Project quality tracing and early warning method based on data analysis

Abstract

The invention provides a project quality tracing and early warning method based on data analysis, and relates to the technical field of data analysis, and the method comprises the steps: obtaining a project quality data set, carrying out the traversal matching of the data set according to a key factor corresponding to a project stage, carrying out the replacement and recombination of a predetermined project quality data set through key feature parameters, and obtaining a project quality data set; taking the real-time data set as input information of a project quality prediction middle station and obtaining output information of the project quality prediction middle station, if the real-time prediction quality does not meet a preset project quality threshold value, sending out a primary early warning signal, and building a completed project stage based on the primary early warning signal; and a predetermined tracing early warning mechanism is called to perform tracing analysis on the completed project stage to obtain a real-time quality abnormity category, and a fine-grained early warning signal of the predetermined project is sent out, so that the problem that the discovery and control of the project quality problem lags behind the construction process and potential quality risks cannot be timely and accurately identified is solved; and the effect of improving the timeliness, accuracy and comprehensiveness of project quality management is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and in particular to an engineering quality tracing and early warning method based on data analysis. Background Art

[0002] In the construction process of engineering projects, quality control is always a key link. With the continuous expansion of engineering scale and the advancement of construction technology, traditional engineering quality management methods have gradually revealed their limitations, especially in terms of quality monitoring and early warning, facing problems such as lag, one-sidedness and inaccuracy. At present, most engineering quality supervision work still relies on traditional manual inspection and report generation, and quality inspection is mainly concentrated in the late construction or project completion stage. This method mainly relies on on-site spot checks, step-by-step inspections and witness sampling, and cannot predict and warn quality risks in real time. The method of collecting engineering quality data mainly relies on manual records, lacking comprehensive data integration and systematic analysis across stages. Although some projects will use quality inspection equipment for sampling and monitoring, the use of these equipment is often limited to certain specific project stages, and the collected data cannot fully cover all quality-related factors, and the data at different stages are often isolated, lacking effective multi-dimensional correlation analysis. In addition, under the traditional quality management model of inspection and testing business, quality problems are usually discovered and handled only after the problems have appeared, and the quality inspection cycle is long, which makes some quality problems fail to be identified early in the construction process, resulting in a significant increase in the cost of later repairs, and may even affect the progress and safety of the entire project.

[0003] In summary, the existing project quality management often has lagging project quality supervision methods, lacks comprehensive, multi-dimensional data retrieval statistics and chain reaction analysis and real-time early warning mechanisms during the construction process, and is unable to timely and accurately identify potential quality risks and conduct predictive early warnings, resulting in the discovery and control of quality problems lagging behind the construction process, causing irreversible technical problems in risk management. Summary of the invention

[0004] This application provides a project quality traceability and early warning method based on data analysis, which is used to solve the problems existing in the existing project quality management, such as the lag in project quality supervision methods, the lack of comprehensive, multi-dimensional data retrieval statistics and chain reaction analysis and real-time early warning mechanism during the construction process, the inability to timely and accurately identify potential quality risks and conduct predictive early warning, resulting in the discovery and control of quality problems lagging behind the construction process, causing irreversible technical problems in risk management and control.

[0005] The engineering quality tracing and early warning method based on data analysis provided in this application includes: A target engineering quality data set is obtained, wherein the target engineering quality data set refers to a data set obtained by collecting multi-dimensional quality information of a target engineering stage of a predetermined engineering; the target engineering quality data set is traversed and matched according to the target key factors corresponding to the target engineering stage to obtain target key feature parameters; the predetermined engineering quality data set is replaced and reorganized with the target key feature parameters to obtain a target real-time data set; the target real-time data set is used as input information of an engineering quality prediction center, and output information of the engineering quality prediction center is obtained, wherein the output information includes the target real-time prediction quality of the target engineering stage; if the target real-time prediction quality does not meet a predetermined engineering quality threshold, a primary warning signal is issued; a completed engineering stage is formed based on the primary warning signal, wherein the completed engineering stage includes all engineering stages before the target engineering stage and the target engineering stage; a predetermined traceability warning mechanism is called to trace and analyze the completed engineering stage to obtain a real-time quality abnormality category, and a fine-grained warning signal of the real-time quality abnormality category of the predetermined engineering is issued.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages: The engineering quality tracing and early warning method based on data analysis provided by the present application obtains a target engineering quality data set, wherein the target engineering quality data set refers to a data set obtained by collecting multi-dimensional quality information of a target engineering stage of a predetermined engineering; the target engineering quality data set is traversed and matched according to the target key factors corresponding to the target engineering stage to obtain target key feature parameters; the predetermined engineering quality data set is replaced and reorganized with the target key feature parameters to obtain a target real-time data set; the target real-time data set is used as input information of an engineering quality prediction middle station, and output information of the engineering quality prediction middle station is obtained, wherein the output information includes the target real-time predicted quality of the target engineering stage; if the target real-time predicted quality does not meet the predetermined engineering quality threshold, a primary early warning signal is issued; and a completed prediction system is established based on the primary early warning signal. Engineering stage, wherein the completed engineering stage includes all engineering stages before the target engineering stage and the target engineering stage; the predetermined traceability and early warning mechanism is called to trace back and analyze the completed engineering stage to obtain real-time quality abnormality categories, and a fine-grained early warning signal of the real-time quality abnormality category of the predetermined engineering is issued, which solves the problems existing in the existing engineering quality management, such as the lag in engineering quality supervision methods, the lack of comprehensive and multi-dimensional data retrieval statistics and chain reaction analysis and real-time early warning mechanism in the construction process, the inability to timely and accurately identify potential quality risks and make predictions and early warnings, resulting in the discovery and control of quality problems lagging behind the construction process, causing irreversible technical problems in risk management and control, and achieving the technical effects of real-time monitoring, prediction and accurate early warning of engineering quality, timely identification of potential quality risks, and improvement of the timeliness, accuracy and comprehensiveness of engineering quality management. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] Figure 1 A flow chart of the engineering quality tracing and early warning method based on data analysis is provided for this application.

