A data interaction and collaborative management method of an intelligent control system of a pipe cutting machine

Abstract

This invention relates to the field of intelligent control technology and discloses a data interaction and collaborative management method for an intelligent control system of a pipe cutting machine. The method includes: locating defects in the inner wall image data stream of the pipe within the target intelligent control system to obtain inner wall quality status data; spatiotemporally aligning the inner wall quality status data with the real-time length measurement data stream of the pipe within the target intelligent control system to construct a continuous pipe status map; locating the pipe segment at the target cutting point in the continuous pipe status map and making collaborative cutting decisions for the pipe segment to obtain cutting control commands; transmitting the cutting control commands to the cutting execution stage of the target intelligent control system and simultaneously collecting actual cutting result data and transmission status data; and storing the obtained data in the central database of the target intelligent control system to achieve subsequent collaborative management. This invention can improve the efficiency of data interaction and collaborative management in the intelligent control system of a pipe cutting machine.

Description

Technical Field

[0001] This invention relates to the field of intelligent control technology, and in particular to a data interaction and collaborative management method for an intelligent control system of a pipe cutting machine. Background Technology

[0002] In the actual operation of the intelligent control system for pipe cutting machines, the data interaction process lacks an efficient multi-source data integration mechanism. Pipe length measurement data and inner wall quality inspection data are often scattered, making it difficult to achieve accurate spatiotemporal correlation. This results in the system being unable to comprehensively and in real-time grasp the overall status of the pipe. Consequently, the cutting decision-making process lacks reliable comprehensive data support, making it difficult to formulate appropriate cutting strategies for the quality differences in different sections of the pipe, thus affecting the accuracy and rationality of the cutting operation.

[0003] Under the existing collaborative management model, the generation and execution feedback of cutting control commands lack a closed-loop connection. Key information such as cutting result data and transmission status data cannot be effectively stored and reused by the system, resulting in a lack of data basis for subsequent equipment debugging and process optimization. At the same time, the accuracy of identifying defects on the inner wall of the pipe during data processing is insufficient, and the judgment of abnormal patterns lacks scientific quantitative standards, which easily leads to missed detections or misjudgments. This further reduces the overall efficiency of the system's collaborative management and fails to meet the requirements of high-precision and high-efficiency pipe cutting operations. Summary of the Invention

[0004] This invention provides a data interaction and collaborative management method for an intelligent control system of a pipe cutting machine to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides a data interaction and collaborative management method for an intelligent control system of a pipe cutting machine, comprising: S1. Acquire real-time length measurement data stream and inner wall image data stream of the pipe during the operation of the target intelligent control system; S2. Defect location is performed on the inner wall image data stream of the pipe to obtain the inner wall quality status data of the inner wall image data stream of the pipe; S3. The inner wall quality status data and the real-time length measurement data stream of the pipe are spatiotemporally correlated and aligned to construct a continuous pipe status map of the target intelligent control system. S4. Locate the pipe section with the target cutting point in the continuous pipe status map, and make a collaborative cutting decision on the pipe section based on the quality label attribute of the pipe section to obtain the cutting control command of the target intelligent control system. S5. Transmit the cutting control command to the cutting execution stage of the target intelligent control system, and simultaneously collect the actual cutting result data and transmission status data fed back by the cutting execution stage; S6. The continuous pipe status map, the cutting control command, the actual cutting result data and the transmission status data are stored in the central database of the target intelligent control system to realize subsequent collaborative management of the target intelligent control system.

[0006] In a preferred embodiment, the real-time length measurement data stream and the inner wall image data stream of the pipe during the operation of the intelligent control system for acquiring the target include: Continuous non-contact measurement of the pipe material during the conveying process of the target intelligent control system is performed to obtain the original length measurement signal of the target intelligent control system; The original length measurement signal is converted from analog to digital to obtain the digital signal sequence of the target intelligent control system; The digital signal sequence is timestamped to obtain the real-time pipe length measurement data stream of the target intelligent control system. Based on the real-time length measurement data stream of the pipe, the inner wall image data stream of the pipe of the target intelligent control system is collected synchronously.

[0007] In a preferred embodiment, the step of locating defects in the pipe inner wall image data stream to obtain inner wall quality status data of the pipe inner wall image data stream includes: Contrast enhancement is performed on the inner wall image data stream of the pipe to obtain a standardized inner wall image sequence of the inner wall image data stream; Extract the texture and edge features of the standardized inner wall image sequence to obtain the inner wall feature descriptor subset of the pipe inner wall image data stream; Anomaly pattern recognition is performed on the inner wall feature description subset to obtain preliminary defect area identifiers for the inner wall image data stream of the pipe. Based on the preliminary defect area identification, the defect classification and determination are performed on the inner wall image data stream of the pipe to obtain the inner wall quality status data of the inner wall image data stream of the pipe.

[0008] In a preferred embodiment, the step of performing abnormal pattern recognition on the subset of inner wall feature descriptions to obtain preliminary defect region identifiers for the inner wall image data stream of the pipe includes: The feature space distribution evolution of the inner wall feature descriptor subset is performed to obtain the feature distribution map of the inner wall feature descriptor subset; Based on the feature distribution map and the preset reference feature vector, the deviation of the inner wall feature descriptor subset is calculated to obtain the anomaly confidence of the feature vectors in the inner wall feature descriptor subset. The formula for calculating the anomaly confidence is as follows: ; In the formula, This indicates the anomaly confidence level. This indicates the number of the reference feature vectors. This represents the feature vector in the subset of inner wall feature descriptions. Indicates the first The mean vector of the reference eigenvectors, Indicates the first The standard deviation of each reference feature vector This represents the preset local similarity adjustment factor. Indicates the first One reference feature vector, The Euclidean norm of a vector. This represents a similarity measurement function between vectors; Adaptive threshold segmentation is performed on the abnormal confidence level to obtain abnormal feature points in the inner wall image data stream of the pipe. Spatial clustering of the abnormal feature points yields preliminary defect region identifiers for the inner wall image data stream of the pipe.

[0009] In a preferred embodiment, the step of spatiotemporally correlating and aligning the inner wall quality status data with the real-time length measurement data stream of the pipe to construct a continuous pipe status map of the target intelligent control system includes: Based on the timestamp information of the target intelligent control system, the inner wall quality status data and the real-time length measurement data stream of the pipe are time-series synchronized and aligned to obtain the time-series associated data stream of the target intelligent control system. Based on the time-series associated data stream and the pipe delivery mapping relationship of the target intelligent control system, the real-time length measurement data stream of the pipe is mapped to the corresponding position in the inner wall quality status data; Based on the mapping results, a location index mapping table for the target intelligent control system is established; Based on the location index mapping table, the defect location results in the inner wall quality status data are associated with the absolute length position of the pipe in the target intelligent control system to obtain the quality label sequence of the target intelligent control system. Multi-source fusion analysis is performed on the quality label sequence and the real-time length measurement data stream of the pipe to obtain the continuous state entries of the target intelligent control system; According to the preset data structure, the continuous state entries are topologically constructed to obtain the continuous pipe state map of the target intelligent control system.

