CAD drawing error detection method and system

Abstract

The present invention discloses a CAD drawing error detection method and system, and relates to data processing related fields. The method comprises: constructing a multi-dimensional detection block; obtaining the dimension coverage of the CAD drawing, the dimension coverage being all the detection dimensions contained in the CAD drawing; matching the multi-dimensional detection block using the dimension coverage, wherein the detection block is activated for the existing dimension and disabled for the non-existent dimension; extracting data from the CAD drawing by activating the detection block, and performing error judgment on the extracted data using the preset abnormal recognition model of each dimension detection block to obtain dimension error information; and performing global error integration using the dimension error information to obtain the drawing error detection result. The method solves the technical problems of insufficient detection accuracy and efficiency of existing CAD drawing error detection, and achieves the technical effect of improving the accuracy and efficiency of error detection.

Description

Technical Field

[0001] The present application relates to data processing related fields, and in particular to a CAD drawing error detection method and system. Background Art

[0002] In the process of making and using CAD (computer-aided design) drawings, it is crucial to ensure the accuracy and completeness of the drawings. However, due to the complexity of CAD drawings and various factors in the design process, various errors often occur in the drawings. At present, manual inspection and rule-based detection methods are mainly used. Manual inspection relies on the naked eyes and experience of designers or engineers to find errors in drawings, which is inefficient; rule-based detection methods set a series of rules and use software to automatically check whether the drawings comply with these rules, thereby discovering potential errors. However, due to the diversity and complexity of CAD drawings, it is very difficult to formulate comprehensive rules, resulting in insufficient comprehensiveness of error detection.

[0003] In the current related technologies, CAD drawing error detection methods have technical problems of insufficient detection accuracy and efficiency. Summary of the invention

[0004] The present application provides a CAD drawing error detection method and system, which adopts the technical means of constructing multi-dimensional detection blocks, activating corresponding detection blocks according to the dimensional coverage of the drawing, and using a preset anomaly recognition model to perform error judgment on the extracted data, integrating the error information detected in each dimension to form a complete drawing error detection result, thereby achieving the technical effect of improving the accuracy and efficiency of error detection.

[0005] This application provides a CAD drawing error detection method, including:

[0006] Construct a multi-dimensional detection block, wherein the detection dimensions include geometric dimensions, layer dimensions, annotation dimensions, and external reference dimensions; obtain the dimensional coverage of the CAD drawing, wherein the dimensional coverage is all the detection dimensions contained in the CAD drawing; use the dimensional coverage to match the multi-dimensional detection block, wherein the detection block is activated for the existing dimensions and disabled for the non-existent dimensions; perform data matching and extraction on the CAD drawing by activating the detection block, and perform error judgment on the extracted data using the preset abnormality recognition model of each dimensional detection block to obtain dimensional error information; use the dimensional error information to perform global error integration to obtain the drawing error detection result.

[0007] In a possible implementation, the dimensional coverage of the CAD drawing is obtained by performing the following processing:

[0008] According to the source information of the CAD drawing, the identification channel is matched, wherein the source information includes drawing by local software, drawing by external software with a collaborative interface, and drawing by external software without a collaborative interface, and the identification channel corresponds to the source information one by one; when the source information is drawing by local software or drawing by external software with a collaborative interface, the operation log is obtained through the identification channel, and dimension identification is performed according to the operation log to obtain the dimension coverage; when the source information is drawing by external software without a collaborative interface, the drawing dimension is parsed through the built-in format parsing module in the identification channel to obtain the dimension coverage.

[0009] In a possible implementation, the built-in format parsing module in the recognition path performs drawing dimension parsing to obtain the dimension coverage, and performs the following processing:

[0010] The format reading submodule is used to read drawing files according to different formats; the read drawing files are imported into the file parsing submodule to obtain the header information, object table, and layer table of the file; the detection dimension is identified according to the header information, object table, and layer table of the file to obtain the dimension coverage, wherein the header information includes the file format, unit setting, version information, and existing dimension types; the object table is used to identify the geometric dimension and the annotation dimension; the layer table is used to identify the layer dimension of the drawing; and the external reference dimension is extracted through the reference record in the object table or the layer table.

[0011] In a possible implementation, the abnormal recognition model preset in each dimension detection block is used to perform error judgment on the extracted data, obtain dimension error information, and perform the following processing:

[0012] Activate the pre-trained anomaly recognition model, wherein the pre-trained anomaly recognition model is a training data set and a test data set constructed according to a historical error sample set, and perform anomaly discriminator training through a machine learning algorithm. The machine learning model obtained after reaching the convergence target is used to identify and output the error features of each dimension; use the extracted data corresponding to the detection dimension as an input variable, perform error detection through the pre-trained anomaly recognition model, and obtain the error information of the dimension.

[0013] In a possible implementation, the dimension error information is obtained by performing the following processing:

[0014] When the source information of the CAD drawing to be detected is drawn by local software or external software with a collaborative interface, obtain the operation log of the CAD drawing to be detected; perform operation regression decomposition of each detection dimension according to the operation log to obtain a dimensional operation data stream; perform error detection on the CAD drawing to be detected according to the dimensional operation data stream to obtain abnormal operation data; and locate the dimensional error information according to the abnormal operation data.

