Construction quality information automatic evaluation method and system

Abstract

The present application relates to the technical field of building engineering quality information management, and specifically discloses a construction quality information automatic evaluation method and system. The present application collects multi-dimensional quality sensing data of to-be-measured points in the construction process by deploying multi-source sensing devices in real time, and performs time sequence alignment processing based on timestamps and point identifiers. Combined with an abnormality detection rule library, the present application automatically marks quality deviation points and abnormal event information, analyzes deviation propagation paths and influence values using a multi-layer weighted directed graph model, and generates process-level quality comprehensive deviation indexes and construction quality comprehensive evaluation information, thereby solving the technical problems of low data collection efficiency, information transmission lag, insufficient data integrity and accuracy, and scattered historical quality data that is difficult to trace and analyze under the traditional manual inspection and paper record mode, and improving the construction quality control level and the degree of engineering quality management refinement.

Description

Technical Field

[0001] This invention relates to the field of information management technology for construction engineering quality, and in particular to an automatic evaluation method and system for construction quality information. Background Technology

[0002] In the field of construction engineering, construction quality control is a core element in ensuring project safety and durability. Currently, the collection of quality information at construction sites mainly relies on a combination of manual inspections and paper records. Quality inspectors must carry testing instruments to measure point by point and manually fill out reports, which are then compiled offline to form a quality assessment report. This traditional model suffers from low data collection efficiency and severe information delays. On-site test data often takes hours or even days to be fed back to the project management platform, resulting in quality problems not being detected and addressed in a timely manner. More importantly, manual recording methods cannot guarantee the completeness and accuracy of data, leading to frequent omissions and misrecording. Furthermore, historical quality data is scattered across different departments, lacking a unified digital management platform, making it impossible to achieve dynamic traceability and systematic analysis of construction quality information, which seriously restricts the improvement of project quality management.

[0003] Therefore, there is an urgent need to provide a technical solution to address the above problems. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides an automatic assessment method and system for construction quality information.

[0005] Firstly, the present invention provides an automatic assessment method for construction quality information, the technical solution of which is as follows: Obtain the quality control items corresponding to the construction process, determine at least one test point and at least one quality indicator corresponding to each test point based on the quality control items, and establish the correspondence between the test point and the quality indicator. The system receives multi-dimensional quality sensing data collected in real time from multi-source sensing devices deployed at the test points. The multi-dimensional quality sensing data includes at least structural mechanical parameters, material state parameters, and environmental parameters. The system also records the timestamp and point identifier corresponding to the multi-dimensional quality sensing data. Based on the correspondence, the multidimensional quality sensing data is time-series aligned according to the timestamp and the location identifier to construct a quality time-series data sequence for each location to be measured. The system calls a preset anomaly detection rule library, which stores the mapping relationship between the quality indicators and threshold conditions. The system matches the quality time series data sequence with the threshold conditions of the corresponding quality indicators, marks the test points with data anomalies as quality deviation points, and records the point anomaly event information corresponding to the quality deviation points. The point anomaly event information includes the deviation quality indicator, the deviation value, and the time of anomaly occurrence. The quality index deviation of the quality deviation point is calculated based on the abnormal event information of the point, and the quality index deviation is input into a pre-established multi-layer weighted directed graph model. The multi-layer weighted directed graph model determines the propagation path of the influence of the quality index deviation on the quality index of at least one related process and the propagation influence value corresponding to the propagation path. The quality index deviation of all test points under the same construction process is summarized, along with the influence propagation path and propagation influence value of all quality deviation points, to generate a process-level comprehensive quality deviation index. Based on the process-level quality comprehensive deviation index, a preset evaluation level mapping table is queried to determine the corresponding quality level and handling suggestions. The quality level, the handling suggestions, and the deviation quality indicators, deviation values, anomaly occurrence time, and location identifiers in the location anomaly event information are combined to generate construction quality comprehensive evaluation information containing text descriptions and structured data fields.