[0008] Figure 2 A flow chart of obtaining target key characteristic parameters in an engineering quality traceability and early warning method based on data analysis is provided for this application. DETAILED DESCRIPTION

[0009] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0010] Examples, such as Figure 1 As shown, the present application provides a project quality tracing and early warning method based on data analysis, the method comprising: A target engineering quality data set is obtained, wherein the target engineering quality data set refers to a data set obtained by collecting multi-dimensional quality information of a target engineering stage of a predetermined engineering.

[0011] The target engineering quality data set is traversed and matched according to the target key factors corresponding to the target engineering stage to obtain target key feature parameters.

[0012] Specifically, the scheduled projects include environmental monitoring, construction projects, agricultural materials projects, water conservancy projects, power projects, environmental projects, communication projects, fire protection projects, etc. The target project quality data set refers to a comprehensive data set formed by collecting various types of data related to the target project stage, including design data, material data, construction data, etc. First, determine the target project stage. For example, in a construction project, the target project stage may be the "foundation construction stage" or the "main structure construction stage"; while in a power project, it may be the "equipment installation stage" or the "electrical commissioning stage". Each project stage contains specific quality information, which comes from multiple dimensions such as construction process, design standards, material testing, and environmental monitoring. In other words, by dynamically collecting data related to the project quality of the scheduled project at each stage of the project, a full data foundation is provided for subsequent project quality management, and the accuracy of quality management is provided. Then, quality data is collected from multiple angles and dimensions. Therefore, the target project quality data set includes not only real-time data during the construction process, but also data from different dimensions such as material properties, construction control, design specifications, and environmental conditions.

[0013] Next, the target engineering quality data set is traversed and matched according to the target key factors corresponding to the target engineering stage to obtain the target key feature parameters. The core task of this step is to screen out the key factors closely related to the engineering quality from the target engineering quality data set. The target key factors refer to the main influencing factors that determine the quality in a certain target engineering stage, such as construction technology, material type, environmental factors, etc. The key factors affecting quality may be different in different engineering stages. For example, the key factors that may affect the quality in the construction stage of a building project include "construction process control", "material quality standards" and "construction personnel operation specifications"; while the key factors in the equipment installation stage of a power project may be "equipment installation accuracy", "temperature and humidity control of the installation environment" and "standardization of electrical wiring". By traversing and matching the target engineering quality data set, the key feature parameters related to the quality of the target engineering stage can be automatically identified. The target key feature parameters are parameters with significant influence selected from the target engineering quality data set, such as concrete strength, material humidity, temperature, etc. For example, the key factors that may need to be matched in the foundation construction stage of a building project include "foundation soil type" and "foundation bearing capacity". Through this matching, characteristic parameters that have a significant impact on quality can be accurately extracted, providing basic data for subsequent analysis and prediction.

[0014] Through the above steps, the target project quality data set is acquired and key feature parameters are extracted. Through in-depth analysis of the target project stage, the key factors affecting quality can be accurately extracted, and detailed basic data is provided for subsequent quality traceability and early warning analysis. The automation and precision of this process effectively improves the sophistication and real-time nature of quality management, laying a solid foundation for the prediction and control of project quality risks.

[0015] The predetermined engineering quality data set is replaced and reorganized with the target key characteristic parameters to obtain a target real-time data set.

[0016] The target real-time data set is used as input information of the engineering quality prediction center, and output information of the engineering quality prediction center is obtained, wherein the output information includes the target real-time prediction quality of the target engineering stage.

[0017] Furthermore, the target key characteristic parameters extracted from the target engineering quality data set are applied to the predetermined engineering quality data set to generate a real-time, dynamically updated data set, thereby providing more accurate data support for subsequent quality prediction. The target key characteristic parameters, i.e. the extracted parameters that have a significant impact on the engineering quality, will directly replace the relevant parameters in the original quality data set. This replacement process is based on the real-time data of the current engineering stage. For example, if the target engineering stage is "mid-construction", the corresponding part of the original data set is updated according to the real-time data of the construction site (such as temperature, humidity, construction method, etc.). This process ensures that the data set used can reflect the actual situation of the current engineering stage and ensure the timeliness and accuracy of the data. Among them, the predetermined engineering quality data set refers to a complete data set collected before the start of the project according to design standards, material requirements, etc., which contains engineering quality data of each stage; the target real-time data set is the latest data set obtained by replacing and reorganizing the real-time data of the target engineering stage, reflecting the specific situation of the current construction or inspection stage.

[0018] After the target real-time data set is generated, this data set will be passed as input information to the engineering quality prediction middle platform, which is an integrated data analysis platform that can receive input from multiple data sources and perform centralized processing and analysis. In this embodiment, the target real-time data set is used as the input of the prediction model, providing real-time information from the engineering site. Through this input, the engineering quality prediction middle platform can perform data fusion, analysis and calculation to obtain the quality prediction results for the target engineering stage. The engineering quality prediction middle platform outputs the target real-time prediction quality based on the input target real-time data set after analysis and processing. This output information contains the quality status assessment of the target engineering stage, which is the result of predicting whether the quality of the current stage of the project meets the predetermined standards. For example, in the "foundation construction stage" of a construction project, the output information may include "foundation bearing capacity prediction value" and "soil moisture impact prediction", etc., to comprehensively evaluate the quality risk of the target stage. If the predicted value is lower than the predetermined quality standard, an early warning signal will be issued to the engineering management personnel. That is, the target real-time prediction quality refers to the predicted value of the quality status of the current target engineering stage obtained by analyzing the engineering quality prediction middle platform. This predicted value can reflect whether the quality of the engineering stage meets the predetermined design standards or safety requirements. The output information, i.e. the predicted results, includes quality data, risk assessment, possible quality issues, etc. After generating the target real-time predicted quality, the information is used for subsequent quality control and management decisions. Through real-time prediction results, project managers can timely understand the quality risks at the current stage and take effective quality control measures. If the predicted quality fails to meet the predetermined standards, managers can adjust the construction method, optimize the use of materials, or strengthen quality inspections to ensure that the project quality meets the design requirements.

[0019] By combining the target key characteristic parameters with the predetermined engineering quality data set to generate the target real-time data set, and passing this data set as input information to the engineering quality prediction center, the center generates the target real-time predicted quality after data analysis and processing. This output information provides an important basis for real-time monitoring, risk prediction and control of engineering quality. Through this real-time data update and quality prediction, the quality management efficiency of engineering projects can be greatly improved, potential quality problems can be identified in advance, and preventive and corrective measures can be taken in time to ensure the stability and safety of engineering quality.

