balance beam in the plurality of hoisting monitoring key frames to obtain a sequence of center of gravity data;

[0010] differences between each two adjacent barycenter data in the sequence of the barycenter data to obtain a sequence of difference barycenter data, and arrange the sequence of the difference barycenter data as a difference barycenter data sequence vector;

[0011] pass the difference barycenter data sequence vector through a multi-scale neighborhood feature extraction module to obtain a multi-scale barycenter change feature vector;

[0012] correct a feature distribution of the multi-scale barycenter change feature vector based on the difference barycenter data sequence vector to obtain a classification feature vector; and

[0013] pass the classification feature vector through a classifier to obtain a classification result, the classification result being used to represent whether to generate a hoisting instability early warning prompt of the hoisted adjustable balance beam.

[0014] In the construction method of the fabricated stair hoisted adjustable balance beam, the extracting a plurality of hoisting monitoring key frames from the hoisting monitoring video comprises: extracting the plurality of hoisting monitoring key frames from the hoisting monitoring video at a predetermined sampling frequency.

[0015] In the construction method of the fabricated stair hoisted adjustable balance beam, the extracting barycenter data of the hoisted adjustable balance beam in the plurality of hoisting monitoring key frames to obtain a sequence of barycenter data comprises: converting the plurality of hoisting monitoring key frames from camera perspectives into global perspectives to obtain a plurality of mapped hoisting monitoring key frames; and extracting barycenter data of image frames of the plurality of mapped hoisting monitoring key frames as the barycenter data of the hoisted adjustable balance beam to obtain the sequence of barycenter data.

[0016] In the construction method of the fabricated stair hoisted adjustable balance beam, the passing the difference barycenter data sequence vector through a multi-scale neighborhood feature extraction module to obtain a multi-scale barycenter change feature vector comprises: inputting the difference barycenter data sequence vector into a first convolutional layer of the multi-scale neighborhood feature extraction module to obtain a first-scale barycenter change feature vector, wherein the first convolutional layer has a first one-dimensional convolutional kernel of a first length; inputting the difference barycenter data sequence vector into a second convolutional layer of the multi-scale neighborhood feature extraction module to obtain a second-scale barycenter change feature vector, wherein the second convolutional layer has a second one-dimensional convolutional kernel of a second length, and the first length is different from the second length; and concatenating the first-scale barycenter change feature vector and the second-scale barycenter change feature vector to obtain the multi-scale barycenter change feature vector.

[0017] In the construction method of the assembled stair lifting adjustable balance beam, the first convolution layer of the multi-scale neighborhood feature extraction module is used to one-dimensionally convolve and encode the difference center of gravity data sequence vector according to the following formula to obtain the first scale center of gravity change feature vector; wherein the formula is:

[0018]

[0019] Wherein a is the width of the first convolution kernel in the x direction, F(a) is the first convolution kernel parameter vector, G(x-a) is the local vector matrix operated with the convolution kernel function, w is the size of the first convolution kernel, and X represents the difference center of gravity data sequence vector.

[0020] In the construction method of the assembled stair lifting adjustable balance beam, the second convolution layer of the multi-scale neighborhood feature extraction module is used to one-dimensionally convolve and encode the difference center of gravity data sequence vector according to the following formula to obtain the second scale center of gravity change feature vector; wherein the formula is:

[0021]

[0022] Wherein b is the width of the second convolution kernel in the x direction, F(b) is the second convolution kernel parameter vector, G(x-b) is the local vector matrix operated with the convolution kernel function, m is the size of the second convolution kernel, and X represents the difference center of gravity data sequence vector.

[0023] In the construction method of the assembled stair lifting adjustable balance beam, the feature distribution of the multi-scale center of gravity change feature vector is corrected based on the difference center of gravity data sequence vector to obtain a classification feature vector, which includes: calculating a small-scale local derivative vector between the multi-scale center of gravity change feature vector and the difference center of gravity data sequence vector, the small-scale local derivative vector being related to the absolute value of the positional difference between the multi-scale center of gravity change feature vector and the difference center of gravity data sequence vector; and multiplying the multi-scale center of gravity change feature vector by the small-scale local derivative vector as a weighted feature vector to obtain the classification feature vector.

[0024] In the construction method of the assembled stair lifting adjustable balance beam, the calculation of the small-scale local derivative vector between the multi-scale center of gravity change feature vector and the difference center of gravity data sequence vector comprises: calculating the small-scale local derivative vector between the multi-scale center of gravity change feature vector and the difference center of gravity data sequence vector according to the following formula: wherein, the formula is:

[0025]

[0026] wherein v 1i , v 2i and v wi are the vector values of each position of the multi-scale center of gravity change feature vector, the difference center of gravity data sequence vector and the small-scale local derivative vector respectively, and |v 1i -v 2i | represents the absolute value of the difference between the corresponding positions of the vectors.

[0027] In the construction method of the assembled stair lifting adjustable balance beam, the classification feature vector is input into the classifier to obtain a classification result, and the classification result is used to indicate whether a lifting instability early warning prompt of the lifted adjustable balance beam is generated, which comprises: using a full connection layer of the classifier to perform full connection coding on the classification feature vector to obtain a coded classification feature vector; and inputting the coded classification feature vector into a Softmax classification function of the classifier to obtain the classification result.

[0028] According to another aspect of the present application, a construction system of an assembled stair lifting adjustable balance beam is provided, comprising:

[0029] A monitoring module is configured to acquire a lifting monitoring video of the lifted adjustable balance beam.

[0030] A sampling module is configured to extract a plurality of lifting monitoring key frames from the lifting monitoring video.

[0031] A center of gravity acquisition module is configured to extract center of gravity data of the lifted adjustable balance beam in the plurality of lifting monitoring key frames to obtain a sequence of the center of gravity data.

[0032] A difference and arrangement module is configured to calculate the difference between each two adjacent center of gravity data in the sequence of the center of gravity data to obtain a sequence of difference center of gravity data, and arrange the sequence of difference center of gravity data into a difference center of gravity data sequence vector.

[0033] A multi-scale coding module is configured to input the difference center of gravity data sequence vector into a multi-scale neighborhood feature extraction module to obtain a multi-scale center of gravity change feature vector.

[0034] a feature distribution correction module, configured to correct a feature distribution of the multi-scale center of gravity change feature vector based on the differential center of gravity data sequence vector to obtain a classification feature vector; and

[0035] a pre-warning result generation module, configured to pass the classification feature vector through a classifier to obtain a classification result, the classification result being used to represent whether to generate a hoisting instability pre-warning prompt of the hoisted adjustable balance beam.