[0006] Secondly, the present invention provides an automatic assessment system for construction quality information, the technical solution of which is as follows: The acquisition module is used to acquire the quality control items corresponding to the construction process, determine at least one test point and at least one quality indicator corresponding to each test point based on the quality control items, and establish the correspondence between the test point and the quality indicator. The receiving module is used to receive multi-dimensional quality sensing data collected in real time by multi-source sensing devices deployed at the test point. The multi-dimensional quality sensing data includes at least structural mechanical parameters, material state parameters and environmental parameters, and records the timestamp and point identifier corresponding to the multi-dimensional quality sensing data. The construction module is used to perform time-series alignment processing on the multidimensional quality sensing data according to the timestamp and the location identifier based on the correspondence, and construct a quality time-series data sequence for each of the locations to be measured; The matching module is used to call a preset anomaly detection rule library. The anomaly detection rule library stores the mapping relationship between the quality indicators and threshold conditions. It matches the quality time series data sequence with the threshold conditions of the corresponding quality indicators, marks the test points with data anomalies as quality deviation points, and records the point anomaly event information corresponding to the quality deviation points. The point anomaly event information includes the deviation quality indicator, the deviation value, and the time of anomaly occurrence. The generation module is used to calculate the quality index deviation of the quality deviation point based on the point abnormal event information, and input the quality index deviation into a pre-established multi-layer weighted directed graph model. The multi-layer weighted directed graph model determines the influence propagation path of the quality index deviation on the quality index of at least one related process and the propagation influence value corresponding to the influence propagation path. The summary module is used to summarize the quality index deviation of all the test points under the same construction process, the influence propagation path and propagation influence value corresponding to all the quality deviation points, and generate a process-level comprehensive quality deviation index. The evaluation module is used to query a preset evaluation level mapping table based on the comprehensive deviation index of the process-level quality, determine the corresponding quality level and handling suggestions, and combine the quality level, the handling suggestions and the deviation quality index, deviation value, abnormal occurrence time and location identifier in the location abnormal event information to generate comprehensive construction quality evaluation information containing text description and structured data fields.

[0007] The technical solution of this invention collects multi-dimensional quality perception data of the test points in the construction process in real time by deploying multi-source sensing devices, and performs time-series alignment processing based on timestamps and point identifiers. Combined with the anomaly detection rule base, it automatically marks the quality deviation points and abnormal event information, and uses a multi-layer weighted directed graph model to analyze the deviation propagation path and impact value. It summarizes and generates a process-level comprehensive quality deviation index and comprehensive construction quality assessment information, which solves the technical problems of low data collection efficiency, delayed information transmission, insufficient data integrity and accuracy, and difficulty in tracing and analyzing historical quality data under the traditional manual inspection and paper record mode. It improves the level of construction quality control and the degree of refinement of engineering quality management.

[0008] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and in order to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0010] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 This is a flowchart illustrating an embodiment of an automatic construction quality information assessment method according to the present invention. Detailed Implementation

[0011] Exemplary embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention can be implemented in various forms and should not be limited to the embodiments set forth herein.

[0012] Figure 1 This diagram illustrates a flowchart of an embodiment of an automatic construction quality information assessment method provided by the present invention, executed by a control terminal. Figure 1 As shown, the automatic assessment method for construction quality information includes the following steps: S1. Obtain the quality control items corresponding to the construction process, determine at least one test point and at least one quality indicator corresponding to each test point based on the quality control items, and establish the correspondence between the test point and the quality indicator.

[0013] Construction procedures refer to the distinct stages of a building's construction process, defined according to a predetermined workflow, each with its own independent work content and quality acceptance requirements. For example, in the construction of a standard floor of Building A in a high-rise residential project, the construction procedures sequentially include surveying and setting out, column reinforcement binding, column formwork installation, beam and slab formwork installation, beam and slab reinforcement binding, and concrete pouring. Quality control items refer to the inspection categories that require focused monitoring during specific construction procedures to ensure the finished product meets design specifications and acceptance standards. For example, in the concrete pouring process, quality control items include concrete slump control, vibration compaction control, and pouring thickness control.

[0014] Among them, the test points refer to the specific physical locations on the construction site used to collect quality perception data according to the requirements of the quality control project; for example, in the core tube shear wall area of ​​the standard floor during the concrete pouring process, the corner of the wall, the middle of the wall, and the edge of the door and window openings are selected as three test points for concrete strain monitoring. Quality indicators refer to measurable physical quantities or process parameters used to quantitatively evaluate whether the construction quality at the test points meets the standards; for example, for the above three test points, the quality indicator is the concrete strain value, which reflects the changes in internal stress of the concrete during the hardening process.

[0015] The correspondence refers to the mapping relationship established between the test points and the quality indicators, which is used to clarify the quality indicators that need to be monitored for each test point. For example, the test points at the corners of the core tube shear wall correspond to two quality indicators: concrete strain value and internal hydration heat temperature of concrete, while the test points in the middle of the wall correspond to one quality indicator: concrete strain value.