[0015] In a possible implementation, error detection is performed on the CAD drawing to be detected according to the dimension operation data stream, abnormal operation data is obtained, and the following processing is performed:

[0016] According to the input target parameters of the CAD drawing to be detected, the target operation data stream is parsed; the dimensional operation data stream is aligned and identified using the target operation data stream to obtain a positive exception identification result; an abnormal operation data stream is obtained, and the abnormal operation data stream is obtained according to the historical erroneous operation samples of each detection dimension; the dimensional operation data stream is reversely identified using the abnormal operation data stream to obtain a reverse exception identification result; the forward exception identification result and the reverse exception identification result are integrated to obtain the abnormal operation data.

[0017] In a possible implementation, the dimension error information is used to perform global error integration, obtain a drawing error detection result, and perform the following processing:

[0018] The operation timestamp is located according to the abnormal operation data, and time integration is performed based on the dimensional error information to construct a timeline view; the operation superposition effect analysis is performed according to the timeline view to track the drawing error data; the drawing error data is added to the timeline view to locate the position of all error information and its operation timeline, and the drawing error detection result is generated.

[0019] In a possible implementation, the following processing is performed:

[0020] The abnormal operation data is clustered according to time features or operation type features to construct an error operation set; operation specification analysis is performed based on the error operation set to generate operation reminder data, and a mapping association between the operation reminder data and user information is established.

[0021] This application also provides a CAD drawing error detection system, including:

[0022] A multi-dimensional detection construction module, the multi-dimensional detection construction module is used to construct a multi-dimensional detection block, wherein the detection dimensions include geometric dimensions, layer dimensions, annotation dimensions, and external reference dimensions; a drawing dimension coverage acquisition module, the drawing dimension coverage acquisition module is used to obtain the dimension coverage of the CAD drawing, and the dimension coverage is all the detection dimensions contained in the CAD drawing; a dimension detection matching module, the dimension detection matching module is used to match the multi-dimensional detection block using the dimension coverage, wherein the detection block is activated for the existing dimension and disabled for the non-existent dimension; an error judgment module, the error judgment module is used to perform data matching and extraction on the CAD drawing by activating the detection block, and perform error judgment on the extracted data using the preset abnormality recognition model of each dimension detection block to obtain dimension error information; a drawing error detection result acquisition module, the drawing error detection result acquisition module is used to perform global error integration using the dimension error information to obtain a drawing error detection result.

[0023] The CAD drawing error detection method and system proposed in the present application first construct a multi-dimensional detection block, wherein the detection dimensions include geometric dimensions, layer dimensions, annotation dimensions, and external reference dimensions, and then obtain the dimensional coverage of the CAD drawing, wherein the dimensional coverage is all the detection dimensions contained in the CAD drawing, and then use the dimensional coverage to match the multi-dimensional detection block, wherein the detection block is activated for the existing dimensions and disabled for the non-existent dimensions, and the CAD drawing is matched and extracted by activating the detection block, and the extracted data is judged for errors using the preset abnormality recognition model of each dimensional detection block to obtain dimensional error information, and finally the dimensional error information is used for global error integration to obtain the drawing error detection result, thereby achieving the technical effect of improving the accuracy and efficiency of error detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the technical solution of the embodiment of the present invention, the accompanying drawings of the embodiment of the present invention will be briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the system according to the embodiment of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.

[0025] Figure 1 A flow chart of a CAD drawing error detection method provided in an embodiment of the present application.

[0026] Figure 2 A schematic diagram of the structure of a CAD drawing error detection system provided in an embodiment of the present application.

[0027] Explanation of the reference numerals: multi-dimensional detection construction module 10, drawing dimension coverage range acquisition module 20, dimension detection matching module 30, error judgment module 40, drawing error detection result acquisition module 50. DETAILED DESCRIPTION

[0028] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.

[0029] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0030] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.

[0031] The present application embodiment provides a CAD drawing error detection method, such as Figure 1 As shown, the method includes:

[0032] Step S100, construct a multi-dimensional detection block, wherein the detection dimensions include geometric dimensions, layer dimensions, annotation dimensions, and external reference dimensions. Specifically, according to the common error types and detection requirements of CAD drawings, the detection dimensions are determined, including geometric dimensions (used to detect errors in geometric objects, such as overlaps, unclosed paths, incorrect shape connections, etc.), layer dimensions (check the use of layers, detect unused layers, irregular layer management, and incorrect layer classification, etc.), annotation dimensions (check the correctness of annotation size, angle and other information, including checking whether the annotation is associated with the geometric object and whether the annotation format is correct) and external reference dimensions (used to check the reference and integrity of the external reference, to ensure that the external file path is correct and the referenced drawings are not lost or wrong). For each detection dimension, a corresponding detection block is constructed. These detection blocks are independent modules containing preset anomaly recognition models, which are used to perform error detection on data of specific dimensions.

[0033] Step S200, obtain the dimension coverage of the CAD drawing, and the dimension coverage is all the detection dimensions contained in the CAD drawing. Specifically, according to the source information of the CAD drawing, the corresponding identification channel is matched, and the dimension is identified through the identification channel to determine the dimension coverage of the drawing, that is, the dimensions corresponding to all error types involved in the CAD drawing.