[0036] In the construction system for hoisting the adjustable balance beam of the fabricated staircase, the sampling module is configured to extract the plurality of hoisting monitoring key frames from the hoisting monitoring video at a predetermined sampling frequency.

[0037] In the construction system for hoisting the adjustable balance beam of the fabricated staircase, the center of gravity acquisition module is further configured to: convert the plurality of hoisting monitoring key frames from camera perspectives into global perspectives to obtain a plurality of mapped hoisting monitoring key frames; and extract center of gravity data of image frames of the plurality of mapped hoisting monitoring key frames as the center of gravity data of the hoisted adjustable balance beam to obtain a sequence of the center of gravity data.

[0038] In the construction system for hoisting the adjustable balance beam of the fabricated staircase, the multi-scale coding module is further configured to: input the differential center of gravity data sequence vector into a first convolutional layer of the multi-scale neighborhood feature extraction module to obtain a first-scale center of gravity change feature vector, wherein the first convolutional layer has a first one-dimensional convolutional kernel with a first length; input the differential center of gravity data sequence vector into a second convolutional layer of the multi-scale neighborhood feature extraction module to obtain a second-scale center of gravity change feature vector, wherein the second convolutional layer has a second one-dimensional convolutional kernel with a second length, and the first length is different from the second length; and concatenate the first-scale center of gravity change feature vector and the second-scale center of gravity change feature vector to obtain the multi-scale center of gravity change feature vector.

[0039] In the construction system for hoisting the adjustable balance beam of the fabricated staircase, the feature distribution correction module is further configured to: calculate the small-scale local derivative vector between the multi-scale center of gravity change feature vector and the differential center of gravity data sequence vector according to the following formula:

[0040]

[0041] wherein v 1i , v 2i and v wi are vector values of respective positions of the multi-scale center of gravity change feature vector, the differential center of gravity data sequence vector and the small-scale local derivative vector respectively, and |v 1i -v2i | represents the absolute value of the difference between the corresponding positions of the vectors; and the multiscale center of gravity change feature vector is multiplied by the small-scale local derivative vector as a weighted feature vector to obtain the classification feature vector.

[0042] In the construction system of the assembled stair lifting adjustable balance beam, the early warning result generation module is further configured to: perform full connection coding on the classification feature vector using a full connection layer of the classifier to obtain an encoded classification feature vector; and input the encoded classification feature vector into a Softmax classification function of the classifier to obtain the classification result.

[0043] According to still another aspect of the present application, an electronic device is provided, comprising: a processor; and a memory having computer program instructions stored therein, which, when executed by the processor, cause the processor to perform the construction method of the assembled stair lifting adjustable balance beam as described above.

[0044] According to still another aspect of the present application, a computer readable medium is provided, having computer program instructions stored thereon, which, when executed by a processor, cause the processor to perform the construction method of the assembled stair lifting adjustable balance beam as described above.

[0045] Compared with the prior art, the construction method of the assembled stair lifting adjustable balance beam and the system thereof provided by the present application adopt artificial intelligence detection technology based on machine vision to focus on the center of gravity data distribution information about the hoisted adjustable balance beam by extracting the center data of each key frame in the hoisting monitoring video of the hoisted adjustable balance beam, and extract the multiscale correlation feature distribution information of the relative change characteristics of the center of gravity data in time sequence, so as to detect the hoisting stability of the hoisted adjustable balance beam. In this way, the hoisting stability of the hoisted adjustable balance beam can be accurately detected and judged, and an unstable early warning prompt is generated when instability is detected, thereby ensuring the safety of construction. BRIEF DESCRIPTION OF DRAWINGS

[0046] The above and other objects, features and advantages of the present application will become more apparent from the following detailed description of embodiments of the present application taken in conjunction with the accompanying drawings. The accompanying drawings are intended to provide a further understanding of embodiments of the present application and are incorporated in and constitute a part of this specification, illustrate embodiments of the present application and serve to explain the present application, but do not limit the present application. In the drawings, like reference numerals refer to like parts or steps throughout.

[0047] Figure 1 The application scenario diagram of the construction method of the assembled stair lifting adjustable balance beam according to the embodiments of the present application.

[0048] Figure 2 A flowchart of a construction method of an adjustable balance beam of an assembled stair hoist according to an embodiment of the present application.

[0049] Figure 3 An architectural diagram of a construction method of an adjustable balance beam of an assembled stair hoist according to an embodiment of the present application.

[0050] Figure 4 A flowchart of a construction method of an adjustable balance beam of an assembled stair hoist according to an embodiment of the present application.

[0051] Figure 5 A block diagram of a construction system of an adjustable balance beam of an assembled stair hoist according to an embodiment of the present application.

[0052] Figure 6 A block diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0053] Hereinafter, example embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only a part but not all of the embodiments of the present application, and thus the present application can not be limited to the described embodiments. It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory and are intended to provide further explanation of the application as claimed.

[0054] SUMMARY

[0055] As described in the above background, in the field construction process in the field of construction, the stairs in the building are mainly cast-in-place or steel structure, and the construction method of cast-in-place stairs consumes a lot of resources such as formwork, scaffold, energy and manpower, and the quality is uneven, the construction progress is slow, and the construction period is long. With the iteration of construction methods, assembled stairs are widely used in construction, and existing assembled stairs need to be split during installation, which raises many new technical problems.

[0056] For example, in the construction process of the adjustable balance beam of the assembled stair hoist, the adjustable balance beam in the hoisting process is a dangerous source, which may fall and cause damage and injury when it fluctuates. Therefore, an optimized construction scheme of the adjustable balance beam of the assembled stair hoist is expected.

[0057] At present, deep learning and neural networks have been widely used in computer vision, natural language processing, speech signal processing and other fields. In addition, deep learning and neural networks have also shown a level close to or even surpassing that of humans in image classification, object detection, semantic segmentation, text translation and other fields.

[0058] In recent years, the development of deep learning and neural networks provides a new solution for the intelligent monitoring of the adjustable balance beam of the assembled stair hoisting.