[0016] S2. Receive multi-dimensional quality sensing data collected in real time by multi-source sensing devices deployed at the test point. The multi-dimensional quality sensing data includes at least structural mechanical parameters, material state parameters, and environmental parameters. Record the timestamp and point identifier corresponding to the multi-dimensional quality sensing data.

[0017] Among them, multi-source sensing equipment acquisition refers to the process of simultaneously acquiring multiple physical quantity data through different types of sensors deployed at the construction site; for example, in the concrete pouring process, vibrating wire strain sensors are deployed at the test points at the corners of shear walls to collect concrete strain values, temperature and humidity sensors are deployed to collect ambient temperature and humidity, and at the same time, embedded temperature sensors are used to collect the internal hydration heat temperature of the concrete.

[0018] Structural mechanics parameters refer to physical quantities that reflect the stress and deformation state of building structural components under load or their own deformation; for example, the strain value of concrete collected by a vibrating wire strain sensor is 3.2 × 10⁻⁶. -4 This indicates the presence of micro-strain within the concrete. Material state parameters refer to physical quantities that reflect the physicochemical properties and current state of building materials; for example, the hydration heat temperature inside the concrete collected by the embedded temperature sensor is 65℃, indicating that the concrete is in a stage of intense hydration reaction. Environmental parameters refer to physical quantities that reflect the natural environmental conditions surrounding the construction site; for example, the ambient temperature collected by the temperature and humidity sensor is 28℃, the ambient humidity is 75%, and the wind speed is 3.5m / s.

[0019] The timestamp refers to the time stamp recorded by the multi-source sensing device when collecting multi-dimensional quality sensing data, used to calibrate the time information of data generation; for example, if a vibrating wire strain sensor collects concrete strain values ​​at 14:30:15 on January 1, 2026, the corresponding timestamp record is 20260101143015. The location identifier refers to the coded information used to uniquely distinguish different test locations on the construction site; for example, the location identifier for a test location at the corner of the core tube shear wall is P01, the location identifier for a test location in the middle of the wall is P02, and the location identifier for a test location at the edge of door and window openings is P03.

[0020] S3. Based on the correspondence, the multidimensional quality sensing data is time-series aligned according to the timestamp and the location identifier to construct a quality time-series data sequence for each location to be measured.

[0021] Among them, the quality time series data sequence refers to the data set formed by arranging the same test point in the same quality index according to the order of timestamps; for example, the concrete strain value of test point P01 is collected every 15 minutes within 72 hours after the concrete is poured, and all the collected values ​​are arranged in the order of timestamps to form the quality time series data sequence of concrete strain value.

[0022] S4. Call the preset anomaly detection rule library, which stores the mapping relationship between the quality indicators and threshold conditions. Match the quality time series data sequence with the threshold conditions of the corresponding quality indicators, mark the test points with data anomalies as quality deviation points, and record the point anomaly event information corresponding to the quality deviation points. The point anomaly event information includes the deviation quality indicator, the deviation value, and the time of anomaly occurrence.

[0023] The anomaly detection rule base refers to a pre-established set of rules used to determine whether anomalies occur in a quality time-series data sequence. This rule set stores the correspondence between quality indicators and threshold conditions. For example, the anomaly detection rule base stores the relationship between concrete strain values ​​and the static threshold condition "not exceeding 3.5 × 10⁻⁶". -4 The mapping relationship between "" and "" and the mapping relationship between the internal hydration heat temperature of concrete and the dynamic threshold condition "temperature rise rate not exceeding 5℃ / h". The threshold condition refers to the boundary value or boundary interval used to determine whether the collected value corresponding to the quality index is within the normal range; for example, the threshold condition for concrete strain value is a fixed upper limit of 3.5×10. -4 The threshold condition for the internal hydration heat temperature of concrete is a temperature rise rate range that fluctuates according to changes in ambient temperature. The mapping relationship refers to a one-to-one correspondence between individual quality indicators and individual threshold conditions in the anomaly detection rule base; for example, the quality indicator "concrete strain value" and the threshold condition "not exceeding 3.5 × 10⁻⁶".-4 This forms one mapping relationship, and the quality indicator "internal hydration heat temperature of concrete" and the threshold condition "temperature rise rate not exceeding 5℃ / h" form another mapping relationship.