[0010] In a preferred embodiment, the step of performing multi-source fusion analysis on the quality tag sequence and the real-time length measurement data stream of the pipe to obtain continuous state entries of the target intelligent control system includes: The pipe length ranges marked as defects in the quality label sequence are integrated into the defect length range set of the target intelligent control system; The real-time length measurement data stream of the pipe is discretized and segmented at preset length intervals to obtain the basic length interval sequence of the target intelligent control system; The defect length interval set is associated and matched with the basic length interval sequence, and based on the matching result, the basic length interval is labeled with a quality status label to obtain the length interval segment of the target intelligent control system. Logically link the length interval segments according to the increasing order of pipe length to obtain the continuous state entries of the target intelligent control system.

[0011] In a preferred embodiment, the step of performing topology construction on the continuous state entries according to a preset data structure to obtain the continuous pipe state map of the target intelligent control system includes: The continuous state entries are instantiated into nodes to obtain the graph nodes of the target intelligent control system; Based on the length attribute of the graph nodes, directed edges are linked to adjacent graph nodes to obtain the edge relationships of the target intelligent control system. Based on the quality state labels in the continuous state entries, attribute values ​​are assigned to the graph nodes to obtain the optimized graph nodes of the target intelligent control system. Based on the optimized graph nodes and the edge relationships, a continuous pipe state graph of the target intelligent control system is constructed.

[0012] In a preferred embodiment, the step of locating the pipe segment with the target cutting point in the continuous pipe state map, and making a collaborative cutting decision on the pipe segment based on the quality label attributes of the pipe segment to obtain the cutting control command of the target intelligent control system includes: Based on preset cutting task parameters, the nodes in the continuous pipe state map are traversed to filter out candidate pipe sections in the continuous pipe state map that meet the cutting task parameters. Based on the candidate pipe sections, the quality label attributes of the candidate pipe sections are parsed to obtain the candidate section quality attribute set of the candidate pipe sections; Based on preset cutting quality standards and target cutting points, the set of quality attributes of the candidate sections is matched for cutting priority to obtain the selected pipe section of the target cutting point and the corresponding cutting decision parameters. The selected pipe section and its corresponding cutting decision parameters are structured and encoded to obtain the cutting control command of the target intelligent control system.

[0013] In a preferred embodiment, the step of matching the cutting priority of the candidate segment quality attribute set based on preset cutting quality standards and target cutting points to obtain the selected pipe segment of the target cutting point and the corresponding cutting decision parameters includes: Based on the cutting quality standard and the set of candidate segment quality attributes, the candidate pipe segments are screened for quality level to obtain a priority queue of candidate pipe segments. Based on the defect severity attribute of the segments in the candidate segment priority queue and the position offset of adjacent segments, the candidate segment priority queue is dynamically prioritized and rearranged. Based on the location attributes of the target cutting point, the uniqueness of the rearranged candidate segment priority queue is determined to obtain the selected pipe segment of the target cutting point. The quality label attributes and geometric parameters of the selected pipe section are encapsulated into cutting decision parameters to obtain the cutting decision parameters of the target cutting point.

[0014] In a preferred embodiment, the step of transmitting the cutting control command to the cutting execution stage of the target intelligent control system, and simultaneously collecting the actual cutting result data and transmission status data fed back by the cutting execution stage, includes: The cutting control commands are subjected to protocol adaptation and conversion to obtain the command adaptation data of the target intelligent control system; The instruction adaptation data is integrated and encapsulated to obtain the instruction data packet of the target intelligent control system; The instruction data packet is transmitted to the cutting execution stage of the target intelligent control system to trigger the cutting execution stage to complete the cutting operation; The system synchronously receives the feedback data stream reported in real time by the cutting execution stage; The feedback data stream is subjected to multi-channel data separation to obtain the actual cutting result data of the target intelligent control system and the transmission status data.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention acquires real-time pipe length data through continuous non-contact measurement, and achieves precise location of inner wall defects by combining image enhancement, feature extraction, and abnormal confidence quantification. Then, a continuous pipe status map is constructed through spatiotemporal correlation alignment, enabling efficient fusion of multi-source data and providing comprehensive and accurate pipe status support for cutting decisions. Based on the priority matching mechanism of quality label attributes, cutting control instructions adapted to different pipe sections can be generated, greatly improving the accuracy and rationality of cutting operations.

[0016] 2. This invention establishes a closed-loop mechanism for cutting control, execution feedback, and data storage, synchronously collecting and storing key data such as cutting results and transmission status. This provides reliable data for subsequent equipment debugging and process optimization, achieving collaborative management throughout the entire process. Simultaneously, the standardized data processing flow and structured storage method improve data reuse rate and system interaction efficiency, enhancing the scalability and ease of operation and maintenance of the intelligent control system. Attached Figure Description

[0017] Figure 1 A flowchart illustrating a data interaction and collaborative management method for an intelligent control system of a pipe cutting machine according to an embodiment of the present invention; The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0018] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0019] This application provides a data interaction and collaborative management method for an intelligent control system of a pipe cutting machine. The executing entity of this data interaction and collaborative management method for an intelligent control system of a pipe cutting machine includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the data interaction and collaborative management method for an intelligent control system of a pipe cutting machine can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0020] Reference Figure 1The diagram shown is a flowchart illustrating a data interaction and collaborative management method for an intelligent control system of a pipe cutting machine according to an embodiment of the present invention. In this embodiment, the data interaction and collaborative management method for the intelligent control system of a pipe cutting machine includes: S1. Acquire real-time length measurement data stream and inner wall image data stream of the pipe during the operation of the target intelligent control system; In this embodiment of the invention, the real-time length measurement data stream and the inner wall image data stream of the pipe during the operation of the intelligent control system for acquiring the target include: Continuous non-contact measurement of the pipe material during the conveying process of the target intelligent control system is performed to obtain the original length measurement signal of the target intelligent control system; The original length measurement signal is converted from analog to digital to obtain the digital signal sequence of the target intelligent control system; The digital signal sequence is timestamped to obtain the real-time pipe length measurement data stream of the target intelligent control system. Based on the real-time length measurement data stream of the pipe, the inner wall image data stream of the pipe of the target intelligent control system is collected synchronously.