[0059] Correspondingly, in the actual construction process of the hoisting adjustable balance beam of the assembled stair, the hoisting process of the adjustable balance beam needs to be monitored to determine whether it is stable, which can be realized by analyzing and processing the monitoring video collected by the camera. However, when actually detecting the stability of the hoisted adjustable balance beam, since each image frame in the monitoring video has a large amount of information, and the position of the camera is different, causing the distribution of the image information about the hoisted adjustable balance beam to be different, which brings difficulties to the stability detection of hoisting.

[0060] Based on this, in the technical scheme of the present application, an artificial intelligence detection technology based on machine vision is used to focus on the center of gravity data distribution information of the hoisted adjustable balance beam in each key frame of the hoisting monitoring video of the hoisted adjustable balance beam by extracting the center data of each key frame, and to extract the multi-scale correlation feature distribution information of the relative change characteristics of the center of gravity data in time sequence, so as to detect the hoisting stability of the hoisted adjustable balance beam. In this way, the hoisting stability of the hoisted adjustable balance beam can be accurately detected and judged, and an unstable warning prompt can be generated when instability is detected, ensuring the safety of construction.

[0061] Specifically, in the technical scheme of the present application, first, the hoisting monitoring video of the hoisted adjustable balance beam is obtained. Then, considering that the dynamic change characteristics of the hoisted adjustable balance beam can be represented by the difference between adjacent monitoring frames in the hoisting monitoring video, that is, the dynamic change of the hoisted adjustable balance beam can be represented by the image representation of adjacent image frames, but considering that the difference between adjacent frames in the hoisting monitoring video is small, there is a lot of data redundancy, therefore, in order to reduce the measurement and avoid the adverse effects of data redundancy on detection, the hoisting monitoring video is sampled at a predetermined sampling frequency to extract a plurality of hoisting monitoring key frames. Here, it is worth mentioning that the sampling frequency can be adjusted based on the application requirements of the actual scene, rather than a default value.

[0062] Then, considering that the stability of the hoisted adjustable balance beam is related to the center of gravity of the balance beam when detecting the hoisting stability of the hoisted adjustable balance beam, the change information of the center of gravity feature distribution of the hoisted adjustable balance beam should be focused on when performing image feature extraction of the monitoring key frame, that is, the center of gravity data of the hoisted adjustable balance beam in the plurality of hoisting monitoring key frames is extracted to obtain a sequence of center of gravity data. Specifically, the plurality of hoisting monitoring key frames are respectively converted from a camera perspective to a global perspective to obtain a plurality of mapped hoisting monitoring key frames, and the center of gravity data of the image frame of the plurality of mapped hoisting monitoring key frames is extracted as the center of gravity data of the hoisted adjustable balance beam to obtain the sequence of center of gravity data.

[0063] Further, in order to mine the dynamic change features of the hoisting stability of the hoisted adjustable balance beam, the center of gravity change features of the hoisted adjustable balance beam need to be extracted. Considering that if the center of gravity change features of the hoisted adjustable balance beam are extracted in absolute change information, not only will the calculation amount be large, causing overfitting, but also the subtle change features of the center of gravity on a small scale will be difficult to be perceived, thereby affecting the accuracy of subsequent classification. Therefore, in the technical solution of the present application, it is expected to use the relative change features of the center of gravity of the hoisted adjustable balance beam to detect the hoisting stability of the balance beam. Specifically, first, the difference between each two adjacent center of gravity data in the sequence of center of gravity data is calculated to obtain a sequence of differential center of gravity data, and the sequence of differential center of gravity data is arranged as a differential center of gravity data sequence vector.

[0064] Then, considering that the change features of the center of gravity data of the hoisted adjustable balance beam under different time period spans have different mode information, in the technical solution of the present application, the differential center of gravity data sequence vector is subjected to feature mining through a multi-scale neighborhood feature extraction module to extract dynamic multi-scale neighborhood correlation change features of the center of gravity data of the hoisted adjustable balance beam under different time spans, thereby obtaining a multi-scale center of gravity change feature vector.

[0065] Then, the classification feature vector is further subjected to a classifier to obtain a classification result, which is used to represent whether to produce a hoisting instability warning prompt of the hoisted adjustable balance beam. That is, the classification feature vector having the dynamic multi-scale neighborhood correlation change features of the center of gravity data of the hoisted adjustable balance beam under different time spans is subjected to classification processing in a classifier to obtain a classification result for representing whether to produce a hoisting instability warning prompt of the hoisted adjustable balance beam. In this way, the hoisting stability of the hoisted adjustable balance beam can be accurately detected to produce an instability warning prompt when instability is detected.

[0066] In particular, in the technical solution of the present application, by passing the sequence of the differential barycenter data through the multi-scale neighborhood feature extraction module to obtain the multi-scale barycenter change feature vector, the neighborhood time sequence correlation features of the differential barycenter data sequence at multiple scales can be extracted. However, considering that the multi-scale neighborhood feature extraction module obtains neighborhood time sequence correlation features at relatively small scales, in order to make the distribution of the multi-scale barycenter change feature vector in the time sequence dimension as consistent as possible with the time sequence dimension distribution of the differential barycenter data sequence as the source data, it is expected to improve the consistency of the multi-scale barycenter change feature vector at small scales for the local correlation feature distribution of the differential barycenter data sequence.

[0067] Therefore, the small-scale local derivative vector of the multi-scale barycenter change feature vector, such as V1, and the differential barycenter data sequence vector, such as V2, is calculated as a weighted feature vector, denoted as:

[0068]

[0069] v 1i 、v 2i and v wi are the vector values of each position of the multi-scale barycenter change feature vector V1, the differential barycenter data sequence vector V2, and the small-scale local derivative vector, such as V w .

[0070] Here, by calculating the small-scale local derivative features between the multi-scale barycenter change feature vector V1 and the differential barycenter data sequence vector V2, the physical nature of mutual expression between data sequences can be simulated based on the geometric approximation of the corresponding positions between the multi-scale barycenter change feature vector V1 and the differential barycenter data sequence vector V2, thereby enhancing the local nonlinear dependence across feature domain positions with point-by-point regression between data sequences. In this way, by point multiplying the multi-scale barycenter change feature vector V1 with the weighted feature vector V w , the consistency of the multi-scale barycenter change feature vector V1 at small scales for the local correlation feature distribution of the differential barycenter data sequence can be improved, and the accuracy of the classification result of the multi-scale barycenter change feature vector V1 can be improved. In this way, the hoisting stability of the hoisted adjustable balance beam can be accurately detected and judged, and an instability warning prompt can be generated when instability is detected, thereby ensuring the safety of construction.