[0024] Among them, quality deviation points refer to the test points in the quality time series data sequence whose collected values ​​exceed the normal range defined by the corresponding threshold conditions; for example, the concrete strain collected at test point P01 at 14:45 is 3.8 × 10 -4 Exceeding the threshold condition of 3.5 × 10 -4 The test point P01 is marked as a quality deviation point. Point anomaly information refers to descriptive information recording events where quality indicators exceed limits at the quality deviation point; for example, the point anomaly information for quality deviation point P01 includes the deviation quality indicator being the concrete strain value and the deviation value being 3.8 × 10⁻⁶. -4 The anomaly occurred at 14:45 on January 1, 2026. The deviation quality index refers to the name of the quality index that was determined to exceed the threshold condition at the quality deviation point; for example, in the anomaly event information for the measured point P01, the deviation quality index is the concrete strain value. The deviation value refers to the specific collected value of the quality index that was determined to exceed the threshold condition at the quality deviation point; for example, in the anomaly event information for the measured point P01, the deviation value is 3.8 × 10⁻⁶. -4 The time of anomaly occurrence refers to the specific moment when the quality index at the quality deviation point is determined to exceed the threshold condition; for example, the time of anomaly occurrence in the abnormal event information of the test point P01 is 14:45 on January 1, 2026.

[0025] S5. Calculate the quality index deviation of the quality deviation point based on the abnormal event information of the point, and input the quality index deviation into a pre-established multi-layer weighted directed graph model. The multi-layer weighted directed graph model determines the propagation path of the influence of the quality index deviation on the quality index of at least one related process and the propagation influence value corresponding to the propagation path.

[0026] Among them, the quality index deviation refers to the quantified deviation value obtained by quantifying the degree to which the deviation value at the quality deviation point deviates from the benchmark threshold; for example, based on the deviation value of the test point P01, which is 3.8 × 10 -4 and baseline threshold 3.5×10 -4 The quality index deviation D is calculated by combining the rate of change factor and the process sensitivity weighting factor. 11 It is 0.21.

[0027] The multi-layer weighted directed graph model refers to a pre-constructed mathematical model used to characterize the influence transmission relationship between different construction quality indicators. This model includes a process layer and an indicator layer. Nodes in the process layer represent construction processes, and nodes in the indicator layer represent construction quality indicators. Directed edges in the multi-layer weighted directed graph model represent the influence transmission relationship between different construction processes and construction quality indicators. Each directed edge corresponds to an influence propagation weight. The influence propagation weights are pre-determined in the following ways: First, samples of quality indicator deviations from different construction processes are collected from historical construction quality data. Using the deviation quality indicator nodes as independent variables and the actual deviation values ​​of the associated process quality indicator nodes as dependent variables, a multiple linear regression method is used to fit the regression coefficients of each directed edge, which are then used as the influence propagation weights. Second, multiple engineering experts score the influence degree of each directed edge, and the arithmetic mean of the scores is normalized and used as the influence propagation weight.

[0028] Among them, the quality index of the associated process refers to the quality index of other construction processes that have an influence transmission relationship with the quality index that has deviated in the multi-layer weighted directed graph model; for example, if the concrete strain value deviates in the concrete pouring process, the quality index of the associated process is determined to be the axial force value of the formwork support process according to the multi-layer weighted directed graph model.

[0029] In this context, the influence propagation path refers to the sequence of nodes traversed in a multi-layered weighted directed graph model, starting from the quality index node where the deviation occurs and proceeding along the directed edges to the quality index node of the associated process. For example, starting from the concrete strain value node, passing through the concrete internal stress node, and reaching the support axial force value node, the sequence of nodes traversed constitutes the influence propagation path. The propagation influence value refers to the quantified value of the degree of influence generated by the quality index deviation being transmitted along the influence propagation path to the quality index node of the associated process. For example, using the quality index deviation of 0.21 as the initial disturbance value, multiplying it sequentially by the influence propagation weights corresponding to each directed edge on the influence propagation path, the propagation influence value C1 of the support axial force value node is obtained as 0.14.

[0030] S6. Summarize the quality index deviation of all test points under the same construction process, the influence propagation path and propagation influence value corresponding to all quality deviation points, and generate a process-level comprehensive quality deviation index.

[0031] Among them, the process-level quality comprehensive deviation index refers to a comprehensive index used to characterize the overall quality deviation of a construction process, which is calculated by combining the quality index deviation of all test points under the same construction process and the propagation influence value corresponding to all quality deviation points. For example, the process-level quality comprehensive deviation index E of the concrete pouring process is calculated as 1.78 by multiplying the quality index deviation of all test points under the concrete pouring process by the corresponding position weight coefficient and summing them, and then adding the sum of the propagation influence values ​​corresponding to the three influence propagation paths after attenuation penalty.