[0021] The specific implementation process for continuously non-contactly measuring the pipe material during the conveying process of the target intelligent control system to obtain the original length measurement signal of the target intelligent control system is as follows: A laser ranging device is used to perform continuous non-contact measurement operations. The transmitting end and receiving end of the laser ranging device are fixed on both sides of the pipe material conveying path of the target intelligent control system, and the line connecting the transmitting end and the receiving end is perpendicular to the pipe material conveying direction. As the pipe material moves with the conveying mechanism, the transmitting end of the laser ranging device continuously emits a laser beam towards the surface of the pipe material. After the laser beam contacts the surface of the pipe material, it is reflected. The reflected laser beam is captured by the receiving end in real time. The receiving end converts the laser beam into a corresponding electrical signal according to the time difference between the emission and reception of the laser beam. This electrical signal is continuously generated as the position of the pipe material changes during the conveying process, forming a continuous original length measurement signal.

[0022] The specific implementation process of performing analog-to-digital conversion on the original length measurement signal to obtain the digital signal sequence of the target intelligent control system is as follows: the original length measurement signal is input into the analog-to-digital conversion module. The analog-to-digital conversion module has a built-in sample-and-hold circuit and a quantization encoding circuit. The sample-and-hold circuit discretizes the continuously changing original length measurement signal according to a preset timing sequence, keeping the signal amplitude of each interception node stable during the interception process. The quantization encoding circuit compares the amplitude of the discrete analog signal output by the sample-and-hold circuit with the preset standard voltage amplitude range point by point, mapping the amplitude of each discrete analog signal to the corresponding binary code. All binary codes are arranged sequentially according to the sampling timing sequence to form an ordered digital signal sequence.

[0023] The specific implementation process of timestamping the digital signal sequence to obtain the real-time pipe length measurement data stream of the target intelligent control system is as follows: the digital signal sequence is transmitted to the timestamp module. The timestamp module has a built-in real-time clock unit, which continuously outputs standard time information. The timestamp module matches the corresponding time information output by the real-time clock unit to each binary coded data point of the digital signal sequence one by one according to the generation order of the digital signal sequence. Each data point carries a unique corresponding time information. The digital signal sequence carrying time information is arranged and integrated according to the generation time sequence to form the real-time pipe length measurement data stream.

[0024] The specific implementation process of synchronously acquiring the pipe inner wall image data stream of the target intelligent control system based on the real-time pipe length measurement data stream is as follows: The real-time pipe length measurement data stream is transmitted to the synchronization control module. The synchronization control module extracts the time information in the real-time pipe length measurement data stream as a trigger reference. The synchronization control module establishes a signal connection with the high-speed industrial endoscope image acquisition device. When the synchronization control module detects the time information of a certain data point in the real-time pipe length measurement data stream, it immediately sends an acquisition trigger signal to the high-speed industrial endoscope image acquisition device. The probe of the high-speed industrial endoscope image acquisition device is pre-extended into the inner wall of the pipe of the target intelligent control system. After receiving the trigger signal, it immediately captures the inner wall image of the pipe at the corresponding time. Each trigger signal corresponds to one frame of inner wall image. All inner wall images are arranged in the time order of the trigger signals, and each frame of image carries the same time information as the real-time pipe length measurement data stream, forming a pipe inner wall image data stream synchronized with the real-time pipe length measurement data stream.

[0025] The beneficial effects are: non-contact laser measurement is used to obtain continuous and stable original length measurement signals, avoiding the damage and errors of contact measurement; analog-to-digital conversion is used to achieve accurate conversion of analog signals into digital signal sequences, solving the problem of analog signals being susceptible to interference; a traceable real-time pipe length measurement data stream is formed with the help of timestamp marking; and inner wall images are synchronously acquired based on the time reference of this data stream to realize the linkage between length data and inner wall status data, thereby improving the efficiency of intelligent control system in managing and detecting pipes.

[0026] S2. Defect location is performed on the inner wall image data stream of the pipe to obtain the inner wall quality status data of the inner wall image data stream of the pipe; In this embodiment of the invention, the step of locating defects in the pipe inner wall image data stream to obtain the inner wall quality status data of the pipe inner wall image data stream includes: Contrast enhancement is performed on the inner wall image data stream of the pipe to obtain a standardized inner wall image sequence of the inner wall image data stream; Extract the texture and edge features of the standardized inner wall image sequence to obtain the inner wall feature descriptor subset of the pipe inner wall image data stream; Anomaly pattern recognition is performed on the inner wall feature description subset to obtain preliminary defect area identifiers for the inner wall image data stream of the pipe. Based on the preliminary defect area identification, the defect classification and determination are performed on the inner wall image data stream of the pipe to obtain the inner wall quality status data of the inner wall image data stream of the pipe.

[0027] The step of performing abnormal pattern recognition on the subset of inner wall feature descriptions to obtain preliminary defect region identifiers for the inner wall image data stream of the pipe includes: The feature space distribution evolution of the inner wall feature descriptor subset is performed to obtain the feature distribution map of the inner wall feature descriptor subset; Based on the feature distribution map and the preset reference feature vector, the deviation of the inner wall feature descriptor subset is calculated to obtain the anomaly confidence of the feature vectors in the inner wall feature descriptor subset. The formula for calculating the anomaly confidence is as follows: ; In the formula, This indicates the anomaly confidence level. This indicates the number of the reference feature vectors. This represents the feature vector in the subset of inner wall feature descriptions. Indicates the first The mean vector of the reference eigenvectors, Indicates the first The standard deviation of each reference feature vector This represents the preset local similarity adjustment factor. Indicates the first One reference feature vector, The Euclidean norm of a vector. This represents a similarity measurement function between vectors; Adaptive threshold segmentation is performed on the abnormal confidence level to obtain abnormal feature points in the inner wall image data stream of the pipe. Spatial clustering of the abnormal feature points yields preliminary defect region identifiers for the inner wall image data stream of the pipe.

[0028] The pipe inner wall image data stream is transmitted to the image contrast enhancement module. This module has a built-in pixel grayscale adjustment unit. The grayscale adjustment unit reads all pixel grayscale information of each frame of the inner wall image in the data stream, identifies the maximum and minimum grayscale values ​​of each frame, and stretches and adjusts all pixel grayscale values ​​between the maximum and minimum grayscale values ​​according to the grayscale distribution range, so that the grayscale difference between the bright and dark areas in the image is significantly increased. After the grayscale stretching adjustment, all the inner wall images are arranged and integrated in their original time order to form a standardized inner wall image sequence of the pipe inner wall image data stream.

[0029] The standardized inner wall image sequence is input into the image feature extraction module, which has a built-in texture feature recognition unit and an edge feature recognition unit. The texture feature recognition unit scans the images in the standardized inner wall image sequence frame by frame, captures the arrangement and distribution of pixels in the images, and extracts feature information reflecting the roughness and texture direction of the inner wall surface. The edge feature recognition unit detects the images in the standardized inner wall image sequence frame by frame, captures the pixel positions where the gray value changes abruptly in the images, and extracts feature information reflecting the inner wall contour and possible defect edges. The extracted texture feature information and edge feature information are integrated according to the time sequence of the corresponding images to form the inner wall feature descriptor subset of the pipe inner wall image data stream.