[0034] In a possible implementation, the dimensional coverage of the CAD drawing is obtained, and step S200 further includes step S210, matching the identification channel according to the source information of the CAD drawing, wherein the source information includes drawing by local software, drawing by external software with a collaborative interface, and drawing by external software without a collaborative interface, and the identification channel corresponds to the source information one by one. Specifically, the source of the CAD drawing, that is, the creation or source method of the CAD drawing, is determined by checking the metadata, file extension, or user input of the drawing file. Once the source of the drawing is determined, the drawing is matched with the corresponding identification channel according to the preset configuration. These identification channels are a series of processing flows defined within the system, which are used to use appropriate processing methods according to the source information of the drawing. Step S220, when the source information is drawn by local software or drawn by external software with a collaborative interface, the operation log is obtained through the identification channel, and the dimension identification is performed according to the operation log to obtain the dimensional coverage. Specifically, for drawings drawn by local software, the operation log (log file recording the operations performed by the user in the CAD software) is directly extracted from the log file of the software; for drawings drawn by external software with a collaborative interface, the operation log is remotely obtained through the collaborative interface. The obtained operation log is parsed to extract information related to the drawing dimensions, such as geometric dimensions, layer dimensions, annotation dimensions, etc. Step S230, when the source information is drawn by external software without a collaborative interface, the drawing dimensions are parsed by the built-in format parsing module in the identification path to obtain the dimension coverage. Specifically, for drawings drawn by external software without a collaborative interface, the drawing file is first read, and the drawing file is parsed using the built-in format parsing module (a module used to parse CAD drawings of different formats within the system) to extract the header information, object table, layer table and other key information of the drawing. According to the parsed drawing information, the dimensions contained in the drawing are identified. This implementation method can flexibly process drawings from different sources by matching the identification channel according to the source of the drawing, ensuring accurate acquisition of the dimension information of the drawing.

[0035] In a possible implementation, the dimension parsing of the drawing is performed by the built-in format parsing module in the identification path to obtain the dimension coverage, and step S230 further includes step S231, using the format reading submodule to read the drawing file according to different formats. Specifically, the format reading submodule is a part of the built-in format parsing module in the identification path, and is responsible for reading the drawing file according to different file formats (such as DWG, DXF, etc.). In the format reading submodule, the format of the drawing file is identified by checking the file extension, reading the header information of the file, or using the file format detection algorithm. Step S232, importing the read drawing file into the file parsing submodule to obtain the header information, object table, and layer table of the file. Specifically, the file parsing submodule receives the content of the drawing file read by the format reading submodule, and selects a corresponding reading strategy, including determining how to parse the header information, object table and layer table of the file, and how to process special structures or data in the file. For example, if the file is in DWG format, a binary parsing method is used to extract the object table and layer table. According to the selected reading strategy, the file parsing submodule parses the header information of the drawing file to extract key information such as the file format, unit setting, version information, and existing dimension types. Parse the object table and layer table in the drawing file, wherein the object table contains the definition and properties of all objects in the drawing, such as lines, shapes, text, etc.; the layer table records the definition and properties of all layers in the drawing, such as layer name, color, line type, etc. Step S233, identify the detection dimension according to the header information, object table, and layer table of the file, and obtain the dimension coverage range, wherein the header information contains the file format, unit setting, version information, and existing dimension types, the object table is used to identify the geometric dimension and the annotation dimension, the layer table is used to identify the layer dimension of the drawing, and the external reference dimension is extracted through the reference record in the object table or the layer table. Specifically, the header information contains the global properties of the drawing file, wherein the file format is used to determine the parsing method, the unit setting contains the unit used by the file, such as millimeters, inches, etc., which is used to parse the annotation dimension, and the version information contains the version number of the CAD file or the extended function mark. If the file uses a higher version of the extended function (such as 3D geometry, external reference enhancement), it is prompted that the geometric dimension or external reference dimension needs to be parsed in detail.

[0036] The object table is the part of the CAD file that records all drawing elements such as geometric objects, annotations, and block references. By parsing the object table, the existence and complexity of geometric dimensions and annotation dimensions are identified. For example, the geometric object types (such as line segments, polylines, circles, arcs, block references, regions, 3D entities, etc.) are extracted from the object table to count the number and type of geometric objects, and determine whether a large number of complex geometric elements (such as 3D entities, nested blocks, etc.) are included. If there are a large number of three-dimensional geometric objects or complex regions in the object table, the complexity of the geometric dimension is high, indicating that more detailed geometric error detection is required. By traversing the object table, the geometric type, geometric structure, and reference relationship of the object are analyzed. If a large number of complex geometric objects are found, the geometric dimension is activated and marked as high complexity. Similarly, objects related to annotations (such as linear dimension annotations, angular annotations, radius annotations, text annotations, etc.) are extracted from the object table to check whether the annotations are associated with geometric objects, whether the references of the annotation objects are correct, and whether the annotations meet the design standards of the drawings. If the number of annotation objects in the object table is large or the association between the annotations and geometric objects is complex, the complexity of the annotation dimension is high. By parsing the type and reference relationship of the annotation object, the existence and complexity of the annotation dimension are determined. When the annotation is not correctly associated with the geometric object, the annotation dimension is activated and needs to be checked in detail.