[0032] S7. Based on the comprehensive deviation index of the process-level quality, query the preset evaluation level mapping table to determine the corresponding quality level and handling suggestions. Combine the quality level, the handling suggestions, and the deviation quality indicators, deviation values, abnormal occurrence time, and location identifiers in the location anomaly event information to generate comprehensive construction quality evaluation information containing text descriptions and structured data fields.

[0033] The assessment level mapping table refers to a data table that pre-stores the correspondence between the comprehensive deviation index range of the process level and the quality level and handling suggestions. For example, the mapping table sets the index range [0, 1.0) to "qualified" level and handling suggestion "normal monitoring", [1.0, 2.5) to "general deviation" level and handling suggestion "strengthen inspection", and [2.5, ∞) to "serious deviation" level and handling suggestion "immediate stoppage for rectification".

[0034] It should be noted that when determining the corresponding quality level and handling recommendations, the calculated process-level comprehensive quality deviation index is matched with the index range in the mapping table. If E=1.78, then the range [1.0, 2.5) is matched, determining the quality level as "general deviation" and the handling recommendation as "strengthening inspection". When generating evaluation information, the determined quality level, handling recommendations, and deviation quality indicators, deviation values, anomaly occurrence time, and location identifiers in the location anomaly event information are structurally combined. For example, the text description "The quality level of the concrete pouring process is general deviation, and it is recommended to strengthen inspection; the concrete strain value at anomaly point P01 exceeds the limit (3.8 × 10⁻⁶)" is generated. -4 The data, which occurred on 2026-01-01 at 14:45, along with the corresponding structured data fields (such as grade codes, a list of recommended measures, and a list of abnormal locations), forms a complete comprehensive assessment of construction quality.

[0035] Among them, the comprehensive construction quality assessment information refers to the assessment conclusions generated based on the comprehensive deviation index of process-level quality and the information on abnormal events at the test points, which comprehensively reflect the construction quality status. For example, the comprehensive construction quality assessment information includes the comprehensive deviation index of process-level quality of concrete pouring process (1.78), the abnormal event record of the concrete strain value exceeding the limit at the test point P01 at 14:45 on January 1, 2026, and the assessment conclusion that it is recommended to strengthen the curing of concrete in the shear wall area of ​​the core tube of Building A and to review the adjacent formwork support system.

[0036] The technical solution of this embodiment collects multi-dimensional quality perception data of the test points in the construction process in real time by deploying multi-source sensing devices, and performs time-series alignment processing based on timestamps and point identifiers. Combined with the anomaly detection rule base, it automatically marks the quality deviation points and abnormal event information, and uses a multi-layer weighted directed graph model to analyze the deviation propagation path and impact value. It summarizes and generates a process-level comprehensive quality deviation index and comprehensive construction quality assessment information, which solves the technical problems of low data collection efficiency, delayed information transmission, insufficient data integrity and accuracy, and difficulty in tracing and analyzing historical quality data under the traditional manual inspection and paper record mode. It improves the level of construction quality control and the degree of refinement of engineering quality management.

[0037] In one optional embodiment, the multi-source sensing device includes a vibrating wire strain sensor deployed at the test point on the concrete structure, an ultrasonic stress sensor deployed at the test point on the steel structure, a temperature and humidity sensor deployed within a set range around the test point, and a data acquisition terminal. The data acquisition terminal is used to collect the structural mechanical parameters collected by the vibrating wire strain sensor or ultrasonic stress sensor corresponding to the same test point and the environmental parameters collected by the temperature and humidity sensor, and to add corresponding timestamps and point identifiers to the collected multi-dimensional quality sensing data.

[0038] In the above-mentioned optional methods, mechanical parameters of concrete and steel structures are further collected by vibrating wire strain sensors and ultrasonic stress sensors, respectively, and environmental parameters are obtained by temperature and humidity sensors. The data acquisition terminal realizes the synchronous collection and spatiotemporal marking of multi-source data, which solves the technical problem of insufficient monitoring dimensions of single sensors and improves the integrity and reliability of construction quality perception data.

[0039] In one alternative approach, the material state parameters in the multidimensional quality sensing data include the internal hydration heat temperature of concrete, the surface moisture content of concrete, and the thickness of the steel reinforcement protective layer; the environmental parameters include ambient temperature, ambient humidity, and wind speed; and the structural mechanical parameters include concrete strain values, steel structure stress values, and support axial force values.