[0030] The inner wall feature descriptor subset is transmitted to the feature space distribution evolution module. This module has a built-in feature space mapping unit. The feature space mapping unit assigns corresponding spatial dimension coordinates to each feature information in the inner wall feature descriptor subset, maps the texture feature information and edge feature information to the preset feature space respectively, marks the corresponding position points in the feature space according to the attribute association relationship of each feature information, and connects and distributes all position points according to feature attributes to form the feature distribution map of the inner wall feature descriptor subset.

[0031] The feature distribution map and the preset reference feature vector are input into the deviation calculation module. This module has a built-in feature comparison unit. The feature comparison unit extracts the feature attribute information corresponding to each position point in the feature distribution map, and compares the feature attribute information with the standard feature attribute information corresponding to the preset reference feature vector point by point. It determines the offset of the feature attribute information of each position point relative to the reference feature vector, and determines the deviation degree quantification result corresponding to each feature attribute information based on the offset. The quantification result is the abnormal confidence of the feature vector in the inner wall feature descriptor subset.

[0032] In the formula for calculating the anomaly confidence level, the number of reference feature vectors comes from a pre-defined set of reference feature vectors. This set is pre-constructed based on the feature data of the inner wall of qualified pipes. During construction, multiple defect-free inner wall samples of pipes are selected, their feature vectors are extracted, and organized into a set. The feature vectors within the set must cover the typical feature states of the inner wall of qualified pipes. The feature vectors in the inner wall feature description subset come from the inner wall feature description subset, which is obtained by extracting texture and edge features from a standardized inner wall image sequence. The extraction process must ensure coverage of all effective areas of the image, and that the feature information corresponds consistently with the temporal sequence of the image. The mean vector of the nth reference eigenvector is the... The value is obtained by averaging the feature vectors of all samples corresponding to the reference feature vector. The calculation must use all sample data under the same type of reference feature vector; no samples should be omitted or added. The standard deviation of the nth reference eigenvector is based on the nth reference eigenvector. The value of each reference feature vector is calculated by examining the numerical dispersion of all sample feature vectors corresponding to that vector. This calculation must be based on the same sample set as the mean vector. The preset local similarity adjustment factor is pre-set based on the degree of difference in common characteristics of pipe inner wall defects, and its setting must match the quality control standards for pipe inner walls in actual production. The first reference feature vector comes from a pre-defined set of reference feature vectors, the construction of which must conform to the feature specifications of the inner wall of qualified pipe materials. The Euclidean norm of the vector is obtained by calculating the spatial distance between the feature vector and the mean vector, and the calculation must be based on the numerical values ​​of the feature vector and the corresponding mean vector. The similarity measure function between vectors is obtained by comparing the feature vector with the first reference feature vector. The degree of attribute matching of a reference feature vector is obtained, and the comparison needs to cover all attribute dimensions of the feature vector.

[0033] When the feature vector in the inner wall feature descriptor subset is related to the first The greater the spatial distance between the mean vectors of the nth reference feature vectors, the greater the anomaly confidence. The larger the standard deviation of the reference feature vectors, the smaller the variation in anomaly confidence. Conversely, the lower the matching degree corresponding to the similarity metric function between vectors, the larger the anomaly confidence. Furthermore, the larger the preset local similarity adjustment factor, the greater the influence of the vector similarity metric function on anomaly confidence.

[0034] The abnormal confidence scores are transmitted to the adaptive threshold segmentation module, which has a built-in threshold generation unit and feature filtering unit. The threshold generation unit generates an appropriate segmentation threshold based on the overall distribution of abnormal confidence scores. The feature filtering unit compares the abnormal confidence scores of each feature vector with the generated segmentation threshold and filters out the location points corresponding to feature vectors whose abnormal confidence scores exceed the segmentation threshold. These location points are the abnormal feature points of the inner wall image data stream of the pipe.

[0035] The abnormal feature points are transmitted to the spatial clustering module, which has a built-in feature point aggregation unit. The feature point aggregation unit identifies the location distribution of abnormal feature points in the feature space, aggregates and classifies abnormal feature points that are adjacent in location and have similar feature attributes, and assigns a corresponding region identifier to each type of aggregated abnormal feature points. These region identifiers are the preliminary defect region identifiers of the inner wall image data stream of the pipe.

[0036] The preliminary defect area identifier is input into the defect classification and determination module. This module has a built-in defect type matching unit. The defect type matching unit extracts the feature attribute information corresponding to the preliminary defect area identifier, and matches the feature attribute information with the preset defect standard feature information one by one. Based on the matching result, the defect type corresponding to each preliminary defect area identifier is determined. All defect type information and corresponding area location information are integrated to form the inner wall quality status data of the pipe inner wall image data stream.

[0037] The beneficial effects include enhancing image contrast to highlight inner wall details, extracting texture and edge features to form feature support, intuitively presenting feature distribution through feature space evolution, quantifying the degree of anomalies by combining reference vectors, locating anomaly points through threshold segmentation, obtaining defect area identifiers through clustering, and finally classifying and determining to generate inner wall quality status data, thereby achieving accurate detection of pipe inner walls.

[0038] S3. The inner wall quality status data and the real-time length measurement data stream of the pipe are spatiotemporally correlated and aligned to construct a continuous pipe status map of the target intelligent control system. In this embodiment of the invention, the step of spatiotemporally correlating and aligning the inner wall quality status data with the real-time length measurement data stream of the pipe to construct a continuous pipe status map of the target intelligent control system includes: Based on the timestamp information of the target intelligent control system, the inner wall quality status data and the real-time length measurement data stream of the pipe are time-series synchronized and aligned to obtain the time-series associated data stream of the target intelligent control system. Based on the time-series associated data stream and the pipe delivery mapping relationship of the target intelligent control system, the real-time length measurement data stream of the pipe is mapped to the corresponding position in the inner wall quality status data; Based on the mapping results, a location index mapping table for the target intelligent control system is established; Based on the location index mapping table, the defect location results in the inner wall quality status data are associated with the absolute length position of the pipe in the target intelligent control system to obtain the quality label sequence of the target intelligent control system. Multi-source fusion analysis is performed on the quality label sequence and the real-time length measurement data stream of the pipe to obtain the continuous state entries of the target intelligent control system; According to the preset data structure, the continuous state entries are topologically constructed to obtain the continuous pipe state map of the target intelligent control system.

[0039] The multi-source fusion analysis of the quality label sequence and the real-time length measurement data stream of the pipe to obtain the continuous state entries of the target intelligent control system includes: The pipe length ranges marked as defects in the quality label sequence are integrated into the defect length range set of the target intelligent control system; The real-time length measurement data stream of the pipe is discretized and segmented at preset length intervals to obtain the basic length interval sequence of the target intelligent control system; The defect length interval set is associated and matched with the basic length interval sequence, and based on the matching result, the basic length interval is labeled with a quality status label to obtain the length interval segment of the target intelligent control system. Logically link the length interval segments according to the increasing order of pipe length to obtain the continuous state entries of the target intelligent control system.