[0020] If the target real-time prediction quality does not meet the predetermined engineering quality threshold, a primary warning signal is issued.

[0021] A completed engineering stage is established based on the primary warning signal, wherein the completed engineering stage includes all engineering stages before the target engineering stage and the target engineering stage.

[0022] Optionally, the generated target real-time prediction quality is evaluated and compared with a predetermined engineering quality threshold. The target real-time prediction quality refers to the quality prediction value of the target engineering stage generated by the engineering quality prediction middle platform. These prediction values ​​reflect the gap between the actual quality of the project and the design standard or safety requirement. If the target real-time prediction quality fails to meet the predetermined quality standard or threshold, it will be considered that there is a potential quality problem in the engineering stage. If it is found through comparison that the target real-time prediction quality fails to meet the predetermined engineering quality threshold, a primary warning signal is automatically issued. The warning signal is a preliminary warning, indicating that there is a potential quality problem in the current target engineering stage and may affect the overall engineering progress or safety. The warning signal can be notified to the engineering management personnel in a variety of ways according to the set rules, such as email, SMS notification or system alarm. The issuance of the primary warning signal is intended to attract the attention of engineering management personnel and take timely measures for quality control. Among them, the engineering quality threshold refers to a pre-set quality standard or safety limit, which is used to judge whether the engineering quality is qualified. An exemplary example is to collect historical engineering quality management records of similar projects of the scheduled project based on big data and build an engineering quality data lake, and combine the expert group's evaluation and analysis to determine the quality requirement threshold for the scheduled project to meet application requirements, and then determine the quality threshold of the scheduled project.

[0023] Next, we start to build the completed engineering phase based on the primary warning signal. The completed engineering phase refers to the set of all completed engineering phases with quality data, including the target engineering phase. Specifically, the completed engineering phase includes not only the current engineering phase (target engineering phase), but also all engineering phases that have been completed before the target engineering phase. The data of these engineering phases are combined to form a complete quality traceability framework. Through this step, the quality status of the entire project can be comprehensively evaluated, including quality problems and potential risks in the completed phase. Through traceability analysis, potential quality problems in engineering projects can be tracked and identified, helping managers find the root cause of the problem.

[0024] The predetermined traceability early warning mechanism is called up to perform traceability analysis on the completed engineering stage to obtain real-time quality abnormality categories, and a fine-grained early warning signal of the real-time quality abnormality category of the predetermined engineering is issued.

[0025] Specifically, the data of the completed engineering stage is traced and analyzed by calling the scheduled traceability early warning mechanism. The scheduled traceability early warning mechanism is a method to analyze and identify quality problems based on the historical data of the project and the quality event database. The function of this mechanism is to trace back and locate possible quality anomalies based on the data and quality records of the completed stage. In this process, data mining technology and pattern recognition algorithms are used to conduct in-depth analysis of historical data to automatically identify quality anomaly categories in the completed engineering stage. In the traceability analysis process, multiple factors will be comprehensively considered, such as the construction environment, material quality, construction technology, personnel operating specifications, etc., to conduct a comprehensive assessment of quality anomalies. These abnormal categories may involve different quality problems, such as "construction deviation", "unqualified material strength" or "equipment installation errors".

[0026] After retrospective analysis, real-time quality anomaly categories are generated, that is, all quality anomaly problem categories existing in the current engineering stage and previous stages. These anomaly categories are obtained through in-depth analysis and mining of the data of the completed engineering stages, reflecting the potential risks that affect quality. For example, in construction projects, real-time quality anomaly categories may include "exterior wall cracks" and "uneven concrete pouring", while in power projects, they may include "electrical equipment connection errors" or "irregular cable laying". This step ensures that quality problems existing in different stages of the project can be discovered in a timely manner through retrospective analysis, avoiding potential problems from being ignored or accumulated in the later stages, thereby affecting the overall quality and progress of the project.

[0027] Once the real-time quality anomaly categories are identified, fine-grained early warning signals will be issued based on these anomaly categories. Fine-grained early warning signals are detailed warnings issued for specific quality anomaly categories, aiming to provide accurate and specific quality problem warnings to project managers. These early warning signals not only indicate the category of quality anomaly, but also the specific cause of the anomaly, the possible scope of risk impact, and the improvement measures to be taken. For example, if "exterior wall cracks" are identified as quality anomaly categories, fine-grained early warning signals may include detailed information such as the specific location of the cracks, the width of the cracks, and the possible impact of the cracks on the building structure. In this way, managers can quickly locate the problem through this signal and take effective measures, such as strengthening the repair of the cracks and further investigating the causes of the cracks. The application of fine-grained early warning signals not only improves the response speed of project quality management, but also enhances the pertinence and effectiveness of early warning measures, thereby providing a more scientific quality control method for engineering projects.

[0028] In addition, in a specific embodiment, the associated dimension table can be processed with dimension reduction based on the real-time change of data and the records of a core fine-grained fact table as the main line. The fact table stores the quality quantitative data in the construction process, while the dimension table stores the contextual description data of the data in the fact table. By real-time monitoring of the data changes in the fact table, the data changes can be responded to in time, and the dimension table associated with the core fact table can be processed with dimension reduction. Dimension reduction refers to simplifying the structure of the dimension table, reducing the number of fields in the table or adjusting the storage format of the field to improve query efficiency and data processing speed. The associated fact table is processed with association resolution, and a data quality problem table is formed for data with quality problems, which is convenient for subsequent tracking and closed-loop processing. Association resolution refers to identifying and integrating repeated or redundant data records in multiple fact tables to maintain data consistency and accuracy. For the quality problem data that still exists after the above processing, a special data quality problem table is formed, which not only records the specific information of the data problem, but also includes the source of the problem data, the type of problem and the scope of influence, etc., which is convenient for subsequent problem tracking and closed-loop processing. The above steps strengthen the monitoring and management of quality issues, help to quickly locate and solve data problems, and ensure the accuracy and reliability of data.