[0040] Among the above-mentioned optional methods, the parameter composition of multi-dimensional quality perception data is further refined. Hydration heat temperature, surface moisture content, and protective layer thickness are included in the scope of material condition monitoring, ambient temperature, humidity, and wind speed are included in the scope of environmental monitoring, and strain value, stress value, and support axial force value are included in the scope of structural mechanics monitoring. This solves the technical problem of incomplete coverage of quality indicators and improves the precision of construction quality assessment.

[0041] In one optional approach, the step of performing time-series alignment processing on the multidimensional quality sensing data according to the timestamp and the location identifier includes interpolating the missing quality sensing data of the same location to be tested between two adjacent timestamps. The interpolation method adopts a weighted moving average interpolation algorithm based on the historical quality time-series data sequence change trend of the same location to be tested.

[0042] Among the above-mentioned optional methods, a weighted moving average interpolation algorithm is further introduced in the time series alignment process. Based on the historical data change trend of the same test point, the missing data between adjacent timestamps is intelligently filled, which solves the technical problem of time series breakage caused by missing data and improves the continuity and availability of quality time series data sequences.

[0043] In one optional approach, the mapping relationship between the quality indicators and threshold conditions stored in the anomaly detection rule base includes a static threshold mapping relationship and a dynamic threshold mapping relationship. The threshold conditions in the static threshold mapping relationship are fixed upper or lower limits that do not change with environmental parameters, while the threshold conditions in the dynamic threshold mapping relationship are floating threshold ranges that are adjusted according to changes in environmental parameters based on a preset functional relationship.

[0044] Among the above-mentioned optional methods, an anomaly detection rule library containing static thresholds and dynamic thresholds is further constructed. The static threshold is suitable for environmentally insensitive indicators, while the dynamic threshold is automatically adjusted according to the changes in environmental parameters based on a preset functional relationship. This solves the technical problem that fixed thresholds are difficult to adapt to complex construction environments and improves the accuracy and environmental adaptability of anomaly detection.

[0045] In one alternative approach, the influence propagation weights corresponding to the directed edges in the multi-layer weighted directed graph model are predetermined by: collecting samples of quality index deviations for different construction processes from historical construction quality data; using the deviation quality index nodes as independent variables and the actual deviation values ​​of the quality index nodes of the associated processes as dependent variables; and using a multiple linear regression method to fit the regression coefficients of each directed edge as influence propagation weights; or, having multiple engineering experts score the degree of influence of each directed edge, and normalizing the arithmetic mean of the scores as influence propagation weights.

[0046] In the step of determining the influence propagation path of the quality index deviation on at least one related process quality index and the propagation influence value corresponding to the influence propagation path by the multi-layer weighted directed graph model, the quality index deviation is used as the initial perturbation value of the corresponding index layer node in the multi-layer weighted directed graph model. The related process quality index nodes are traversed layer by layer along the direction of the directed edge. The node sequence passed during the traversal is recorded as the influence propagation path. The initial perturbation value is multiplied by the influence propagation weight corresponding to each directed edge passed through to obtain the propagation influence value of the related process quality index node.

[0047] Among the above-mentioned optional methods, a multi-layer weighted directed graph model containing process layer and index layer is further constructed. Nodes represent construction processes and quality indicators, and directed edges represent the cross-process quality influence transmission relationship and are assigned propagation weights. This solves the technical problem that it is difficult to quantify and trace the cross-process propagation path of quality deviation and improves the systematic nature of quality influence analysis.

[0048] In one optional approach, in the step of determining the influence propagation path of the quality index deviation on at least one related process quality index and the propagation influence value corresponding to the influence propagation path by the multi-layer weighted directed graph model, the quality index deviation is used as the initial perturbation value of the corresponding index layer node in the multi-layer weighted directed graph model. The related process quality index nodes are traversed layer by layer along the direction of the directed edges. The sequence of nodes passed during the traversal is recorded as the influence propagation path. The initial perturbation value is multiplied by the influence propagation weight corresponding to each directed edge passed through to obtain the propagation influence value of the related process quality index node.

[0049] In the above-mentioned optional methods, the quality index deviation is further used as the initial perturbation value of the corresponding index layer node in the multi-layer weighted directed graph model. The quality index nodes of the associated process are traversed layer by layer along the direction of the directed edge, and the node sequence is recorded as the influence propagation path and the propagation influence value is calculated. This solves the technical problem that the influence range of quality deviation is difficult to track dynamically and improves the accuracy of cross-process quality risk assessment.