[0040] The step of constructing a topology for the continuous state entries according to a preset data structure to obtain a continuous pipe state map of the target intelligent control system includes: The continuous state entries are instantiated into nodes to obtain the graph nodes of the target intelligent control system; Based on the length attribute of the graph nodes, directed edges are linked to adjacent graph nodes to obtain the edge relationships of the target intelligent control system. Based on the quality state labels in the continuous state entries, attribute values ​​are assigned to the graph nodes to obtain the optimized graph nodes of the target intelligent control system. Based on the optimized graph nodes and the edge relationships, a continuous pipe state graph of the target intelligent control system is constructed.

[0041] Based on the timestamp information of the target intelligent control system, the specific implementation process of synchronizing and aligning the inner wall quality status data with the real-time length measurement data stream of the pipe to obtain the time-series associated data stream of the target intelligent control system is as follows: extract the timestamp information carried by the inner wall quality status data and the real-time length measurement data stream of the pipe, input these timestamp information into the time alignment module, the module reads the timestamp content corresponding to each data, and pairs and integrates the entries whose timestamps in the inner wall quality status data are completely consistent with the timestamps in the real-time length measurement data stream of the pipe to form a time-series associated data stream containing inner wall quality information and length measurement information under the same timestamp.

[0042] The specific implementation process of mapping the real-time length measurement data stream of the pipe to the corresponding position in the inner wall quality status data according to the pipe transportation mapping relationship of the time-series associated data stream and the target intelligent control system is as follows: obtain the pipe transportation mapping relationship preset by the target intelligent control system. This relationship clarifies the correspondence rules between length change and time during pipe transportation. The real-time length measurement data of the pipe in the time-series associated data stream is mapped one by one to the entry position corresponding to the same timestamp in the inner wall quality status data according to the rules corresponding to the mapping relationship, so that each inner wall quality status data entry is associated with the corresponding pipe length measurement information.

[0043] Based on the mapping results, the specific implementation process for establishing the location index mapping table of the target intelligent control system is as follows: organize the correspondence between the mapped inner wall quality status data and the real-time pipe length measurement data, take the pipe length information corresponding to each inner wall quality status data as the index item, and take the entry identifier of the inner wall quality status data as the corresponding item. Arrange and record the index items and corresponding items in the order of the pipe length information to form a location index mapping table containing the correspondence between length information and inner wall quality data entry identifier.

[0044] Based on the location index mapping table, the specific implementation process of associating the defect location results in the inner wall quality status data with the absolute length position of the pipe in the target intelligent control system to obtain the quality label sequence of the target intelligent control system is as follows: extract the defect location results from the inner wall quality status data, find the absolute length position of the pipe corresponding to the defect location results through the location index mapping table, bind the defect location information with the corresponding absolute length position, and arrange the defect and qualified status information of all bound length positions in ascending order of pipe length to form a quality label sequence.

[0045] The specific implementation process of integrating the pipe length intervals marked as defects in the quality label sequence into the defect length interval set of the target intelligent control system is as follows: identify the pipe length position corresponding to all entries marked as defects in the quality label sequence, merge adjacent and continuous defect length positions into one length interval, and organize and record all independent defect length intervals in sequence to form a defect length interval set.

[0046] The specific implementation process of discretizing and segmenting the real-time length measurement data stream of the pipe with a preset length interval to obtain the basic length interval sequence of the target intelligent control system is as follows: determine the preset fixed length interval of the target intelligent control system, start from the initial length position of the real-time length measurement data stream of the pipe, divide the entire pipe length data according to the fixed length interval, each divided length range is a basic length interval, and arrange all basic length intervals in length order to form a basic length interval sequence.

[0047] The specific implementation process of associating and matching the set of defect length intervals with the sequence of basic length intervals, and labeling the basic length intervals with quality status based on the matching results to obtain the length interval segments of the target intelligent control system is as follows: Each interval in the set of defect length intervals is compared with the intervals in the sequence of basic length intervals. If the basic length interval is completely or partially contained within the defect length interval, a defect status label is marked for the basic length interval. If the basic length interval does not overlap with any defect length interval, a qualified status label is marked. Each basic length interval after marking is a length interval segment.

[0048] The specific implementation process of logically linking the length interval segments according to the increasing order of pipe length to obtain the continuous state entries of the target intelligent control system is as follows: arrange all length interval segments in order from the beginning to the end according to the corresponding pipe length, connect adjacent length interval segments through the connection relationship of length position, so that the end length of each segment corresponds to the initial length of the next segment, forming a continuous state information sequence covering the entire pipe length, which is the continuous state entry.

[0049] The specific implementation process of instantiating the continuous state entries into nodes to obtain the graph nodes of the target intelligent control system is as follows: each length interval segment in the continuous state entry is treated as an independent instance object, a unique node identifier is assigned to each instance object, and the length range and quality status label information corresponding to the segment are recorded. These instance objects containing identifiers, length ranges and quality status are the graph nodes.

[0050] The specific implementation process of linking adjacent graph nodes with directed edges based on the length attribute of the graph nodes to obtain the edge relationship of the target intelligent control system is as follows: extract the length range information corresponding to each graph node, identify the connection relationship between the end length of the previous node and the initial length of the next node in adjacent graph nodes, establish a directed link edge between these two adjacent nodes, with the direction of the edge pointing from the node with smaller length to the node with larger length, and the link edges between all adjacent nodes together constitute the edge relationship.

[0051] The specific implementation process of assigning attribute values ​​to the graph nodes based on the quality status labels in the continuous state entries to obtain the optimized graph nodes of the target intelligent control system is as follows: extract the quality status labels corresponding to each length interval segment from the continuous state entries, assign the label information to the corresponding graph nodes, so that each graph node carries the corresponding quality status attributes in addition to the original identifier and length range information. The graph nodes after the assignment is completed are the optimized graph nodes.

[0052] The specific implementation process for constructing the continuous pipe state map of the target intelligent control system based on the optimized map nodes and the edge relationships is as follows: arrange all optimized map nodes in the order of their corresponding lengths, and connect adjacent optimized map nodes with directed edges according to the edge relationships, so that all nodes and edges together form an overall structure that includes pipe length distribution, quality status information and connection relationships between nodes. This structure is the continuous pipe state map.

[0053] The beneficial effects include achieving precise temporal correlation between inner wall quality and length measurement data, establishing a location index and binding defects with the absolute length of the pipe, generating quality labels and defect intervals, matching the basic length interval to mark the quality status, and finally constructing a continuous pipe status map to intuitively present the correspondence between pipe length and quality.