[0071] Based on this, this application proposes a construction method for lifting an adjustable balance beam in a prefabricated staircase, comprising: acquiring a lifting monitoring video of the adjustable balance beam being lifted; extracting multiple key frames from the lifting monitoring video; extracting the center of gravity data of the adjustable balance beam being lifted from the multiple key frames to obtain a sequence of center of gravity data; calculating the difference between every two adjacent center of gravity data in the sequence of center of gravity data to obtain a sequence of differential center of gravity data, and arranging the sequence of differential center of gravity data into a differential center of gravity data sequence vector; passing the differential center of gravity data sequence vector through a multi-scale neighborhood feature extraction module to obtain a multi-scale center of gravity change feature vector; correcting the feature distribution of the multi-scale center of gravity change feature vector based on the differential center of gravity data sequence vector to obtain a classification feature vector; and passing the classification feature vector through a classifier to obtain a classification result, wherein the classification result is used to indicate whether an instability warning for the lifting of the adjustable balance beam being lifted is generated.

[0072] Figure 1 This diagram illustrates an application scenario of the construction method for a prefabricated staircase with an adjustable balance beam, according to an embodiment of this application. Figure 1 As shown, in this application scenario, firstly, the camera (e.g., such as...) Figure 1 As shown in C), the adjustable balance beam being hoisted (e.g., as shown in C) is obtained. Figure 1 The hoisting monitoring video of B) shown in the diagram. Then, the hoisting monitoring video of the adjustable balance beam being hoisted is input to a server deployed with a construction algorithm for hoisting the adjustable balance beam of the prefabricated staircase (e.g., such as...). Figure 1 In the illustrated S), the server is able to process the hoisting monitoring video of the hoisted adjustable balance beam based on the construction algorithm of the prefabricated staircase hoisting adjustable balance beam, so as to obtain a classification result indicating whether a hoisting instability warning prompt is generated for the hoisted adjustable balance beam.

[0073] After introducing the basic principles of this application, various non-limiting embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0074] Exemplary methods

[0075] Figure 2 This is a flowchart illustrating the construction method of an adjustable balance beam for lifting a prefabricated staircase according to an embodiment of this application. Figure 2As shown, the construction method for lifting an adjustable balance beam for a prefabricated staircase according to an embodiment of this application includes: S110, acquiring a lifting monitoring video of the adjustable balance beam being lifted; S120, extracting multiple lifting monitoring key frames from the lifting monitoring video; S130, extracting the center of gravity data of the adjustable balance beam being lifted from the multiple lifting monitoring key frames to obtain a sequence of center of gravity data; S140, calculating the difference between every two adjacent center of gravity data in the sequence of center of gravity data to obtain a sequence of differential center of gravity data, and then combining the differential center of gravity data... The centroid data sequence is arranged into a differential centroid data sequence vector; S150, the differential centroid data sequence vector is processed by a multi-scale neighborhood feature extraction module to obtain a multi-scale centroid change feature vector; S160, based on the differential centroid data sequence vector, the feature distribution of the multi-scale centroid change feature vector is corrected to obtain a classification feature vector; and S170, the classification feature vector is processed by a classifier to obtain a classification result, the classification result being used to indicate whether an instability warning for the hoisted adjustable balance beam is generated.

[0076] Figure 3 This is a structural diagram of a construction method for a prefabricated staircase with an adjustable balance beam, according to an embodiment of this application. Figure 3 As shown, in this architecture, firstly, a hoisting monitoring video of the adjustable balance beam being hoisted is acquired, and multiple hoisting monitoring keyframes are extracted from the video. Next, the center of gravity data of the adjustable balance beam being hoisted is extracted from the multiple keyframes to obtain a sequence of center of gravity data. Then, the difference between every two adjacent center of gravity data points in the sequence is calculated to obtain a sequence of differential center of gravity data, and this sequence is arranged into a differential center of gravity data sequence vector. Furthermore, the differential center of gravity data sequence vector is processed through a multi-scale neighborhood feature extraction module to obtain a multi-scale center of gravity change feature vector. Next, based on the differential center of gravity data sequence vector, the feature distribution of the multi-scale center of gravity change feature vector is corrected to obtain a classification feature vector. Further, the classification feature vector is processed through a classifier to obtain a classification result, which indicates whether an instability warning for the hoisted adjustable balance beam is generated.

[0077] In step S110, the hoisting monitoring video of the hoisted adjustable balance beam is acquired. As described in the background, in the field construction process in the construction field, the stairs in the building are mainly cast-in-place or steel structure, and the construction method of cast-in-place stairs consumes a lot of resources such as formwork, scaffold, energy and manpower, and the quality is uneven, the construction progress is slow, and the construction period is long. With the iteration of the construction method, the fabricated stairs are widely used in the construction, and the existing fabricated stairs need to be split during installation, which causes many new technical problems. For example, in the hoisting process of the hoisted adjustable balance beam of the fabricated stairs, the adjustable balance beam in the hoisting process is a dangerous source, and when it fluctuates, it may fall and cause damage and injury. Therefore, an optimized construction scheme for hoisting the adjustable balance beam of the fabricated stairs is expected.

[0078] Correspondingly, in the actual construction process of hoisting the adjustable balance beam of the fabricated stairs, the adjustable balance beam in the hoisting process needs to be monitored to determine whether it is stable, which can be realized by analyzing and processing the monitoring video collected by the camera. However, when actually detecting and determining the stability of the hoisted adjustable balance beam, each image frame in the monitoring video has a large amount of information, and the position of the camera is different, causing the distribution of the image information about the adjustable balance beam in the hoisting process to be different, which brings difficulties to the stability detection of the hoisting.

[0079] Based on this, in the technical scheme of the present application, an artificial intelligence detection technology based on machine vision is used to focus on the center of gravity data distribution information about the hoisted adjustable balance beam in each key frame of the hoisted adjustable balance beam by extracting the center data of each key frame in the hoisted adjustable balance beam, and extracting the multi-scale correlation feature distribution information of the relative change characteristics of the center of gravity data in time sequence, so as to detect the stability of the hoisted adjustable balance beam. In this way, the hoisting stability of the hoisted adjustable balance beam can be accurately detected and determined, and an unstable warning prompt is generated when instability is detected, ensuring the safety of construction. Specifically, in the technical scheme of the present application, first, the camera acquires the hoisting monitoring video of the hoisted adjustable balance beam.