[0050] In one alternative approach, in the step of calculating the quality index deviation of the quality deviation point based on the point anomaly event information, the quality index deviation is calculated according to the following formula: ; In the formula, Indicates the first The number of test points corresponds to the first... The deviation of a quality indicator; This indicates that the corresponding number in the multidimensional quality perception data is... Real-time collected values ​​of quality indicators; This indicates that the corresponding anomaly detection rule in the rule base is... The benchmark threshold for each quality indicator; Indicates the first The real-time rate of change of a quality indicator between the current sampling time and the previous sampling time; Indicates the first The maximum allowable rate of change threshold for a quality indicator under the same sampling interval; and Let be the weight coefficient, and satisfy... ; It is a preset minimum positive number; Indicates the first The process sensitivity weighting factor for each quality indicator is preset based on the structural safety level of the current construction process.

[0051] Among them, the preset minimum positive number Used to prevent the denominator from being zero, and The value of is further used to adjust the sensitivity of the quality indicator deviation to small fluctuations in the rate of change. As the value increases, the response of the dynamic deviation component to fluctuations in the rate of change tends to be smoother. As the value decreases, the dynamic deviation component becomes more sensitive to fluctuations in the rate of change.

[0052] It should be noted that the formula for calculating the deviation of quality indicators is constructed by simultaneously considering the static deviation of real-time collected values ​​relative to a benchmark threshold and the dynamic deviation of the real-time rate of change relative to the maximum allowable rate of change. The static and dynamic deviation components are weighted and synthesized using preset weighting coefficients, and then multiplied by a process sensitivity weighting factor. In the formula, the static deviation component represents the instantaneous exceedance of the quality indicator as the ratio of the absolute value of the difference between the real-time collected value and the benchmark threshold to the benchmark threshold. The dynamic deviation component represents the abnormally drastic change trend of the quality indicator as the ratio of the absolute value of the real-time rate of change to the sum of the maximum rate of change threshold and the minimum positive number preventing zero change. The weighting coefficients are used to adjust the contribution ratio of static and dynamic deviations in the deviation calculation, and the process sensitivity weighting factor is used to adjust the final value of the deviation based on the structural safety level of the current construction process. The function of the above formula is to transform a single exceedance event at a quality deviation point into a quantitative deviation indicator that comprehensively considers the deviation magnitude, change trend, and process importance, providing a unified and comparable deviation input value for subsequent process-level comprehensive quality assessment.

[0053] Among the above-mentioned optional methods, a comprehensive algorithm that integrates relative deviation and rate of change deviation is further adopted to calculate the deviation of quality indicators. The process sensitivity weight factor is introduced to weight and correct the calculation results, which solves the technical problem that single-dimensional deviation assessment is difficult to reflect the actual quality risk and improves the scientificity and pertinence of quality deviation point identification.

[0054] In one optional approach, in the step of summarizing the quality index deviations of all test points under the same construction process, the influence propagation paths and propagation influence values ​​corresponding to all quality deviation points, and generating a process-level comprehensive quality deviation index, the process-level comprehensive quality deviation index is calculated according to the following formula: ; In the formula, This represents the overall deviation index of process-level quality. Indicates the first The number of test points corresponds to the first... The deviation of a quality indicator; Indicates the first The number of test points corresponds to the first... The positional weight coefficient of each quality indicator in the construction process; This represents the total number of test points under the aforementioned construction procedure; Indicates the first The total number of quality indicators corresponding to each test point; Indicates the penalty factor for cross-process impact; Indicates the first The propagation impact value corresponding to each propagation path; Indicates the propagation distance attenuation coefficient; Indicates the first The number of nodes that an influence propagation path passes through in the multi-layer weighted directed graph model; This represents the total number of all activated influence propagation paths under the same construction procedure.

[0055] Among them, the cross-process impact penalty factor Attenuation coefficient with propagation distance Synergistic effect Used to control the weight of cross-process quality deviations in the overall evaluation of this process. It is used to finely adjust the rate at which the influence value on a single influence propagation path decays as the number of propagation nodes increases. Together, these two factors enable the process-level quality comprehensive deviation index to take into account both the global importance of influence propagation and the differences in local path lengths.