[0029] By calling the predetermined traceability early warning mechanism to conduct traceability analysis on the quality data of the completed engineering stage, it is possible to identify real-time quality abnormality categories, and issue fine-grained early warning signals based on these abnormal categories, providing managers with accurate quality problem warnings, and indicating possible risks and countermeasures to be taken. Through this process, potential quality problems in the project can be identified and resolved in advance, improving the accuracy, timeliness and reliability of quality management, and providing a strong guarantee for the smooth completion of the project.

[0030] Furthermore, if Figure 2 As shown, the target engineering quality data set is traversed and matched according to the target key factors corresponding to the target engineering stage to obtain target key feature parameters, including: A project quality database of similar projects to the predetermined project is obtained, wherein the project quality database includes a plurality of project quality events.

[0031] Any engineering quality event corresponding to the first category among the multiple engineering quality events is extracted, and the arbitrary engineering quality event includes any data set.

[0032] A correlation analysis is performed on the arbitrary data set and the first category to obtain an arbitrary significant factor set, and a key factor set is constructed based on the arbitrary significant factor set.

[0033] The target key factors of the target engineering stage are traversed in the key factor set.

[0034] Exemplarily, a historical background of quality data is provided for the target project by obtaining the engineering quality database of similar projects of the predetermined project. The engineering quality database contains multiple engineering quality events, which are data records related to quality problems that occurred in past engineering projects. These events can include various quality defects, such as material problems, improper construction processes, environmental factors, etc. By analyzing these historical engineering quality events, common quality problems in similar engineering types can be understood, thereby providing a comprehensive quality reference framework for the target project. Specifically, the engineering quality database of similar projects refers to a collection of quality data of multiple engineering projects similar to the current target project. By analyzing these data, quality problems and solutions that occur in similar projects can be identified; and engineering quality events refer to specific quality problems or events that occur during the engineering process. Each quality event contains relevant quality data and impact descriptions, such as the type of quality defects, causes of occurrence, and solutions.

[0035] After obtaining data from a similar engineering quality database, any engineering quality event corresponding to the first category in the multiple engineering quality events is extracted. The first category generally refers to a representative category in the classification of quality problems. For example, in a construction project, the first category may be "building structure" or "construction technology", while in a power project it may be "equipment installation" or "electrical connection". Each quality event belongs to a specific category according to its classification, and then any quality event related to the target engineering quality problem is selected for analysis. Next, a correlation analysis is performed on any extracted data set and the first category to identify factors with significant influence in engineering quality events. Correlation analysis refers to analyzing the relationship between two or more data sets by statistical methods to determine which factors frequently appear in historical quality events and have a greater impact on quality. Through this analysis, any significant factor set can be screened out, that is, factors that appear repeatedly in different engineering quality events and have a greater impact on quality. These factors are usually the root causes of quality problems and have strong predictability. These factors may include construction methods, material quality, environmental control, etc. The extraction of a significant factor set helps to determine the quality control points that need special attention in the current engineering stage.

[0036] Subsequently, based on the above analysis results, any significant factor sets extracted are integrated to form a key factor set. The key factor set refers to a collection of all factors that can have a significant impact on the quality of the target engineering stage, which are extracted through multiple correlation analyses during the analysis process. The key factor set may include factors in various aspects such as design, construction, and materials, and provide important basic data for the next step of quality prediction and early warning. Finally, the key factor set will be traversed to screen out the target key factors related to the current target engineering stage. These target key factors are the most important factors in the quality analysis of the target engineering stage, and usually directly affect the quality level of this stage. By traversing the key factor set, the quality control points that need to be focused on in the current stage can be accurately identified, such as "construction process quality" in construction projects, or "equipment installation accuracy" in power projects.

[0037] By obtaining the engineering quality database of similar projects, we can refer to the quality events of multiple historical engineering projects, extract the significant factor set based on correlation analysis, and finally construct the key factor set; by traversing the key factor set, we can accurately identify the target key factors of the target engineering stage, providing an accurate data basis for subsequent quality analysis and early warning. This process ensures that the quality analysis of the target project is not only based on real-time data, but also can use the lessons learned from historical data to further improve the accuracy and practicality of quality prediction.

[0038] Furthermore, the predetermined engineering quality data set is replaced and reorganized with the target key characteristic parameters to obtain the target real-time data set, including: Obtain an engineering design plan for the predetermined project.

[0039] A first design plan corresponding to the first design project in the engineering design solution is extracted.

[0040] The first design upper limit in the first design plan is extracted to form the predetermined engineering quality data set.

[0041] Optionally, the engineering design plan of the predetermined project is obtained, which is the basis for constructing the target real-time data set. The engineering design plan refers to a detailed design document compiled by the engineering design unit according to the project requirements before the project starts, which contains key contents such as the overall structural design, construction plan, material specifications, and engineering standards of the project. By obtaining the engineering design plan, we can fully understand the design goals, standards, and specific requirements of the project. These contents provide a key basis for the matching and updating of subsequent quality data. This step ensures that the target real-time data set can be replaced and updated according to the requirements of the design plan, ensuring that the actual implementation of the project is in line with the original design intention.

[0042] Next, extract the first design plan corresponding to the first design project in the engineering design plan. The first design project refers to the first key design task or project in the engineering design plan. It is usually the core part of the entire project and has an important impact on subsequent projects. The design plan refers to the detailed implementation plan proposed for each design project in the engineering design plan. The first design plan refers to the specific construction plan proposed for the first design project, which includes detailed implementation steps, standards and requirements. In actual projects, design plans are often divided into multiple design projects, and each design project corresponds to a specific construction plan or technical standard. The first design project can be the largest project in the project. For example, in a construction project, it may be "infrastructure construction", and the first design plan is a detailed implementation plan for the project, which clarifies the various construction standards, quality requirements, material selection and other contents of the design project.

[0043] Extract the first design upper limit in the first design plan. The first design upper limit refers to the upper limit value of the quality standard, technical indicator or construction requirement set in the first design plan. For example, in a construction project, the first design upper limit may involve the upper limit requirement of "foundation bearing capacity" or the maximum standard of "concrete compressive strength". By extracting the upper limit requirements in the first design plan, it is possible to clarify the maximum standard or tolerance limit value that the project quality should reach at this stage. This upper limit provides an important reference for project quality control to ensure that the quality does not exceed the reasonable range of design requirements.