[0080] In step S120, a plurality of hoisting monitoring key frames are extracted from the hoisting monitoring video. In the process of monitoring the adjustable balance beam during the hoisting process, the dynamic change characteristics of the hoisted adjustable balance beam can be represented by the difference between adjacent monitoring frames in the hoisting monitoring video, that is, the dynamic change of the hoisted adjustable balance beam is represented by the image representation of adjacent image frames. However, considering that the difference between adjacent frames in the hoisting monitoring video is small, there is a lot of data redundancy, therefore, in order to reduce the calculation and avoid the adverse effects of data redundancy on detection, the hoisting monitoring video is sampled at a predetermined sampling frequency, thereby extracting a plurality of hoisting monitoring key frames from the hoisting monitoring video. It is worth mentioning that the sampling frequency can be adjusted based on the application requirements of the actual scene, rather than the default value.

[0081] In step S130, the center of gravity data of the hoisted adjustable balance beam in the plurality of hoisting monitoring key frames is extracted to obtain a sequence of center of gravity data. In the detection of the hoisting stability of the hoisted adjustable balance beam, since its stability is related to the center of gravity of the balance beam, when extracting the image features of the monitoring key frames, the focus should be on the change information of the center of gravity feature distribution of the hoisted adjustable balance beam, that is, the center of gravity data of the hoisted adjustable balance beam in the plurality of hoisting monitoring key frames is extracted to obtain a sequence of center of gravity data. Specifically, the plurality of hoisting monitoring key frames are respectively converted from the camera perspective to the global perspective to obtain a plurality of mapped hoisting monitoring key frames, and the center of gravity data of the image frames of the plurality of mapped hoisting monitoring key frames is extracted as the center of gravity data of the hoisted adjustable balance beam to obtain the sequence of center of gravity data.

[0082] In step S140, the difference between every two adjacent center of gravity data in the sequence of center of gravity data is calculated to obtain a sequence of differential center of gravity data, and the sequence of differential center of gravity data is arranged as a differential center of gravity data sequence vector. Further, in order to mine the dynamic change characteristics of the hoisting stability of the hoisted adjustable balance beam, it is necessary to extract the center of gravity change characteristics of the hoisted adjustable balance beam. Considering that if the center of gravity change characteristics of the hoisted adjustable balance beam are extracted in absolute change information, not only will the calculation amount be large, causing overfitting, but also the small-scale subtle change characteristics of the center of gravity will be difficult to be perceived, thereby affecting the accuracy of subsequent classification. Therefore, in the technical solution of the present application, it is expected to use the relative change characteristics of the center of gravity of the hoisted adjustable balance beam to detect the hoisting stability of the balance beam. Specifically, first, the difference between every two adjacent center of gravity data in the sequence of center of gravity data is calculated to obtain a sequence of differential center of gravity data, and the sequence of differential center of gravity data is arranged as a differential center of gravity data sequence vector.

[0083] In step S150, the differential centroid data sequence vector is processed by a multi-scale neighborhood feature extraction module to obtain a multi-scale centroid change feature vector. Considering that the centroid data of the hoisted adjustable balance beam exhibits different pattern information across different time spans, in this application's technical solution, the differential centroid data sequence vector is processed by a multi-scale neighborhood feature extraction module to extract the dynamic multi-scale neighborhood correlation change features of the hoisted adjustable balance beam's centroid data across different time spans, thereby obtaining a multi-scale centroid change feature vector.

[0084] Figure 4 This is a flowchart illustrating how the differential centroid data sequence vector is processed by a multi-scale neighborhood feature extraction module to obtain a multi-scale centroid change feature vector in the construction method of the prefabricated staircase lifting adjustable balance beam according to an embodiment of this application. Figure 4 As shown, the step of obtaining a multi-scale centroid change feature vector by passing the differential centroid data sequence vector through a multi-scale neighborhood feature extraction module includes: S210, inputting the differential centroid data sequence vector into the first convolutional layer of the multi-scale neighborhood feature extraction module to obtain a first-scale centroid change feature vector, wherein the first convolutional layer has a first one-dimensional convolutional kernel of a first length; S220, inputting the differential centroid data sequence vector into the second convolutional layer of the multi-scale neighborhood feature extraction module to obtain a second-scale centroid change feature vector, wherein the second convolutional layer has a second one-dimensional convolutional kernel of a second length, and the first length is different from the second length; and S230, concatenating the first-scale centroid change feature vector and the second-scale centroid change feature vector to obtain the multi-scale centroid change feature vector.

[0085] Specifically, in this embodiment, inputting the differential centroid data sequence vector into the first convolutional layer of the multi-scale neighborhood feature extraction module to obtain the first-scale centroid change feature vector includes: using the first convolutional layer of the multi-scale neighborhood feature extraction module to perform one-dimensional convolutional encoding on the differential centroid data sequence vector using the following formula to obtain the first-scale centroid change feature vector; wherein, the formula is:

[0086]

[0087] Where a is the width of the first convolution kernel in the x direction, F(a) is the parameter vector of the first convolution kernel, G(xa) is the local vector matrix operated with the convolution kernel function, w is the size of the first convolution kernel, and X represents the difference centroid data sequence vector.

[0088] Specifically, in the embodiment of the present application, the second convolutional layer of the multi-scale neighborhood feature extraction module is used to one-dimensionally convolve encode the difference center of gravity data sequence vector according to the following formula to obtain the second scale center of gravity change feature vector; wherein the formula is:

[0089]

[0090] wherein b is the width of the second convolution kernel in the x direction, F(b) is the second convolution kernel parameter vector, G(x-b) is the local vector matrix operated with the convolution kernel function, m is the size of the second convolution kernel, and X represents the difference center of gravity data sequence vector.

[0091] In step S160, the feature distribution of the multi-scale center of gravity change feature vector is corrected based on the difference center of gravity data sequence vector to obtain a classification feature vector. Specifically, in the embodiment of the present application, the feature distribution of the multi-scale center of gravity change feature vector is corrected based on the difference center of gravity data sequence vector to obtain a classification feature vector, first, a small-scale local derivative vector between the multi-scale center of gravity change feature vector and the difference center of gravity data sequence vector is calculated, the small-scale local derivative vector is related to the absolute value of the positional difference between the multi-scale center of gravity change feature vector and the difference center of gravity data sequence vector, and then the small-scale local derivative vector is used as a weighted feature vector to multiply the multi-scale center of gravity change feature vector by position to obtain the classification feature vector.