[0037] The layer table is a structure for managing objects in CAD files. It records the names, states (visibility, lock state, etc.) of all layers in the drawing, and the object types stored on each layer. By parsing the layer table, the layer dimension is identified. For example, the number of layers is obtained from the layer table. If the drawing contains multiple layers and each layer is used to manage different object types (such as geometric objects, annotations, external references, etc.), the layer dimension is activated, and the state of each layer is checked (whether it is visible, whether it is locked, whether there are objects). If some layers are not used or the management is chaotic (such as too many layers without objects, duplicate layers, etc.), there may be problems with the layer dimension. According to the object distribution of each layer in the layer table, it is judged whether the layer management is reasonable (whether the geometric objects are concentrated in the geometric layer, and whether the annotation objects are located in the annotation layer). By traversing the layer table, the usage and object distribution of each layer are analyzed. If the layer management is complex, the layer dimension is activated, and further detection of layer management problems is required.

[0038] External references are a way for CAD files to reference external files (such as other DWG files, PDF files, etc.). External references are used to reference external resources in drawing design. The status of external references is recorded in the object table or layer table, and the external reference dimension is identified by parsing this information. For example, the reference path and status of the external reference file are obtained from the object table or layer table. If the drawing references an external file, the external reference dimension is activated, and the path of the external reference is checked to see if it is valid, whether the external file is missing, and whether the file can be loaded successfully. If the drawing references multiple external files or the referenced file is large and the path is complex, the complexity of the external reference dimension is high, and more detailed error detection is required. By parsing the referenced objects in the object table, analyzing the referenced external file type and path status, if the referenced file path is invalid or the file is missing, there may be a problem with the external reference dimension, which requires further detection. This implementation method uses the format reading submodule and the file parsing submodule to enable the system to read drawing files in different formats and parse out key information such as header information, object table and layer table. By parsing the header information, object table and layer table of the drawing file, it can accurately identify the dimension type contained in the drawing and extract the corresponding dimension information, thereby improving the accuracy of dimension recognition.

[0039] Step S300, using the dimension coverage to match the multi-dimensional detection block, wherein the detection block is activated for the existing dimension and disabled for the non-existent dimension. Specifically, the dimension coverage of the CAD drawing is compared with the multi-dimensional detection block, and according to the dimension coverage of the CAD drawing, the detection block corresponding to the existing dimension is activated for error detection, and the detection block corresponding to the non-existent dimension is disabled.

[0040] Step S400, the CAD drawings are matched and extracted by activating the detection blocks, and the extracted data are judged for errors using the abnormality recognition models preset in the detection blocks of each dimension, so as to obtain dimension error information. Specifically, the activated detection blocks match and extract the data of the CAD drawings, and extract the data related to the detection dimension. The extracted data are judged for errors using the abnormality recognition models preset in the detection blocks of each dimension (machine learning models for identifying error features of specific dimensions). These models are constructed based on historical error sample sets and can identify error features of specific dimensions.

[0041] In a possible implementation, the abnormality recognition model preset in each dimension detection block is used to perform error judgment on the extracted data to obtain dimension error information. Step S400 further includes step S410, activating the pre-trained abnormality recognition model, wherein the pre-trained abnormality recognition model is a training data set and a test data set constructed according to a historical error sample set, and the abnormality discriminator is trained by a machine learning algorithm. The machine learning model obtained after reaching the convergence target is used to identify and output the error features of each dimension. Specifically, the abnormality recognition model is a trained machine learning model, and its function is to identify the error features in the input data. The training data set of the abnormality recognition model is constructed by a historical error sample set, and contains samples of various known errors, such as geometric overlap, unclosed paths, improper layer management, and annotation errors. These samples are collected from previous CAD projects and are confirmed by expert annotation. The test data set is also derived from the historical error sample set, but is independent of the training data set, and is used to verify the performance of the model. The training data set is trained using a supervised learning algorithm, and the model parameters are continuously adjusted during the training process to minimize the error classification rate until the convergence target is reached. The accuracy of the model is verified using the test data set to ensure that the model can accurately identify various errors. Save the trained model, and when error detection is needed, load and activate the pre-trained anomaly recognition model. Step S420, use the extracted data corresponding to the detection dimension as the input variable, perform error detection through the pre-trained anomaly recognition model, and obtain the dimension error information. Specifically, the extracted data corresponding to the detection dimension is used as the input variable. These extracted data are extracted from the CAD drawing and contain information on dimensions such as geometry, layers, annotations, and external references. The input data is input into the pre-trained anomaly recognition model. The model uses the rules learned internally to perform classification and judgment based on the characteristics of the input data, and outputs the dimension error information, that is, the identified error type and its location. This implementation method makes full use of past experience and knowledge by collecting historical error sample sets, thereby improving the accuracy and reliability of the model. The supervised learning algorithm is used for model training to improve the efficiency and accuracy of CAD drawing error recognition.