[0044] Finally, based on the extracted first design upper limit, the existing engineering quality data is replaced and reorganized to generate a predetermined engineering quality data set, which includes not only static data such as design standards and construction specifications, but also real-time monitoring data, such as material strength, construction environment, process execution and other dynamic data. By combining the first design upper limit with the actual quality data, a predetermined engineering quality data set is formed. The predetermined engineering quality data set refers to the quality data set constructed according to the design plan and construction data at the target engineering stage, including static data such as design standards and construction quality requirements and real-time construction monitoring data, reflecting the quality control standards and goals of the current project, and serving as the basis for subsequent quality prediction and early warning.

[0045] By obtaining the engineering design plan, extracting the first design project and its plan, and extracting the first design upper limit, the predetermined engineering quality data set is gradually constructed; by replacing and reorganizing the quality data, a real-time data set reflecting the quality requirements and construction status of the current engineering stage can be dynamically generated, providing an accurate data basis for subsequent quality prediction and early warning analysis. This process ensures that the quality management of the engineering project is based on comprehensive, real-time, and standardized data, and improves the accuracy and timeliness of engineering quality control.

[0046] Further, the target real-time data set is used as input information of the engineering quality prediction middle station, and the output information of the engineering quality prediction middle station is obtained, wherein the output information includes the target real-time prediction quality of the target engineering stage, including: The target real-time data set and the predetermined engineering quality data set are vectorized in turn through the vectorization module of the engineering quality prediction center to obtain the target real-time vector and the predetermined vector respectively.

[0047] The similarity between the target real-time vector and the predetermined vector is calculated using the cosine similarity principle to obtain the target real-time quality similarity.

[0048] The target real-time predicted quality is calculated by taking the predetermined quality corresponding to the predetermined engineering quality data set as 1 and the target real-time quality similarity as a weight.

[0049] Specifically, the target real-time data set and the predetermined engineering quality data set are vectorized in turn through the vectorization module of the engineering quality prediction platform. The vectorization module is a core functional module in the engineering quality prediction platform, which is responsible for converting the original data into a numerical vector for subsequent calculation and analysis. Vectorization is the process of converting the multidimensional information in the data set into a numerical vector for subsequent mathematical analysis. In this step, the real-time data of the target project and the preset design standard data are vectorized to obtain the target real-time vector and the predetermined vector respectively. Among them, the target real-time vector refers to the real-time data (such as temperature, humidity, construction parameters, etc.) of the current target project stage converted into a multidimensional numerical vector, reflecting the actual quality status of the current project stage; and the predetermined vector refers to the data in the predetermined engineering quality data set (such as design requirements, construction specifications, material standards, etc.) converted into a standardized numerical vector, representing the quality requirements of the design stage. This step ensures that the data can be represented in a unified numerical form for subsequent mathematical calculations and quality assessments.

[0050] After vectorization is completed, the cosine similarity principle is used to calculate the similarity between the target real-time vector and the predetermined vector. Cosine similarity is a commonly used algorithm to measure the similarity between two vectors. The similarity between them is evaluated by calculating the cosine value of the angle between the two vectors. Specifically, the value of cosine similarity is between -1 and 1. The closer the value is to 1, the more similar the two vectors are. Conversely, the value close to -1 indicates that the similarity between the two is lower. In this step, the target real-time vector represents the quality data of the current construction stage, and the predetermined vector represents the quality standard of the design stage. The target real-time quality similarity, that is, the degree of match between the current engineering stage and the design requirements, can be obtained through cosine similarity calculation. If the similarity is high, it means that the actual quality is close to the design requirements. Otherwise, it means that there is a quality deviation.

[0051] Finally, the target real-time predicted quality is calculated based on the calculated target real-time quality similarity and the predetermined quality corresponding to the predetermined engineering quality data set is used as the weight. Specifically, the predetermined quality refers to the quality standard specified in the engineering design stage, usually expressed in numerical form, as a benchmark for evaluating the target engineering quality. The predetermined quality (the standard quality value defined in the predetermined engineering quality data set, usually 1) is multiplied by the target real-time quality similarity to obtain the target real-time predicted quality, which can reflect the quality status of the target engineering stage and serve as the basis for subsequent decision-making. For example, assuming that the predetermined quality is 1, which represents the ideal standard of quality in the engineering design stage, and the target real-time quality similarity calculated by cosine similarity is 0.85, indicating that the actual engineering quality matches the design requirements by 85%. The target real-time predicted quality is 0.85 obtained by multiplication, indicating that the current quality status is good, but there is still a certain deviation. Finally, the calculated target real-time predicted quality provides real-time quality assessment results for engineering managers. This predicted quality value can help managers to understand whether the current engineering quality meets the design standards in a timely manner. If the predicted value is lower than the predetermined quality, an early warning can be triggered and corrective measures can be taken.

[0052] The above steps ensure the efficiency and accuracy of project quality prediction, provide a scientific basis for project quality management, and support the implementation of real-time monitoring and early warning mechanisms. Through this method, quality deviations can be discovered in a timely manner to ensure that project quality is always within a controllable range.

[0053] Furthermore, if the target real-time predicted quality does not meet the predetermined engineering quality threshold, a primary warning signal is issued, including: A first design lower limit in the first design plan is extracted.

[0054] A second predetermined engineering quality data set is established based on the first design lower limit.

[0055] A vectorized comparative analysis is performed on the second predetermined engineering quality data set and the predetermined engineering quality data set to obtain a second predetermined quality.

[0056] The predetermined engineering quality threshold is formed according to the second predetermined quality and the predetermined quality.

[0057] Further, the first design lower limit in the first design plan is extracted. In the design plan of each project, the design lower limit refers to the minimum quality requirement allowed under the premise of meeting the project quality standard. Specifically, the first design plan is set according to the project requirements and project characteristics, and the first design lower limit usually represents the minimum tolerance for quality requirements in the design stage. For example, for a construction project, the first design lower limit may include the minimum concrete strength, material thickness or minimum requirements for construction technology. Extracting this design lower limit provides a lower limit standard for subsequent quality analysis as a reference for quality assessment and early warning. After extracting the first design lower limit, a second predetermined project quality data set will be formed based on the lower limit. This data set contains all project data that meet the minimum quality standard, including material specifications, construction parameters, construction technology, etc. This data set is generated to evaluate the actual performance of the target project quality under the minimum standard, especially when the target real-time predicted quality does not reach the predetermined threshold, it is necessary to compare the minimum quality standard with the actual situation. By comparing the minimum quality standard with the actual data, a more conservative evaluation basis for project quality is provided to ensure that all possible risks of project quality control can be covered.