[0092] In particular, here, by passing the sequence of the difference center of gravity data through the multi-scale neighborhood feature extraction module to obtain the multi-scale center of gravity change feature vector, the neighborhood time sequence correlation features of the time sequence of the difference center of gravity data in multiple scales can be extracted. However, considering that the multi-scale neighborhood feature extraction module obtains neighborhood time sequence correlation features in relatively small scales, in order to make the distribution of the multi-scale center of gravity change feature vector in the time sequence dimension as consistent as possible with the time sequence dimension distribution of the difference center of gravity data sequence as the source data, it is expected to improve the consistency of the multi-scale center of gravity change feature vector in small scales for the local correlation feature distribution of the difference center of gravity data sequence. Therefore, the small-scale local derivative vector between the multi-scale center of gravity change feature vector and the difference center of gravity data sequence vector is calculated as a weighted feature vector.

[0093] Specifically, in the embodiment of the present application, the small-scale local derivative vector between the multi-scale center-of-gravity change feature vector and the difference center-of-gravity data sequence vector is calculated, comprising: calculating the small-scale local derivative vector between the multi-scale center-of-gravity change feature vector and the difference center-of-gravity data sequence vector according to the following formula: wherein, the formula is:

[0094]

[0095] wherein v 1i , v 2i and v wi are the vector values of each position of the multi-scale center-of-gravity change feature vector, the difference center-of-gravity data sequence vector and the small-scale local derivative vector respectively, and |v 1i -v 2i | represents the absolute value of the difference between the corresponding positions of the vectors.

[0096] Here, by calculating the small-scale local derivative feature between the multi-scale center-of-gravity change feature vector V1 and the difference center-of-gravity data sequence vector V2, the physical nature of mutual expression between data sequences can be simulated based on the geometric approximation of the corresponding positions between the multi-scale center-of-gravity change feature vector V1 and the difference center-of-gravity data sequence vector V2, thereby enhancing the local nonlinear dependence across the feature domain positions by position-by-position regression between data sequences. In this way, by point multiplying the multi-scale center-of-gravity change feature vector V1 with the weighted feature vector V w , the consistency of the multi-scale center-of-gravity change feature vector V1 with the local correlation feature distribution of the difference center-of-gravity data sequence on a small scale can be improved, and the accuracy of the classification result of the multi-scale center-of-gravity change feature vector V1 can be improved. In this way, the hoisting stability of the hoisted adjustable balance beam can be accurately detected and judged, and an unstable warning prompt can be generated when instability is detected to ensure the safety of construction.

[0097] In step S170, the classification feature vector is passed through a classifier to obtain a classification result, which is used to indicate whether to generate a hoisting instability warning prompt of the hoisted adjustable balance beam. That is, the classification feature vector having the dynamic multi-scale neighborhood correlation change feature of the center-of-gravity data of the hoisted adjustable balance beam under different time spans is classified in a classifier to obtain a classification result used to indicate whether to generate a hoisting instability warning prompt of the hoisted adjustable balance beam. Here, the high-dimensional data manifold of the classification feature vector is divided and determined by the classifier to obtain the classification result. In this way, the hoisting stability of the hoisted adjustable balance beam can be accurately detected, and an unstable warning prompt can be generated when instability is detected.

[0098] Specifically, in this embodiment of the application, the step of passing the classification feature vector through a classifier to obtain a classification result, wherein the classification result is used to indicate whether an unstable hoisting warning is generated for the hoisted adjustable balance beam, includes: using the fully connected layer of the classifier to perform fully connected encoding on the classification feature vector to obtain an encoded classification feature vector; and inputting the encoded classification feature vector into the Softmax classification function of the classifier to obtain the classification result.

[0099] In summary, the construction method for lifting adjustable balance beams for prefabricated staircases based on the embodiments of this application is explained. It employs machine vision-based artificial intelligence detection technology to extract central data from each key frame of the lifting monitoring video of the adjustable balance beam being lifted, focusing on the distribution information of the center of gravity data of the lifted adjustable balance beam. Furthermore, it extracts the multi-scale correlation feature distribution information of the relative change characteristics of these center of gravity data over time, thereby performing lifting stability detection of the adjustable balance beam. This allows for accurate detection and judgment of the lifting stability of the adjustable balance beam, generating an instability warning when instability is detected, ensuring construction safety.

[0100] Exemplary System

[0101] Figure 5 This is a block diagram of a construction system for an adjustable balance beam for lifting prefabricated stairs, according to an embodiment of this application. Figure 5 As shown, the construction system 100 for lifting an adjustable balance beam for a prefabricated staircase according to an embodiment of this application includes: a monitoring module 110 for acquiring a lifting monitoring video of the adjustable balance beam being lifted; a sampling module 120 for extracting multiple lifting monitoring key frames from the lifting monitoring video; a center of gravity acquisition module 130 for extracting the center of gravity data of the adjustable balance beam being lifted from the multiple lifting monitoring key frames to obtain a sequence of center of gravity data; and a difference and arrangement module 140 for calculating the difference between every two adjacent center of gravity data in the sequence of center of gravity data to obtain a sequence of differential center of gravity data, and arranging the difference... The sequence arrangement of the centroid data is a differential centroid data sequence vector; the multi-scale encoding module 150 is used to extract the differential centroid data sequence vector through the multi-scale neighborhood feature extraction module to obtain a multi-scale centroid change feature vector; the feature distribution correction module 160 is used to correct the feature distribution of the multi-scale centroid change feature vector based on the differential centroid data sequence vector to obtain a classification feature vector; and the early warning result generation module 170 is used to pass the classification feature vector through a classifier to obtain a classification result, the classification result being used to indicate whether an early warning prompt for hoisting instability of the hoisted adjustable balance beam is generated.

[0102] In an example, in the construction system 100 for lifting an adjustable balance beam of an assembled stair, the sampling module 120 is configured to extract the plurality of lifting monitoring key frames from the lifting monitoring video at a predetermined sampling frequency.

[0103] In an example, in the construction system 100 for lifting an adjustable balance beam of an assembled stair, the center of gravity obtaining module 130 is further configured to: convert the plurality of lifting monitoring key frames from camera perspectives to global perspectives respectively to obtain a plurality of mapped lifting monitoring key frames; and extract center of gravity data of image frames of the plurality of mapped lifting monitoring key frames as the center of gravity data of the lifted adjustable balance beam to obtain a sequence of the center of gravity data.