[0042] In a possible implementation, dimension error information is obtained, and step S400 further includes step S430, when the source information of the CAD drawing to be detected is drawn by local software or drawn by external software with a collaborative interface, the operation log of the CAD drawing to be detected is obtained. Specifically, the source information of the CAD drawing to be detected is judged. If the drawing is drawn by local software or by external software with a collaborative interface, it indicates that the operation log is supported. At this time, the operation log is read through the corresponding interface or log file. The operation log records a series of operations of the user in the process of drawing the drawing, such as creating geometric objects, modifying layers, adding annotations, and referencing external files. Step S440, according to the operation log, the operation regression decomposition of each detection dimension is performed to obtain the dimension operation data stream. Specifically, the operation log is divided according to the detection dimension (such as geometry, layer, annotation, external reference, etc.), and each dimension corresponds to a series of related operations, and these operations together constitute the operation data stream of the dimension. The operation data stream of each dimension is regressively decomposed, that is, a series of continuous operations are split into independent operation units, and these operation units can be creating or modifying a certain geometric object, changing layer properties, adding or deleting annotations, etc. Step S450, error detection is performed on the CAD drawing to be detected according to the dimension operation data stream to obtain abnormal operation data. Specifically, a series of error rules are defined based on historical experience and expert knowledge, and these rules are used to identify abnormal operation data in the operation data stream. For example, the rules include "it is not allowed to create different types of geometric objects on the same layer" and "the units of annotations must be consistent with the drawing units". The dimension operation data stream is matched with the error rules to find out the abnormal operation data that violates the rules. These abnormal operation data represent errors or inconsistencies in the drawings. Step S460, the dimension error information is located according to the abnormal operation data. Specifically, for each abnormal operation data, the specific error information is located according to the dimension to which it belongs and the operation content, including the error type (such as geometric error, layer error, annotation error, etc.), the error location (such as a specific location or object in the drawing) and possible repair suggestions. This implementation method accurately locates the source of the error by tracing each step of the user's operation using the operation log, thereby improving the accuracy of error detection.

[0043] In a possible implementation, the CAD drawing to be detected is error-detected according to the dimension operation data stream to obtain abnormal operation data, and step S450 further includes step S451, parsing the target operation data stream according to the input target parameters of the CAD drawing to be detected. Specifically, the input target parameters are extracted from the metadata of the CAD drawing to be detected, including the design specifications, size requirements, material types, etc. of the drawing. According to the input target parameters, an idealized target operation data stream is generated, which represents a series of operations that should be performed when the drawing is correctly drawn according to the specifications and requirements. Step S452, the dimension operation data stream is aligned and identified using the target operation data stream to obtain a positive abnormal identification result. Specifically, the dimension operation data stream is aligned with the target operation data stream, including matching each operation node in the dimension operation data stream with the corresponding node in the target operation data stream. In the alignment process, the operation nodes in the dimension operation data stream that do not match the target operation data stream are identified, and these nodes may represent errors or inconsistencies in the drawing, and these inconsistent operation nodes are marked as positive abnormal identification results. Step S453, obtain an abnormal operation data stream, and the abnormal operation data stream is obtained according to the historical error operation samples of each detection dimension. Specifically, collect error operation samples of each detection dimension from historical data, and these samples represent the common error types in previous drawings. Based on the collected historical error operation samples, generate an abnormal operation data stream, which contains various possible error operation modes. Step S454, use the abnormal operation data stream to perform reverse abnormality recognition on the dimension operation data stream to obtain a reverse abnormality recognition result. Specifically, the dimension operation data stream is matched with the abnormal operation data stream, including comparing each operation node in the dimension operation data stream with the error pattern in the abnormal operation data stream. In the matching process, identify the operation nodes in the dimension operation data stream that conform to the error pattern in the abnormal operation data stream, and mark these operation nodes that conform to the error pattern as reverse abnormality recognition results. Step S455, integrate the forward abnormality recognition result and the reverse abnormality recognition result to obtain the abnormal operation data. Specifically, the forward anomaly recognition results and the reverse anomaly recognition results are integrated, including removing duplicate recognized operation nodes and merging operation nodes with similar error characteristics. After integration, preliminary abnormal operation data is generated. The preliminary abnormal operation data is verified using the dimensional error information output by the pre-trained anomaly recognition model, including comparing each abnormal operation node with the error type and pattern output by the model to verify whether it actually represents an error.According to the verification results, the preliminary abnormal operation data is adjusted. If an abnormal operation node is recognized by the model as not an error (i.e., a false positive), it is removed from the abnormal operation data; on the contrary, if the model recognizes a new error (i.e., an error detected by the model but not previously recognized), it is added to the abnormal operation data. After verification and adjustment, the final abnormal operation data is generated, which represents all verified errors or inconsistencies in the CAD drawings to be detected. This implementation method combines forward anomaly recognition and reverse anomaly recognition to perform error detection on the dimensional operation data stream from two different angles, increasing the comprehensiveness and accuracy of error detection. The dimensional error information output by the pre-trained anomaly recognition model is used as verification data to verify the abnormal operation data, further improving the accuracy and reliability of error detection, and ensuring that the final output abnormal operation data is accurate and reliable.

[0044] Step S500, use the dimensional error information to perform global error integration to obtain drawing error detection results. Specifically, the dimensional error information output by each dimensional detection block is integrated to form a global error view. Based on the integrated error information, the error detection results of the CAD drawing are generated, including information such as the type, location, and severity of the error, which is used to guide subsequent drawing correction work. The embodiment of the present application adopts the construction of multi-dimensional detection blocks, activates the corresponding detection blocks according to the dimensional coverage of the drawing, and uses a preset anomaly recognition model to perform error judgment on the extracted data, integrates the error information detected in each dimension, and forms a complete drawing error detection result and other technical means, which achieves the technical effect of improving the accuracy and efficiency of error detection.