[0058] Next, a vectorized comparative analysis is performed on the second predetermined engineering quality data set and the predetermined engineering quality data set. Through this analysis, all relevant parameters in the two data sets are converted into vector form and then compared. The predetermined engineering quality data set is based on the ideal quality standard set in the engineering design phase, while the second predetermined engineering quality data set is based on the minimum quality requirements. In this step, the second predetermined quality is obtained by calculating the difference between the two vectors. This quality value represents the expected engineering quality level based on the minimum standard. The purpose of this step is to compare the engineering quality under the minimum standard with the quality level under the ideal standard, and to evaluate whether there are deviations in the current engineering quality and the impact of these deviations.

[0059] Finally, the predetermined engineering quality threshold is formed according to the result of the second predetermined quality and the predetermined quality. Specifically, the predetermined quality is the ideal quality standard set in the design stage, which is usually 1, and the second predetermined quality is the quality value obtained based on the minimum standard. By combining these two quality values, a new quality threshold will be generated to determine whether the actual quality of the target engineering stage meets the acceptable range. If the target real-time predicted quality is lower than the predetermined engineering quality threshold, a primary warning signal will be issued to remind engineering management personnel to take appropriate quality control measures. This step ensures that the predetermined quality threshold can accurately reflect the quality requirements of the engineering project based on the combination of the minimum design standard and the ideal design standard, and trigger the corresponding early warning mechanism. Through this method, accurate control of engineering quality and risk warning can be achieved, thereby improving the accuracy and timeliness of quality management.

[0060] Furthermore, the predetermined traceability warning mechanism is called to perform traceability analysis on the completed engineering stage to obtain real-time quality abnormality categories, and a fine-grained warning signal of the real-time quality abnormality category of the predetermined engineering is issued, including: A descending list of categories is obtained according to the predetermined traceability warning mechanism.

[0061] The top category in the descending list of categories is extracted, and the top quality prediction channel corresponding to the top category is activated.

[0062] The top quality prediction channel is used to analyze the completed project quality data set of the completed project stage to obtain the top judgment result of the top category.

[0063] When the top analysis result shows that there is a quality abnormality in the top category, the fine-grained early warning signal of the top category is issued.

[0064] Optionally, a descending list of categories is generated by retrospectively analyzing the quality data of the completed engineering phase according to a predetermined traceability and early warning mechanism. The predetermined traceability and early warning mechanism is an analysis model based on historical quality events and existing engineering data, which can extract and filter out important quality-related categories from the completed engineering phase. The descending list of categories is sorted according to the frequency of occurrence or degree of impact of each category in engineering quality issues, where the top category represents the part with the greatest impact on quality. For example, in a construction project, categories may include "foundation problems", "structural problems", "construction process problems", etc., and "foundation problems" may be at the top of the descending list because it usually has a greater impact on building safety.

[0065] After obtaining the descending list of categories, the top category in the descending list of categories will be extracted. The top category represents the most important quality problem category in the current quality analysis and is the part that needs to be focused on in the current engineering stage. Activating the top quality prediction channel means that a special analysis module will be mobilized to conduct in-depth research and judgment on the quality data of this category, predict its possible risks and conduct detailed analysis. For example, if the top category is "foundation problem", the system will enable the quality prediction channel related to "foundation problem" and conduct a detailed analysis of its related data. Among them, the top category refers to the category that ranks high in the descending list of categories, usually representing aspects that have a significant impact on quality or frequently occurring quality problems; and the top quality prediction channel is a quality prediction module specially set for the top category, which is used to conduct in-depth analysis and prediction of the quality data under this category. By activating the top quality prediction channel, we can focus on the most critical quality issues and provide an accurate basis for subsequent quality prediction and early warning.

[0066] Then, the quality data set of the completed project in the completed project phase is analyzed through the top quality prediction channel. The purpose of this analysis is to evaluate the quality performance of this phase and identify potential quality anomalies by reviewing historical data. For example, by analyzing the quality data of the "foundation construction" phase, it may be evaluated whether there are non-compliant construction processes, material problems or environmental impacts. In this process, real-time and historical quality data are combined to evaluate the factors that may cause quality problems. Finally, when the analysis conducted through the top quality prediction channel shows that there are quality anomalies in the top category, a fine-grained early warning signal will be issued. Fine-grained early warning signals are detailed early warning information for specific quality anomalies, which can clearly indicate the specific details such as the cause, location, and scope of impact of the anomaly. For example, if the "foundation problem" is judged as a quality anomaly, the fine-grained early warning signal may include information such as "foundation settlement exceeds the allowable range" or "foundation bearing capacity is insufficient", which helps managers quickly locate the problem and take repair measures. The issuance of fine-grained early warning signals not only helps project managers accurately identify quality problems, but also provides detailed information to facilitate more effective corrective measures.

[0067] Furthermore, obtaining a descending list of categories according to the predetermined traceability warning mechanism includes: The multiple engineering quality events are clustered into categories to obtain event clustering results.

[0068] A first event cluster of the first category in the event clustering result is extracted, and the first event cluster is analyzed to obtain a first total index of the first category.

[0069] According to the predetermined traceability warning mechanism, the category set is sorted in descending order based on the first total index to obtain the category descending list.

[0070] The step of analyzing the first event cluster to obtain a first total index of the first category includes: A construction party event set of the project construction party in the first event cluster is extracted.

[0071] A user event set of the engineering user in the first event cluster is extracted.

[0072] The first total index is obtained by performing a weighted calculation on the builder event frequency corresponding to the builder event set and the user event frequency corresponding to the user event set.

[0073] In a specific embodiment, multiple engineering quality events are clustered into categories to obtain event clustering results. In this process, similar events are classified into one category based on the characteristics of the quality events. For example, all quality events involving "unqualified construction technology" will be clustered into one category, while quality events involving "material problems" will be classified into another category. Cluster analysis is based on multidimensional data, such as event type, frequency of occurrence, problem type, etc., to ensure that similar quality events can be classified together for subsequent analysis. The clustering results provide basic data for subsequent analysis, ensuring that quality events can be effectively managed by category.