[0104] In an example, in the construction system 100 for lifting an adjustable balance beam of an assembled stair, the multi-scale coding module 150 is further configured to: input the sequence vector of the differential center of gravity data into a first convolution layer of the multi-scale neighborhood feature extraction module to obtain a first scale center of gravity change feature vector, wherein the first convolution layer has a first one-dimensional convolution kernel with a first length; input the sequence vector of the differential center of gravity data into a second convolution layer of the multi-scale neighborhood feature extraction module to obtain a second scale center of gravity change feature vector, wherein the second convolution layer has a second one-dimensional convolution kernel with a second length, and the first length is different from the second length; and concatenate the first scale center of gravity change feature vector and the second scale center of gravity change feature vector to obtain the multi-scale center of gravity change feature vector.

[0105] In an example, in the construction system 100 for lifting an adjustable balance beam of an assembled stair, the feature distribution correction module 160 is further configured to: calculate the small-scale local derivative vector between the multi-scale center of gravity change feature vector and the sequence vector of the differential center of gravity data according to the following formula:

[0106]

[0107] wherein v 1i , v 2i and v wi are vector values of respective positions of the multi-scale center of gravity change feature vector, the sequence vector of the differential center of gravity data, and the small-scale local derivative vector respectively, and |v 1i -v 2i | represents an absolute value of a difference between corresponding positions of vectors; and performing point-by-point multiplication on the multi-scale center of gravity change feature vector with the small-scale local derivative vector as a weighted feature vector to obtain the classification feature vector.

[0108] In one example, in the construction system 100 for the assembled stair lifting adjustable balance beam described above, the early warning result generation module 170 is further configured to: perform full connection coding on the classification feature vector using a full connection layer of the classifier to obtain an encoded classification feature vector; and input the encoded classification feature vector into a Softmax classification function of the classifier to obtain the classification result.

[0109] Here, those skilled in the art can understand that the specific functions and operations of each unit and module in the construction system 100 for the assembled stair lifting adjustable balance beam described above have been described in detail above with reference to the construction method for the assembled stair lifting adjustable balance beam Figures 1 to 4 , and therefore the repeated description thereof will be omitted.

[0110] As described above, the construction system 100 for the assembled stair lifting adjustable balance beam according to the embodiments of the present application can be implemented in various terminal devices, such as a server for the construction of the assembled stair lifting adjustable balance beam, etc. In one example, the construction system 100 for the assembled stair lifting adjustable balance beam according to the embodiments of the present application can be integrated into a terminal device as a software module and / or a hardware module. For example, the construction system 100 for the assembled stair lifting adjustable balance beam can be a software module in the operating system of the terminal device, or can be an application program developed for the terminal device; of course, the construction system 100 for the assembled stair lifting adjustable balance beam can also be one of the many hardware modules of the terminal device.

[0111] Alternatively, in another example, the construction system 100 for the assembled stair lifting adjustable balance beam and the terminal device can also be separate devices, and the construction system 100 for the assembled stair lifting adjustable balance beam can be connected to the terminal device through a wired and / or wireless network, and transmit interactive information in accordance with an agreed data format.

[0112] Exemplary electronic device

[0113] Below, the electronic device according to the embodiments of the present application is described with reference to Figure 6 . Figure 6 Fig. 1 illustrates a block diagram of an electronic device according to an embodiment of the present application. As shown in Figure 6 , the electronic device 10 includes one or more processors 11 and a memory 12.

[0114] The processor 11 can be a central processing unit (CPU) or other form of processing unit having data processing and / or instruction execution capabilities, and can control other components in the electronic device 10 to perform desired functions.

[0115] The memory 12 can include one or more computer program products that can include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory, for example, can include random access memory (RAM), cache memory, and / or the like. The non-volatile memory, for example, can include read only memory (ROM), hard disk, flash memory, and / or the like. One or more computer program instructions can be stored on the computer-readable storage media, and the processor 11 can execute the program instructions to implement the functions in the construction method of the fabricated stair hoisting adjustable balance beam of the embodiments of the present application described above and / or other desired functions. Various contents such as hoisting monitoring videos of the hoisted adjustable balance beam can also be stored in the computer-readable storage media.

[0116] In one example, the electronic device 10 can further include an input device 13 and an output device 14, which are interconnected through a bus system and / or other forms of connection mechanism (not shown).

[0117] The input device 13 can include, for example, a keyboard, a mouse, and / or the like.

[0118] The output device 14 can output various information including the decoded value and the like to the outside. The output device 14 can include, for example, a display, a speaker, a printer, a communication network and a remote output device connected thereto, and / or the like.

[0119] Of course, in order to simplify, Figure 6 Only some of the components in the electronic device 10 related to the present application are shown in the block diagram of FIG. 1, and components such as buses, input / output interfaces, and the like are omitted. In addition, the electronic device 10 can include any other appropriate components according to the specific application.

[0120] Exemplary computer program product and computer-readable storage medium

[0121] In addition to the above-described method and device, the embodiments of the present application can be a computer program product including computer program instructions that, when executed by a processor, cause the processor to perform the steps in the functions in the construction method of the fabricated stair hoisting adjustable balance beam according to various embodiments of the present application described in the above "Exemplary Method" section of the specification.

[0122] The computer program product can be written in any combination of one or more programming languages, including an object oriented programming language such as Java, C++ or the like, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computing device, partly on the user's computing device, as a stand-alone software package, partly on the user's computing device and partly on a remote computing device or entirely on the remote computing device or server. The embodiments of methods described herein can be implemented in a computer program product, which can include a computer-readable medium having stored computer program code.

[0123] Furthermore, embodiments of the present application can also be a computer readable storage medium, having stored thereon computer program instructions which, when executed by a processor, cause the processor to carry out the steps described above in the "Exemplary Method" section of the present specification for the method of assembling a hoisting-adjustable balance beam for a prefabricated stairway according to various embodiments of the present application.