[0045] In a possible implementation, the dimension error information is used to perform global error integration to obtain the drawing error detection result. Step S500 further includes step S510, locating the operation timestamp according to the abnormal operation data, performing time integration based on the dimension error information, and constructing a timeline view. Specifically, the abnormal operation data is traversed, and the corresponding timestamp is extracted for each abnormal operation node. This timestamp represents the time point when the abnormal operation occurs during the CAD drawing process. Time integration is performed according to the dimension error information (such as error type, error location, etc.) and timestamp in the abnormal operation data, and the abnormal operation nodes with similar occurrence time are classified and sorted. After time integration, a timeline view is constructed. This timeline view takes the timeline as the baseline, arranges the abnormal operation nodes in the order of their occurrence time, and annotates the error type and location of each node. Step S520, according to the timeline view, an operation superposition effect analysis is performed to track the drawing error data. Specifically, based on the timeline view, an operation superposition effect analysis is performed, including tracking the subsequent errors or impacts that may be caused by each abnormal operation node, as well as the correlation and causal relationship between these errors. By analyzing the time intervals, error types, and locations between abnormal operation nodes, the possible superposition effects between them are evaluated, and the chain reactions or complex error patterns caused by a single error are identified. After analysis, a superposition effect analysis report is generated, which describes the subsequent errors that may be caused by each abnormal operation node and its scope of influence, as well as the correlation and causal relationship between these errors. Step S530, the drawing error data is added to the timeline view, the location of all error information and its operation timeline are located, and the drawing error detection result is generated. Specifically, all drawing error data identified in the superposition effect analysis report (including the original abnormal operation data and the subsequent errors caused by the superposition effect) are added to the timeline view. In the timeline view, the specific location (such as the coordinate point in the CAD drawing) and the operation timeline (i.e., the occurrence time point and duration) of each drawing error data are located. Finally, the final drawing error detection result is generated according to the drawing error data in the timeline view, which includes the location, type, occurrence time, and possible superposition effect of all errors. This implementation method simplifies the error detection process and improves detection efficiency through time integration. By analyzing the operation superposition effect, it identifies chain reactions or complex error patterns caused by a single error, and more accurately evaluates error conditions in the drawings, achieving the technical effect of effectively improving the accuracy and efficiency of CAD drawing error detection.

[0046] In a possible implementation, step S500 further includes step S540, clustering the abnormal operation data by time features or operation type features, and constructing an error operation set. Specifically, for each abnormal operation data, extract its time features (such as the time point of occurrence, time interval, etc.) and operation type features (such as operation name, operation parameters, etc.). Use a clustering algorithm to perform cluster analysis on the extracted features, classify abnormal operation data with similar features into one category, and construct an error operation set based on the clustering results, each error operation set represents a specific type of error operation mode. Step S550, perform operation specification analysis based on the error operation set, generate operation reminder data, and establish a mapping association between the operation reminder data and user information. Specifically, analyze the constructed error operation set to find out the root cause of the abnormal operation. According to the analysis results, formulate targeted operation reminder data to guide users to avoid performing operations that may cause abnormalities. Establish a mapping association between the generated operation reminder data and user information, that is, when a user performs an operation that may trigger an abnormality, a reminder message is automatically sent to the user. This implementation method performs cluster analysis on abnormal operation data, identifies common erroneous operation patterns, and formulates targeted operation reminder data accordingly. When users perform operations that may trigger abnormalities, they can be reminded in time to avoid potential problems.

[0047] In the above, refer to Figure 1 The CAD drawing error detection method according to the embodiment of the present invention is described in detail. Figure 2 A CAD drawing error detection system according to an embodiment of the present invention is described.

[0048] The CAD drawing error detection system according to the embodiment of the present invention is used to solve the technical problems of insufficient detection accuracy and efficiency of existing CAD drawing error detection, and achieve the technical effect of improving the accuracy and efficiency of error detection. The CAD drawing error detection system includes: a multi-dimensional detection construction module 10, a drawing dimension coverage range acquisition module 20, a dimension detection matching module 30, an error determination module 40, and a drawing error detection result acquisition module 50.

[0049] The multi-dimensional detection construction module 10 is used to construct a multi-dimensional detection block, wherein the detection dimensions include geometric dimensions, layer dimensions, annotation dimensions, and external reference dimensions; the drawing dimension coverage acquisition module 20 is used to obtain the dimension coverage of the CAD drawing, and the dimension coverage is all the detection dimensions contained in the CAD drawing; the dimension detection matching module 30 is used to match the multi-dimensional detection block using the dimension coverage, wherein the detection block is activated for the existing dimension and disabled for the non-existent dimension; the error judgment module 40 is used to perform data matching and extraction on the CAD drawing by activating the detection block, and perform error judgment on the extracted data using the preset abnormality recognition model of each dimensional detection block to obtain dimensional error information; the drawing error detection result acquisition module 50 is used to perform global error integration using the dimensional error information to obtain the drawing error detection result.