[0054] S4. Locate the pipe section with the target cutting point in the continuous pipe status map, and make a collaborative cutting decision on the pipe section based on the quality label attribute of the pipe section to obtain the cutting control command of the target intelligent control system. In this embodiment of the invention, the step of locating the pipe segment with the target cutting point in the continuous pipe state map, and making a collaborative cutting decision on the pipe segment based on the quality label attribute of the pipe segment to obtain the cutting control command of the target intelligent control system includes: Based on preset cutting task parameters, the nodes in the continuous pipe state map are traversed to filter out candidate pipe sections in the continuous pipe state map that meet the cutting task parameters. Based on the candidate pipe sections, the quality label attributes of the candidate pipe sections are parsed to obtain the candidate section quality attribute set of the candidate pipe sections; Based on preset cutting quality standards and target cutting points, the set of quality attributes of the candidate sections is matched for cutting priority to obtain the selected pipe section of the target cutting point and the corresponding cutting decision parameters. The selected pipe section and its corresponding cutting decision parameters are structured and encoded to obtain the cutting control command of the target intelligent control system.

[0055] The process involves matching the quality attributes of the candidate sections based on preset cutting quality standards and target cutting points to determine cutting priority, thereby obtaining the selected pipe section at the target cutting point and its corresponding cutting decision parameters. Based on the cutting quality standard and the set of candidate segment quality attributes, the candidate pipe segments are screened for quality level to obtain a priority queue of candidate pipe segments. Based on the defect severity attribute of the segments in the candidate segment priority queue and the position offset of adjacent segments, the candidate segment priority queue is dynamically prioritized and rearranged. Based on the location attributes of the target cutting point, the uniqueness of the rearranged candidate segment priority queue is determined to obtain the selected pipe segment of the target cutting point. The quality label attributes and geometric parameters of the selected pipe section are encapsulated into cutting decision parameters to obtain the cutting decision parameters of the target cutting point.

[0056] The preset cutting task parameters are input into the map node traversal module. This module reads the length and quality status attributes of all optimized map nodes in the continuous pipe status map, and compares the pipe segment information corresponding to each map node one by one according to the length and quality requirements specified in the cutting task parameters. The module then selects the pipe segments corresponding to the map nodes that fully meet all the requirements of the cutting task parameters. These pipe segments are the candidate pipe segments in the continuous pipe status map that meet the cutting task parameters.

[0057] The candidate pipe sections are transmitted to the quality attribute parsing module. This module retrieves the optimized map nodes corresponding to the candidate pipe sections in the continuous pipe status map, extracts all quality-related attribute information such as the quality status label, defect type, and defect distribution range bound to each candidate pipe section, and organizes and collects all the extracted quality attribute information according to the order of the candidate pipe sections to form the candidate section quality attribute set of the candidate pipe sections.

[0058] The cutting quality standard and the set of candidate segment quality attributes are input into the quality level filtering module. This module compares the quality attribute information of each candidate pipe segment in the set of candidate segment quality attributes with the cutting quality standard item by item, determines whether the quality level of each candidate pipe segment meets the requirements of the cutting quality standard, filters out candidate pipe segments whose quality level meets the cutting quality standard, and arranges these candidate pipe segments in descending order of quality level to form the candidate segment priority queue.

[0059] The candidate segment priority queue is input into the dynamic priority reordering module. This module extracts the defect severity attribute and the position offset of adjacent segments for each candidate pipe segment in the candidate segment priority queue. The defect severity attribute determines the basic priority weight of the candidate pipe segment, and the position offset of adjacent segments affects the adjusted priority weight of the candidate pipe segment. The basic priority weight and the adjusted priority weight work together to readjust the original arrangement order in the candidate segment priority queue, thus completing the dynamic priority reordering of the candidate segment priority queue.

[0060] The rearranged priority queue of candidate segments and the position attribute of the target cutting point are input into the uniqueness determination module. This module extracts the absolute length position of the pipe corresponding to the target cutting point, compares this position with the length range of each candidate pipe segment in the rearranged priority queue of candidate segments, and determines whether the position attribute of the target cutting point falls completely within the length range of a certain candidate pipe segment. The module then selects the unique candidate pipe segment that contains the position attribute of the target cutting point, and this candidate pipe segment is the selected pipe segment of the target cutting point.

[0061] The quality label attributes and geometric parameters of the selected pipe section are input into the cutting decision parameterization encapsulation module. This module extracts the defect status and defect level information corresponding to the quality label attributes of the selected pipe section, as well as the length range and section width information corresponding to the geometric parameters. This information is integrated according to the preset parameter encapsulation format and transformed into a standardized parameter form that can be recognized by the target intelligent control system, forming the cutting decision parameters of the target cutting point.

[0062] The selected pipe section and its corresponding cutting decision parameters are input into the structured coding module. This module converts and encodes the length range information of the selected pipe section and the standardized parameter information of the cutting decision parameters according to the instruction coding rules of the target intelligent control system. The encoded information includes all control elements such as the position instruction for cutting execution, the quality judgment instruction, and the parameter execution instruction, forming the cutting control instruction of the target intelligent control system.

[0063] The beneficial effects are: screening candidate pipe sections based on cutting task parameters, analyzing quality attributes and forming a priority queue in combination with cutting quality standards, dynamically rearranging the queue based on defect severity and position offset, determining a unique selected section based on the target cutting point position, and generating cutting control instructions through parameterized encapsulation and structured coding, thereby achieving intelligent decision-making and control of pipe cutting efficiently and accurately.

[0064] S5. Transmit the cutting control command to the cutting execution stage of the target intelligent control system, and simultaneously collect the actual cutting result data and transmission status data fed back by the cutting execution stage; In this embodiment of the invention, the step of transmitting the cutting control command to the cutting execution stage of the target intelligent control system, and simultaneously collecting the actual cutting result data and transmission status data fed back by the cutting execution stage, includes: The cutting control commands are subjected to protocol adaptation and conversion to obtain the command adaptation data of the target intelligent control system; The instruction adaptation data is integrated and encapsulated to obtain the instruction data packet of the target intelligent control system; The instruction data packet is transmitted to the cutting execution stage of the target intelligent control system to trigger the cutting execution stage to complete the cutting operation; The system synchronously receives the feedback data stream reported in real time by the cutting execution stage; The feedback data stream is subjected to multi-channel data separation to obtain the actual cutting result data of the target intelligent control system and the transmission status data.

[0065] The cutting control command is input to the protocol adaptation and conversion module. This module has a built-in communication protocol parsing unit for the cutting execution stage of the target intelligent control system. The parsing unit reads the original data format of the cutting control command and adjusts the field arrangement order and data representation format of the cutting control command according to the communication protocol data format requirements supported by the cutting execution stage. This allows the adjusted command data to be directly recognized and received by the cutting execution stage. The adjusted command data is the command adaptation data of the target intelligent control system.