[0124] The computer readable storage medium can be a combination of one or more computer readable media. The computer readable media can be a computer readable signal medium or a computer readable storage medium. The computer readable storage medium can include, for example, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium include the following: an electrical connection having one or more wires, a portable disc, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

Claims

1. A construction method of an assembled stair hoisting adjustable balance beam, characterized in that, The method comprises: obtaining hoisting monitoring video of a hoisted adjustable balance beam; extracting a plurality of hoisting monitoring key frames from the hoisting monitoring video; extracting center of gravity data of the hoisted adjustable balance beam in the plurality of hoisting monitoring key frames to obtain a sequence of center of gravity data; calculating the difference between each two adjacent center of gravity data in the sequence of center of gravity data to obtain a sequence of differential center of gravity data, and arranging the sequence of differential center of gravity data into a differential center of gravity data sequence vector; passing the differential center of gravity data sequence vector through a multi-scale neighborhood feature extraction module to obtain a multi-scale center of gravity change feature vector; based on the differential center of gravity data sequence vector, correcting the feature distribution of the multi-scale center of gravity change feature vector to obtain a classification feature vector; and passing the classification feature vector through a classifier to obtain a classification result, wherein the classification result is used to indicate whether to generate a hoisting instability warning prompt of the hoisted adjustable balance beam. The method further comprises: calculating a small-scale local derivative vector between the multi-scale center of gravity change feature vector and the differential center of gravity data sequence vector, wherein the small-scale local derivative vector is related to the absolute value of the positional difference between the multi-scale center of gravity change feature vector and the differential center of gravity data sequence vector; and performing positional point multiplication on the multi-scale center of gravity change feature vector by using the small-scale local derivative vector as a weighted feature vector to obtain the classification feature vector. The method further comprises: calculating the small-scale local derivative vector between the multi-scale center of gravity change feature vector and the differential center of gravity data sequence vector by using the following formula: wherein the formula is: The method further comprises: ； wherein , and are the vector values of the respective positions of the multi-scale barycenter variation feature vector, the difference barycenter data sequence vector and the small-scale local derivative vector respectively, denotes the absolute value of the difference between the corresponding positions of the vectors.

2. The construction method of the assembled stair hoisting adjustable balance beam according to claim 1, characterized in that, extracting the plurality of hoisting monitoring key frames from the hoisting monitoring video at a predetermined sampling frequency.

3. The construction method of the assembled stair hoisting adjustable balance beam according to claim 2, characterized in that, The method further comprises: transforming the plurality of hoisting monitoring key frames from a camera perspective into a global perspective to obtain a plurality of mapped hoisting monitoring key frames; and extracting the center of gravity data of the image frames of the plurality of mapped hoisting monitoring key frames as the center of gravity data of the hoisted adjustable balance beam to obtain the sequence of center of gravity data.

4. The construction method of the assembled stair hoisting adjustable balance beam according to claim 3, characterized in that, The method further comprises: inputting the differential center of gravity data sequence vector into a first convolutional layer of the multi-scale neighborhood feature extraction module to obtain a first-scale center of gravity change feature vector, wherein the first convolutional layer has a first one-dimensional convolution kernel with a first length. The method further comprises: inputting the differential center of gravity data sequence vector into a first convolutional layer of the multi-scale neighborhood feature extraction module to obtain a first-scale center of gravity change feature vector, wherein the first convolutional layer has a first one-dimensional convolution kernel with a first length. inputting the differential barycenter data sequence vector into a second convolutional layer of the multiscale neighborhood feature extraction module to obtain a second scale barycenter change feature vector, wherein the second convolutional layer has a second one-dimensional convolution kernel of a second length, and the first length is different from the second length; and concatenating the first scale barycenter change feature vector and the second scale barycenter change feature vector to obtain the multiscale barycenter change feature vector.

5. The construction method of the assembled stair hoisting adjustable balance beam according to claim 4, characterized in that, The inputting the differential barycenter data sequence vector into the first convolutional layer of the multiscale neighborhood feature extraction module to obtain the first scale barycenter change feature vector comprises: performing one-dimensional convolution coding on the differential barycenter data sequence vector using the first convolutional layer of the multiscale neighborhood feature extraction module according to the following formula to obtain the first scale barycenter change feature vector; wherein the formula is: ； wherein a is a width of the first convolution kernel in the x direction, is a first convolution kernel parameter vector, is a local vector matrix operated with the convolution kernel function, w is a size of the first convolution kernel, and X represents the difference center of gravity data sequence vector.

6. The construction method of the assembled stair hoisting adjustable balance beam according to claim 5, characterized in that, The inputting the differential barycenter data sequence vector into the second convolutional layer of the multiscale neighborhood feature extraction module to obtain the second scale barycenter change feature vector comprises: performing one-dimensional convolution coding on the differential barycenter data sequence vector using the second convolutional layer of the multiscale neighborhood feature extraction module according to the following formula to obtain the second scale barycenter change feature vector; wherein the formula is: ； wherein b is the width of the second convolution kernel in the x direction, is a second convolution kernel parameter vector, is a local vector matrix operated with the convolution kernel function, m is the size of the second convolution kernel, and X represents the difference center of gravity data sequence vector.

7. The construction method of the assembled stair hoisting adjustable balance beam according to claim 6, characterized in that, The inputting the classification feature vector through the classifier to obtain a classification result, wherein the classification result is used to represent whether to generate a hoisting instability early warning prompt of the hoisted adjustable balance beam, comprises: performing full connection coding on the classification feature vector using a full connection layer of the classifier to obtain an encoded classification feature vector; and inputting the encoded classification feature vector into a Softmax classification function of the classifier to obtain the classification result.

8. A construction system for assembling a hoisted adjustable balance beam of a prefabricated staircase, characterized in that, The construction system is used to implement the construction method of the fabricated stair hoisting adjustable balance beam according to any one of claims 1-7, which comprises: a monitoring module configured to acquire a hoisting monitoring video of the hoisted adjustable balance beam; a sampling module configured to extract a plurality of hoisting monitoring key frames from the hoisting monitoring video; a barycenter acquisition module configured to extract barycenter data of the hoisted adjustable balance beam in the plurality of hoisting monitoring key frames to obtain a sequence of barycenter data; a differential and arrangement module configured to calculate a difference between each two adjacent barycenter data in the sequence of barycenter data to obtain a sequence of differential barycenter data, and arrange the sequence of differential barycenter data into a differential barycenter data sequence vector; a multiscale coding module configured to input the differential barycenter data sequence vector through a multiscale neighborhood feature extraction module to obtain a multiscale barycenter change feature vector; a feature distribution correction module configured to correct a feature distribution of the multiscale barycenter change feature vector based on the differential barycenter data sequence vector to obtain a classification feature vector; and an early warning result generation module configured to input the classification feature vector through a classifier to obtain a classification result, wherein the classification result is used to represent whether to generate a hoisting instability early warning prompt of the hoisted adjustable balance beam.

Images