[0050] Below, the specific configuration of the drawing dimension coverage acquisition module 20 will be described in detail. As described above, to obtain the dimension coverage of the CAD drawing, the drawing dimension coverage acquisition module 20 may further include: an identification channel matching unit is used to match the identification channel according to the source information of the CAD drawing, wherein the source information includes local software drawing, external software drawing with a collaborative interface, and external software drawing without a collaborative interface, and the identification channel corresponds to the source information one by one; the dimension coverage acquisition unit is used to obtain the operation log through the identification channel when the source information is local software drawing or external software drawing with a collaborative interface, perform dimension identification according to the operation log, and obtain the dimension coverage; when the source information is external software drawing without a collaborative interface, perform drawing dimension parsing through the built-in format parsing module in the identification channel to obtain the dimension coverage.

[0051] Among them, the built-in format parsing module in the identification path performs drawing dimension parsing to obtain the dimension coverage range, and the dimension coverage range acquisition unit may further include: a drawing file reading subunit is used to use the format reading submodule to read the drawing file according to different formats; the file parsing subunit is used to import the read drawing file into the file parsing submodule to obtain the file header information, object table, and layer table; the detection dimension identification subunit is used to identify the detection dimension according to the header information, object table, and layer table of the file to obtain the dimension coverage range, wherein the header information includes file format, unit setting, version information and existing dimension type, the object table is used to identify geometric dimensions and annotation dimensions, the layer table is used to identify the layer dimensions of the drawing, and the external reference dimension is extracted through the reference record in the object table or the layer table.

[0052] The specific configuration of the error determination module 40 will be described in detail below. As described above, the error determination of the extracted data is performed using the preset abnormality recognition model of each dimension detection block to obtain the dimension error information. The error determination module 40 may further include: an abnormality recognition model activation unit is used to activate the pre-trained abnormality recognition model, wherein the pre-trained abnormality recognition model is a training data set and a test data set constructed according to a historical error sample set, and the abnormality discriminator is trained by a machine learning algorithm. The machine learning model obtained after reaching the convergence target is used to identify and output the error features of each dimension; the error detection unit is used to use the extracted data corresponding to the detection dimension as an input variable, and perform error detection through the pre-trained abnormality recognition model to obtain the dimension error information.

[0053] Among them, to obtain dimension error information, the error judgment module 40 may further include: an operation log acquisition unit is used to obtain the operation log of the CAD drawing to be detected when the source information of the CAD drawing to be detected is drawn by local software or external software with a collaborative interface; an operation regression decomposition unit is used to perform operation regression decomposition of each detection dimension according to the operation log to obtain a dimension operation data stream; an abnormal operation data acquisition unit is used to perform error detection on the CAD drawing to be detected according to the dimension operation data stream to obtain abnormal operation data; a dimension error information locating unit is used to locate the dimension error information according to the abnormal operation data.

[0054] Among them, error detection is performed on the CAD drawing to be detected according to the dimensional operation data stream to obtain abnormal operation data, and the abnormal operation data acquisition unit may further include: a target operation data stream parsing subunit is used to parse the target operation data stream according to the input target parameters of the CAD drawing to be detected; a forward abnormality recognition subunit is used to align and recognize the dimensional operation data stream using the target operation data stream to obtain a forward abnormality recognition result; an abnormal operation data stream acquisition subunit is used to obtain an abnormal operation data stream, and the abnormal operation data stream is obtained according to historical error operation samples of each detection dimension; a reverse abnormality recognition subunit is used to perform reverse abnormality recognition on the dimensional operation data stream using the abnormal operation data stream to obtain a reverse abnormality recognition result; an abnormal recognition result integration subunit is used to integrate the forward abnormality recognition result and the reverse abnormality recognition result to obtain the abnormal operation data.

[0055] The specific configuration of the drawing error detection result acquisition module 50 will be described in detail below. As described above, the dimension error information is used to perform global error integration to obtain the drawing error detection result. The drawing error detection result acquisition module 50 may further include: a timeline view construction unit for locating the operation timestamp according to the abnormal operation data, performing time integration based on the dimension error information, and constructing a timeline view; an operation superposition effect analysis unit for performing operation superposition effect analysis according to the timeline view and tracking the drawing error data; and a drawing error detection result generation unit for adding the drawing error data to the timeline view, locating the position of all error information and its operation timeline, and generating the drawing error detection result.

[0056] Among them, the drawing error detection result acquisition module 50 can further include: an error operation set construction unit is used to cluster the abnormal operation data according to time characteristics or operation type characteristics to construct an error operation set; an operation reminder data generation unit is used to perform operation specification analysis based on the error operation set, generate operation reminder data, and establish a mapping association between the operation reminder data and user information.

[0057] The CAD drawing error detection system provided in the embodiment of the present invention can execute the CAD drawing error detection method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0058] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules may be used and run on the user terminal and / or server, and the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.

[0059] The above specific implementation manner does not constitute a limitation to the protection scope of the present application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application should be included in the protection scope of the present application. In some cases, the actions or steps recorded in the present application can be performed in an order different from that in the embodiment and can still achieve the desired results. In addition, the process depicted in the accompanying drawings does not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. A CAD drawing error detection method, characterized in that: The method comprises: Construct a multi-dimensional detection block, where the detection dimensions include geometric dimension, layer dimension, annotation dimension, and external reference dimension; Obtaining the dimensional coverage of the CAD drawing, where the dimensional coverage is all the detection dimensions included in the CAD drawing; Matching the multi-dimensional detection blocks using the dimensional coverage, wherein the detection blocks are activated for the existing dimensions and disabled for the non-existent dimensions; The CAD drawings are matched and extracted by activating the detection blocks, and the extracted data are error-judged by using the abnormality recognition models preset in the detection blocks of each dimension to obtain dimension error information; The dimension error information is used to perform global error integration to obtain a drawing error detection result.