[0066] The instruction adaptation data is transmitted to the integration and encapsulation module. This module has a built-in data verification unit and a data encapsulation unit. The data verification unit verifies the completeness and validity of the instruction adaptation data. After confirming that the data is complete and error-free, the data encapsulation unit integrates the instruction adaptation data and data verification information according to the data packet structure preset by the target intelligent control system. It adds a data header identifier and a data tail identifier to the integrated data to form the instruction data packet of the target intelligent control system.

[0067] The instruction data packet is sent to the cutting execution stage through a dedicated communication link of the target intelligent control system. The cutting execution stage has a built-in instruction receiving unit and instruction execution unit. The instruction receiving unit monitors the data transmission status in the communication link in real time, captures and reads the instruction data packet, and transmits the instruction adaptation data in the data packet to the instruction execution unit. The instruction execution unit drives the cutting execution mechanism to operate according to the requirements of the instruction adaptation data, so as to trigger the cutting execution stage to complete the cutting operation.

[0068] During the cutting execution process, the feedback data receiving module of the target intelligent control system monitors the status output port of the cutting execution process in real time. This port continuously outputs various status and result information during the cutting process. The feedback data receiving module captures this information according to a preset receiving frequency and integrates it into a continuous data sequence, forming the feedback data stream reported in real time by the cutting execution process.

[0069] The feedback data stream is input to the multi-channel data separation module, which has a built-in data feature recognition unit. The data feature recognition unit extracts feature identifiers of different types of data in the feedback data stream. Based on the feature identifiers corresponding to the actual cutting result data, it filters out relevant data such as cutting position and cutting accuracy. Based on the feature identifiers corresponding to the transmission status data, it filters out relevant data such as pipe conveying speed and conveying position. After the two types of data are sorted independently, they form the actual cutting result data and the transmission status data of the target intelligent control system.

[0070] The beneficial effects are that the communication compatibility between cutting control commands and execution links is achieved through protocol adaptation and conversion, the integrity and effectiveness of command data packets are ensured through integration and encapsulation, the transmission of command data packets to the execution link can accurately trigger cutting operations, the synchronous reception of feedback data streams can monitor the operation status in real time, and the multi-path separation of feedback data streams can clearly obtain two types of data: actual cutting results and transmission status, providing reliable support for the closed-loop management of cutting operations in intelligent control systems.

[0071] In this embodiment of the invention, S6, the continuous pipe status map, the cutting control command, the actual cutting result data and the transmission status data are stored in the central database of the target intelligent control system to realize subsequent collaborative management of the target intelligent control system.

[0072] The continuous pipe status diagram, cutting control commands, actual cutting result data, and transmission status data are input into the data preprocessing module of the target intelligent control system. This module has a built-in data format normalization unit and a data validity verification unit. The data format normalization unit adjusts the field arrangement and expression of the four types of data according to the data storage specifications preset by the central database. The data validity verification unit checks the completeness of the four types of data one by one. After confirming that there is no missing data or logical contradiction, it generates standardized data that meets the storage requirements.

[0073] The preprocessed standardized data is transmitted to the central database access module of the target intelligent control system. This module has a built-in data classification unit and a storage path allocation unit. The data classification unit divides the standardized data into graph data, instruction data, result data and status data according to the data attributes. The storage path allocation unit assigns a corresponding dedicated storage path to each type of data to ensure that the storage locations of different types of data are independent of each other and easy to retrieve.

[0074] The central database's write execution unit writes various standardized data sequentially to the corresponding storage areas according to the storage path allocated by the access module. During the writing process, a data storage log is generated synchronously. The log content includes data writing time, data type, and storage location information. After the writing is completed, a data storage success identifier is generated to confirm that the continuous pipe status map, cutting control commands, actual cutting result data, and transmission status data have all been stored in the central database of the target intelligent control system.

[0075] The subsequent management interface module of the central database constructs a data association index based on the four types of data already stored. This index establishes the relationship between the continuous pipe status map and the execution of cutting control commands, the correspondence between cutting control commands and actual cutting result data, and the status matching relationship between actual cutting result data and transmission status data. Through this index, joint query and traceability of the four types of data can be realized to achieve subsequent collaborative management of the target intelligent control system.

[0076] The beneficial effects are that by preprocessing and validating the four types of data, the data stored in the central database is guaranteed to be standardized and complete; the data is managed in an orderly manner through classified storage and path allocation; storage logs are generated synchronously for easy traceability; and the construction of related indexes enables joint queries of multiple types of data, providing comprehensive and reliable data support for the subsequent collaborative management of the intelligent control system.

[0077] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0078] This application embodiment can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0079] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A data interaction and collaborative management method for an intelligent control system of a pipe cutting machine, characterized in that, The method includes: S1. Acquire real-time length measurement data stream and inner wall image data stream of the pipe during the operation of the target intelligent control system; S2. Defect location is performed on the inner wall image data stream of the pipe to obtain the inner wall quality status data of the inner wall image data stream of the pipe; S3. The inner wall quality status data and the real-time length measurement data stream of the pipe are spatiotemporally correlated and aligned to construct a continuous pipe status map of the target intelligent control system. S4. Locate the pipe section with the target cutting point in the continuous pipe status map, and make a collaborative cutting decision on the pipe section based on the quality label attribute of the pipe section to obtain the cutting control command of the target intelligent control system. S5. Transmit the cutting control command to the cutting execution stage of the target intelligent control system, and simultaneously collect the actual cutting result data and transmission status data fed back by the cutting execution stage; S6. The continuous pipe status map, the cutting control command, the actual cutting result data and the transmission status data are stored in the central database of the target intelligent control system to realize subsequent collaborative management of the target intelligent control system.

2. The data interaction and collaborative management method of the intelligent control system for a pipe cutting machine as described in claim 1, characterized in that, The real-time length measurement data stream and inner wall image data stream of the pipe during the operation of the intelligent control system for acquiring the target include: Continuous non-contact measurement of the pipe material during the conveying process of the target intelligent control system is performed to obtain the original length measurement signal of the target intelligent control system; The original length measurement signal is converted from analog to digital to obtain the digital signal sequence of the target intelligent control system; The digital signal sequence is timestamped to obtain the real-time pipe length measurement data stream of the target intelligent control system. Based on the real-time length measurement data stream of the pipe, the inner wall image data stream of the pipe of the target intelligent control system is collected synchronously.

3. The data interaction and collaborative management method of the intelligent control system for a pipe cutting machine as described in claim 1, characterized in that, The step of locating defects in the pipe inner wall image data stream to obtain the inner wall quality status data of the pipe inner wall image data stream includes: Contrast enhancement is performed on the inner wall image data stream of the pipe to obtain a standardized inner wall image sequence of the inner wall image data stream; Extract the texture and edge features of the standardized inner wall image sequence to obtain the inner wall feature descriptor subset of the pipe inner wall image data stream; Anomaly pattern recognition is performed on the inner wall feature description subset to obtain preliminary defect area identifiers for the inner wall image data stream of the pipe. Based on the preliminary defect area identification, the defect classification and determination of the inner wall image data stream of the pipe are performed to obtain the inner wall quality status data of the inner wall image data stream of the pipe.