[0074] Next, the first category in the event clustering result is extracted, that is, the quality problem category that ranks high in the clustering result, which usually represents the most serious or most frequently occurring quality problem. After extracting the first category, the first event cluster in the category is further extracted, that is, the specific event set belonging to the category. Each event cluster will be analyzed in detail to evaluate its impact on the overall project quality. Through this analysis, the specific quality problems under each category can be clarified, thereby providing an accurate basis for subsequent quality evaluation. Then, the first event cluster is analyzed to obtain the first total index of the first category, which is calculated by comprehensively analyzing various quality events occurring in the first event cluster to reflect the severity of the quality problems in the category. For example, the first event cluster may include multiple quality problems, such as foundation settlement, unstable structure, improper construction technology, etc., and a weighted calculation will be performed according to the frequency and impact of each quality problem to obtain a first total index reflecting the severity of the overall problem; or, considering the consistency between the number of engineering quality problems found by the internal quality inspection of the builder and the engineering quality problems found by the owner, whether a project has quality problems needs to be evaluated by multiple parties and different standards. Taking residential building construction as an example, assuming that the first event cluster includes the "foundation settlement" problem found by the internal quality inspection of the builder and the "wall crack" problem reported by the owner, by analyzing the frequency of occurrence of these two events and the severity of each problem, a weighted first total index will be calculated. This index measures the impact of current quality problems on the overall quality of the project.

[0075] After obtaining the first total index of the first category, the category set is sorted in descending order based on the index, thereby obtaining a category descending list. Among them, the category set refers to the category set of all quality problems, and each category contains a group of quality events with similar characteristics. The category descending list refers to a quality problem category ranking from the most serious to the least serious by sorting the total index of each category. This descending list clarifies the degree of influence of each category on the overall project quality according to the total index of each category. The higher the category ranking, the greater its impact on the project quality and needs to be handled first. For example, if the "foundation problem" category ranks first and the "construction process problem" ranks second, the system will prioritize the foundation problem based on this ranking. The generation of the category descending list provides a clear quality priority for project management, allowing managers to take targeted measures. Subsequently, according to the category descending list, the most critical quality problem category can be identified, and a fine-grained early warning signal can be issued based on this analysis. The fine-grained early warning signal is a specific early warning information issued for a specific quality problem, which can clearly indicate the specific type, location and possible impact range of the quality anomaly. For example, if "foundation problems" are identified as the most serious quality problem category, a fine-grained early warning signal including "foundation settlement exceeds the predetermined limit" will be issued to remind management personnel to take timely measures to intervene. Through this process, project quality management can control risks more accurately and timely, thereby ensuring the stability and safety of project quality.

[0076] Further, extracting the top category in the descending list of categories and activating the top quality prediction channel corresponding to the top category includes: Extract any category in the category descending list, and match any event cluster corresponding to the any category in the event clustering result.

[0077] Based on the accident tree principle, accident factor analysis is performed on the arbitrary categories and the arbitrary event clusters, and an arbitrary cut set set is formed.

[0078] An arbitrary training data set is constructed based on the arbitrary cut set set, and supervised training is performed on the arbitrary training data set based on the support vector machine principle to obtain an arbitrary quality prediction channel.

[0079] The top quality prediction channel corresponding to the top category is matched in the arbitrary quality prediction channel.

[0080] Specifically, extract any category in the descending list of categories, that is, select any category from the quality problem categories in descending order for further analysis. Any category may be one of the entire engineering quality problem classification, indicating that there is a potential quality risk under this category. Subsequently, find any event cluster associated with the arbitrary category in the event clustering result. An event cluster is a collection of all events related to a certain quality problem type analyzed according to a clustering algorithm, which has similar characteristics and can help locate the cause of the quality problem. For example, if any category is "construction process problem", the corresponding event cluster may include quality problems such as "substandard construction quality", "construction deviation", and "irregular operation". Through this step, the most critical or frequent quality problems in the current project can be dealt with in a centralized manner, providing basic data for the next step of predictive analysis.

[0081] Next, based on the accident tree principle, accident factor analysis is performed on any category and any event cluster. The accident tree principle is a risk assessment method that helps identify the key factors that are most likely to cause quality problems by systematically analyzing and identifying the causes and paths of potential accidents. In this step, all quality events (such as "substandard construction quality") in any category (such as "construction process problems") are factor analyzed to identify the root cause of the problem. By analyzing and sorting out these factors, a cut set is formed, that is, a set of key factors that jointly or individually cause the occurrence of quality problems. Among them, the cut set refers to the minimum set of factors that cause accidents in the accident tree analysis, which is usually composed of multiple interrelated factors. Through this analysis, the key factors of quality problems can be identified, and the cause-effect relationship of the problem can be displayed through the accident tree model, providing a clear basis for subsequent prediction and processing.

[0082] After identifying the key accident factors, an arbitrary training data set will be formed based on an arbitrary set of cut sets. The training data set contains the specific data of the accident factors and is combined with historical engineering quality events as the input data for training the support vector machine (SVM). Support vector machine (SVM) is a supervised learning method used for classification and regression analysis. In this step, the SVM algorithm will be applied to the training data set to learn to extract quality problem prediction patterns from historical data. Through supervised training, an arbitrary quality prediction channel can be constructed, which can predict future engineering quality problems and provide relevant quality risk assessments. Finally, the arbitrary quality prediction channel is matched in the top quality prediction channel. The top quality prediction channel is a quality prediction channel specifically built for the top category (such as "construction process problems"), which uses the trained SVM model to accurately predict quality problems. By matching in any quality prediction channel, the quality prediction analysis corresponding to the top category can be integrated to provide specific prediction results for the quality problems in this category. For example, if "construction process problems" are identified as the top category, the quality risks that may occur under this category will be calculated in this channel to help managers identify potential problems in advance and take corresponding measures. Through this step, the most important quality issues can be accurately predicted and real-time quality control suggestions can be provided when managers need them.