2. The CAD drawing error detection method according to claim 1, characterized in that: The dimensional coverage of the CAD drawing obtained includes: Matching identification channels according to source information of the CAD drawings, wherein the source information includes drawing by local software, drawing by external software with a collaborative interface, and drawing by external software without a collaborative interface, and the identification channels correspond to the source information one by one; When the source information is drawn by local software or drawn by external software with a collaborative interface, an operation log is obtained through an identification channel, and dimension identification is performed according to the operation log to obtain the dimension coverage; When the source information is drawn by external software without a collaborative interface, the dimension coverage is obtained by performing drawing dimension analysis through a built-in format parsing module in the identification path.

3. The CAD drawing error detection method according to claim 2, characterized in that: The drawing dimension parsing is performed by a built-in format parsing module in the identification path to obtain the dimension coverage, including: Use the format reading submodule to read drawing files in different formats; Import the read drawing file into the file parsing submodule to obtain the file header information, object table, and layer table; According to the header information, object table and layer table of the file, the detection dimension is identified to obtain the dimension coverage, wherein the header information includes the file format, unit setting, version information and the existing dimension type, the object table is used to identify the geometric dimension and the annotation dimension, the layer table is used to identify the layer dimension of the drawing, and the external reference dimension is extracted through the reference record in the object table or the layer table.

4. The CAD drawing error detection method according to claim 2, characterized in that: Use the preset anomaly recognition model of each dimension detection block to make error judgments on the extracted data and obtain dimension error information, including: Activate the pre-trained anomaly recognition model, wherein the pre-trained anomaly recognition model is a training data set and a test data set constructed according to a historical error sample set, and an anomaly discriminator is trained through a machine learning algorithm. The machine learning model obtained after reaching the convergence target is used to identify and output error features of each dimension; The extracted data corresponding to the detection dimension is used as an input variable, and error detection is performed through the pre-trained anomaly recognition model to obtain the dimension error information.

5. The CAD drawing error detection method according to claim 4, characterized in that: The obtaining of dimension error information further includes: When the source information of the CAD drawing to be detected is drawn by local software or by external software with a collaborative interface, obtaining an operation log of the CAD drawing to be detected; Performing operation regression decomposition of each detection dimension according to the operation log to obtain a dimensional operation data stream; Perform error detection on the CAD drawing to be detected according to the dimensional operation data stream to obtain abnormal operation data; The dimension error information is located according to the abnormal operation data.

6. The CAD drawing error detection method according to claim 5, characterized in that: Performing error detection on the CAD drawing to be detected according to the dimension operation data stream to obtain abnormal operation data includes: Parsing a target operation data stream according to input target parameters of the CAD drawing to be detected; Using the target operation data stream to align and identify the dimension operation data stream, to obtain a positive anomaly identification result; Obtaining an abnormal operation data stream, wherein the abnormal operation data stream is obtained according to historical erroneous operation samples of each detection dimension; Using the abnormal operation data stream to perform reverse abnormality identification on the dimension operation data stream to obtain a reverse abnormality identification result; The positive abnormality recognition result and the reverse abnormality recognition result are integrated to obtain the abnormal operation data.

7. The CAD drawing error detection method according to claim 5, characterized in that: The dimension error information is used to perform global error integration to obtain drawing error detection results, including: Locate the operation timestamp according to the abnormal operation data, perform time integration based on the dimension error information, and construct a timeline view; Perform operation stacking effect analysis based on the timeline view and track drawing error data; The drawing error data is added to the timeline view, the positions of all error information and their operation timeline are located, and the drawing error detection result is generated.

8. The CAD drawing error detection method according to claim 7, characterized in that: Also includes: Clustering the abnormal operation data by time features or operation type features to construct an error operation set; An operation specification analysis is performed according to the erroneous operation set, operation reminder data is generated, and a mapping association between the operation reminder data and user information is established. 9.CAD drawing error detection system, characterized in that: The system is used to implement the CAD drawing error detection method according to any one of claims 1 to 8, and the system comprises: A multi-dimensional detection construction module, wherein the multi-dimensional detection construction module is used to construct a multi-dimensional detection block, wherein the detection dimensions include a geometric dimension, a layer dimension, a label dimension, and an external reference dimension; A drawing dimension coverage acquisition module, wherein the drawing dimension coverage acquisition module is used to acquire the dimension coverage of the CAD drawing, wherein the dimension coverage is all the detection dimensions included in the CAD drawing; A dimension detection matching module, the dimension detection matching module is used to match the multi-dimensional detection block using the dimension coverage, wherein the detection block is activated for the existing dimension and disabled for the non-existent dimension; An error determination module, the error determination module is used to perform data matching and extraction on the CAD drawing by activating the detection block, and perform error determination on the extracted data by using the abnormality recognition model preset in each dimension detection block to obtain dimension error information; The drawing error detection result acquisition module is used to use the dimensional error information to perform global error integration to obtain the drawing error detection result.

Images