4. The data interaction and collaborative management method of the intelligent control system for a pipe cutting machine as described in claim 3, characterized in that, The step of performing abnormal pattern recognition on the subset of inner wall feature descriptions to obtain preliminary defect region identifiers for the inner wall image data stream of the pipe includes: The feature space distribution evolution of the inner wall feature descriptor subset is performed to obtain the feature distribution map of the inner wall feature descriptor subset; Based on the feature distribution map and the preset reference feature vector, the deviation of the inner wall feature descriptor subset is calculated to obtain the anomaly confidence of the feature vectors in the inner wall feature descriptor subset. The formula for calculating the anomaly confidence is as follows: ； In the formula, This indicates the anomaly confidence level. This indicates the number of the reference feature vectors. This represents the feature vector in the subset of inner wall feature descriptions. Indicates the first The mean vector of the reference eigenvectors, Indicates the first The standard deviation of each reference feature vector This represents the preset local similarity adjustment factor. Indicates the first One reference feature vector, The Euclidean norm of a vector. This represents a similarity measurement function between vectors; Adaptive threshold segmentation is performed on the abnormal confidence level to obtain abnormal feature points in the inner wall image data stream of the pipe. Spatial clustering of the abnormal feature points yields preliminary defect region identifiers for the inner wall image data stream of the pipe.

5. The data interaction and collaborative management method of the intelligent control system for a pipe cutting machine as described in claim 1, characterized in that, The step of spatiotemporally correlating and aligning the inner wall quality status data with the real-time length measurement data stream of the pipe to construct a continuous pipe status map of the target intelligent control system includes: Based on the timestamp information of the target intelligent control system, the inner wall quality status data and the real-time length measurement data stream of the pipe are time-series synchronized and aligned to obtain the time-series associated data stream of the target intelligent control system. Based on the time-series associated data stream and the pipe material delivery mapping relationship of the target intelligent control system, the real-time length measurement data stream of the pipe material is mapped to the corresponding position in the inner wall quality status data; Based on the mapping results, a location index mapping table for the target intelligent control system is established; Based on the location index mapping table, the defect location results in the inner wall quality status data are associated with the absolute length position of the pipe in the target intelligent control system to obtain the quality label sequence of the target intelligent control system. Multi-source fusion analysis is performed on the quality label sequence and the real-time length measurement data stream of the pipe to obtain the continuous state entries of the target intelligent control system; According to the preset data structure, the continuous state entries are topologically constructed to obtain the continuous pipe state map of the target intelligent control system.

6. The data interaction and collaborative management method of the intelligent control system for a pipe cutting machine as described in claim 5, characterized in that, The multi-source fusion analysis of the quality label sequence and the real-time length measurement data stream of the pipe to obtain the continuous state entries of the target intelligent control system includes: The pipe length ranges marked as defects in the quality label sequence are integrated into the defect length range set of the target intelligent control system; The real-time length measurement data stream of the pipe is discretized and segmented at preset length intervals to obtain the basic length interval sequence of the target intelligent control system; The defect length interval set is associated and matched with the basic length interval sequence, and based on the matching result, the basic length interval is labeled with a quality status label to obtain the length interval segment of the target intelligent control system. Logically link the length interval segments according to the increasing order of pipe length to obtain the continuous state entries of the target intelligent control system.

7. The data interaction and collaborative management method of the intelligent control system for a pipe cutting machine as described in claim 5, characterized in that, The step of constructing a topology for the continuous state entries according to a preset data structure to obtain a continuous pipe state map of the target intelligent control system includes: The continuous state entries are instantiated into nodes to obtain the graph nodes of the target intelligent control system; Based on the length attribute of the graph nodes, directed edges are linked to adjacent graph nodes to obtain the edge relationships of the target intelligent control system. Based on the quality state labels in the continuous state entries, attribute values ​​are assigned to the graph nodes to obtain the optimized graph nodes of the target intelligent control system. Based on the optimized graph nodes and the edge relationships, a continuous pipe state graph of the target intelligent control system is constructed.

8. The data interaction and collaborative management method of the intelligent control system for a pipe cutting machine as described in claim 1, characterized in that, The process of locating the pipe segment with the target cutting point in the continuous pipe state map, and making collaborative cutting decisions for the pipe segment based on the quality label attributes of the pipe segment to obtain the cutting control command of the target intelligent control system includes: Based on preset cutting task parameters, the nodes in the continuous pipe state map are traversed to filter out candidate pipe sections in the continuous pipe state map that meet the cutting task parameters. Based on the candidate pipe sections, the quality label attributes of the candidate pipe sections are parsed to obtain the candidate section quality attribute set of the candidate pipe sections; Based on preset cutting quality standards and target cutting points, the set of quality attributes of the candidate sections is matched for cutting priority to obtain the selected pipe section of the target cutting point and the corresponding cutting decision parameters. The selected pipe section and its corresponding cutting decision parameters are structured and encoded to obtain the cutting control command of the target intelligent control system.

9. The data interaction and collaborative management method of the intelligent control system for a pipe cutting machine as described in claim 8, characterized in that, The process involves matching the quality attributes of the candidate sections based on preset cutting quality standards and target cutting points to determine cutting priority, thereby obtaining the selected pipe section at the target cutting point and its corresponding cutting decision parameters. Based on the cutting quality standard and the set of candidate segment quality attributes, the candidate pipe segments are screened for quality level to obtain a priority queue of candidate pipe segments. Based on the defect severity attribute of the segments in the candidate segment priority queue and the position offset of adjacent segments, the candidate segment priority queue is dynamically prioritized and rearranged. Based on the location attributes of the target cutting point, the uniqueness of the rearranged candidate segment priority queue is determined to obtain the selected pipe segment of the target cutting point. The quality label attributes and geometric parameters of the selected pipe section are encapsulated into cutting decision parameters to obtain the cutting decision parameters of the target cutting point.

10. The data interaction and collaborative management method of the intelligent control system for a pipe cutting machine as described in claim 1, characterized in that, The process of transmitting the cutting control command to the cutting execution stage of the target intelligent control system, and simultaneously collecting the actual cutting result data and transmission status data fed back by the cutting execution stage, includes: The cutting control commands are subjected to protocol adaptation and conversion to obtain the command adaptation data of the target intelligent control system; The instruction adaptation data is integrated and encapsulated to obtain the instruction data packet of the target intelligent control system; The instruction data packet is transmitted to the cutting execution stage of the target intelligent control system to trigger the cutting execution stage to complete the cutting operation; The system synchronously receives the feedback data stream reported in real time by the cutting execution stage; The feedback data stream is subjected to multi-channel data separation to obtain the actual cutting result data of the target intelligent control system and the transmission status data.

Images