[0083] Through the technical solutions of the above-mentioned embodiments, the engineering quality tracing and early warning method based on data analysis provided by the present application solves the problems existing in the existing engineering quality management, such as the lag in engineering quality supervision methods, the lack of comprehensive and multi-dimensional data retrieval statistics and chain reaction analysis and real-time early warning mechanism in the construction process, the inability to timely and accurately identify potential quality risks and conduct predictive early warning, resulting in the discovery and control of quality problems lagging behind the construction process, causing irreversible technical problems in risk management, and achieving the technical effects of real-time monitoring, prediction and accurate early warning of engineering quality, timely identification of potential quality risks, and improvement of the timeliness, accuracy and comprehensiveness of engineering quality management.

[0084] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. The processes depicted in the accompanying drawings do 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.

[0085] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0086] This specification and drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. The engineering quality tracing and early warning method based on data analysis is characterized by: include: Acquire a target engineering quality data set, wherein the target engineering quality data set refers to a data set obtained by collecting multi-dimensional quality information of a target engineering stage of a predetermined engineering; According to the target key factors corresponding to the target engineering stage, the target engineering quality data set is traversed and matched to obtain target key feature parameters; Replacing and reorganizing a predetermined engineering quality data set with the target key characteristic parameters to obtain a target real-time data set; Using the target real-time data set as input information of the engineering quality prediction middle station, and obtaining output information of the engineering quality prediction middle station, wherein the output information includes the target real-time prediction quality of the target engineering stage; If the target real-time predicted quality does not meet the predetermined engineering quality threshold, a primary warning signal is issued; Establishing a completed engineering stage based on the primary warning signal, wherein the completed engineering stage includes all engineering stages before the target engineering stage and the target engineering stage; The predetermined traceability early warning mechanism is called up to perform traceability analysis on the completed engineering stage to obtain real-time quality abnormality categories, and a fine-grained early warning signal of the real-time quality abnormality category of the predetermined engineering is issued.

2. The engineering quality tracing and early warning method based on data analysis according to claim 1 is characterized in that: The target engineering quality data set is traversed and matched according to the target key factors corresponding to the target engineering stage to obtain target key feature parameters, including: Acquire a project quality database of similar projects to the predetermined project, wherein the project quality database includes a plurality of project quality events; Extracting any engineering quality event corresponding to the first category from the plurality of engineering quality events, wherein the any engineering quality event includes any data set; Performing correlation analysis on the arbitrary data set and the first category to obtain any significant factor set, and forming a key factor set based on the arbitrary significant factor set; The target key factors of the target engineering stage are traversed in the key factor set.

3. The engineering quality tracing and early warning method based on data analysis according to claim 1 is characterized in that: The predetermined engineering quality data set is replaced and reorganized with the target key characteristic parameters to obtain the target real-time data set, including: Obtaining an engineering design plan for the predetermined project; Extracting a first design plan corresponding to a first design project in the engineering design plan; The first design upper limit in the first design plan is extracted to form the predetermined engineering quality data set.

4. The engineering quality tracing and early warning method based on data analysis according to claim 1 is characterized in that: The target real-time data set is used as input information of the engineering quality prediction middle station, and output information of the engineering quality prediction middle station is obtained, wherein the output information includes the target real-time prediction quality of the target engineering stage, including: The target real-time data set and the predetermined engineering quality data set are vectorized in sequence by the vectorization module of the engineering quality prediction center, and the target real-time vector and the predetermined vector are obtained respectively; Calculating the similarity between the target real-time vector and the predetermined vector using the cosine similarity principle to obtain the target real-time quality similarity; The target real-time predicted quality is calculated by taking the predetermined quality corresponding to the predetermined engineering quality data set as 1 and the target real-time quality similarity as a weight.

5. The engineering quality tracing and early warning method based on data analysis according to claim 3 is characterized in that: If the target real-time prediction quality does not meet the predetermined engineering quality threshold, a primary warning signal is issued, including: Extracting a first design lower limit in the first design plan; Establishing a second predetermined engineering quality data set based on the first design lower limit; Performing vectorized comparative analysis on the second predetermined engineering quality data set and the predetermined engineering quality data set to obtain a second predetermined quality; The predetermined engineering quality threshold is formed according to the second predetermined quality and the predetermined quality.

6. The engineering quality tracing and early warning method based on data analysis according to claim 2 is characterized in that: The predetermined traceability early warning mechanism is called to perform traceability analysis on the completed engineering stage to obtain real-time quality abnormality categories, and a fine-grained early warning signal of the real-time quality abnormality category of the predetermined engineering is issued, including: Obtain a descending list of categories according to the predetermined traceability warning mechanism; Extracting the top category in the descending list of categories, and activating the top quality prediction channel corresponding to the top category; The completed project quality data set of the completed project stage is analyzed and judged through the top quality prediction channel to obtain the top judgment result of the top category; When the top analysis result shows that there is a quality abnormality in the top category, the fine-grained early warning signal of the top category is issued.

7. The engineering quality tracing and early warning method based on data analysis according to claim 6 is characterized in that: Obtaining a descending list of categories according to the predetermined traceability warning mechanism includes: Clustering the multiple engineering quality events into categories to obtain event clustering results; Extracting a first event cluster of the first category from the event clustering result, and analyzing the first event cluster to obtain a first total index of the first category; According to the predetermined traceability warning mechanism, the category set is sorted in descending order based on the first total index to obtain the category descending list; The step of analyzing the first event cluster to obtain a first total index of the first category includes: Extracting a construction party event set of the engineering construction party in the first event cluster; Extracting a user event set of the engineering user in the first event cluster; The first total index is obtained by performing a weighted calculation on the builder event frequency corresponding to the builder event set and the user event frequency corresponding to the user event set.

8. The engineering quality tracing and early warning method based on data analysis according to claim 7 is characterized in that: Extracting the top category in the descending list of categories and activating the top quality prediction channel corresponding to the top category, including: Extract any category in the category descending list, and match any event cluster corresponding to the any category in the event clustering result; Based on the accident tree principle, an accident factor analysis is performed on the arbitrary categories and the arbitrary event clusters, and an arbitrary cut set set is formed; Based on the arbitrary set of cut sets, an arbitrary training data set is formed, and supervised training is performed on the arbitrary training data set based on the support vector machine principle to obtain an arbitrary quality prediction channel; The top quality prediction channel corresponding to the top category is matched in the arbitrary quality prediction channel.

Images