[0056] It should be noted that the formula for calculating the comprehensive quality deviation index at the process level is constructed by weighting and summing the quality index deviations of all quality indicators at all test points within the construction process according to their corresponding position weight coefficients, and then adding this sum to the exponential attenuation penalty term for the cross-process propagation influence value. The first summation term in the formula is the cumulative value of the deviation of each quality indicator at each test point within the process multiplied by its corresponding position weight coefficient. The position weight coefficient reflects the structural criticality difference of the test point within the construction process. The second summation term in the formula is the cumulative value of the propagation influence value corresponding to all activated influence propagation paths multiplied by the propagation distance attenuation factor expressed as a natural exponential function, and then multiplied by the cross-process influence penalty factor for overall adjustment. The propagation distance attenuation factor is expressed as the product of the propagation influence value and the negative propagation distance attenuation coefficient of the natural constant e raised to the power of the number of path nodes, reflecting the characteristic that the quality deviation influence attenuates with increasing propagation path length. The function of the above formula is to integrate the local quality deviations at multiple points under the same construction process with the cross-process quality influence propagation effect into a single comprehensive value, so as to realize the systematic quantitative expression of the overall quality deviation of the process and provide core evaluation basis for the generation of comprehensive construction quality assessment information.

[0057] Among the above-mentioned optional methods, a comprehensive evaluation model that integrates the deviation of the point in this process and the propagation effect across processes is further constructed. The location weight coefficient, the cross-process influence penalty factor and the propagation distance attenuation coefficient are introduced for weighted calculation, which solves the technical problem that the process-level quality evaluation does not consider the cumulative effect of deviation propagation and improves the comprehensiveness and accuracy of the process-level quality comprehensive deviation index calculation.

[0058] This invention provides an automatic construction quality information evaluation system, which includes: The acquisition module is used to acquire the quality control items corresponding to the construction process, determine at least one test point and at least one quality indicator corresponding to each test point based on the quality control items, and establish the correspondence between the test point and the quality indicator. The receiving module is used to receive multi-dimensional quality sensing data collected in real time by multi-source sensing devices deployed at the test point. The multi-dimensional quality sensing data includes at least structural mechanical parameters, material state parameters and environmental parameters, and records the timestamp and point identifier corresponding to the multi-dimensional quality sensing data. The construction module is used to perform time-series alignment processing on the multidimensional quality sensing data according to the timestamp and the location identifier based on the correspondence, and construct a quality time-series data sequence for each of the locations to be measured; The matching module is used to call a preset anomaly detection rule library. The anomaly detection rule library stores the mapping relationship between the quality indicators and threshold conditions. It matches the quality time series data sequence with the threshold conditions of the corresponding quality indicators, marks the test points with data anomalies as quality deviation points, and records the point anomaly event information corresponding to the quality deviation points. The point anomaly event information includes the deviation quality indicator, the deviation value, and the time of anomaly occurrence. The generation module is used to calculate the quality index deviation of the quality deviation point based on the point abnormal event information, and input the quality index deviation into a pre-established multi-layer weighted directed graph model. The multi-layer weighted directed graph model determines the influence propagation path of the quality index deviation on the quality index of at least one related process and the propagation influence value corresponding to the influence propagation path. The summary module is used to summarize the quality index deviation of all the test points under the same construction process, the influence propagation path and propagation influence value corresponding to all the quality deviation points, and generate a process-level comprehensive quality deviation index. The evaluation module is used to query a preset evaluation level mapping table based on the comprehensive deviation index of the process-level quality, determine the corresponding quality level and handling suggestions, and combine the quality level, the handling suggestions and the deviation quality index, deviation value, abnormal occurrence time and location identifier in the location abnormal event information to generate comprehensive construction quality evaluation information containing text description and structured data fields.

[0059] The technical solution of this embodiment collects multi-dimensional quality perception data of the test points in the construction process in real time by deploying multi-source sensing devices, and performs time-series alignment processing based on timestamps and point identifiers. Combined with the anomaly detection rule base, it automatically marks the quality deviation points and abnormal event information, and uses a multi-layer weighted directed graph model to analyze the deviation propagation path and impact value. It summarizes and generates a process-level comprehensive quality deviation index and comprehensive construction quality assessment information, which solves the technical problems of low data collection efficiency, delayed information transmission, insufficient data integrity and accuracy, and difficulty in tracing and analyzing historical quality data under the traditional manual inspection and paper record mode. It improves the level of construction quality control and the degree of refinement of engineering quality management.

[0060] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for automatically evaluating construction quality information, characterized in that, The method includes: Obtain the quality control items corresponding to the construction process, determine at least one test point and at least one quality indicator corresponding to each test point based on the quality control items, and establish the correspondence between the test point and the quality indicator. The system receives multi-dimensional quality sensing data collected in real time from multi-source sensing devices deployed at the test points. The multi-dimensional quality sensing data includes at least structural mechanical parameters, material state parameters, and environmental parameters. The system also records the timestamp and point identifier corresponding to the multi-dimensional quality sensing data. Based on the correspondence, the multidimensional quality sensing data is time-series aligned according to the timestamp and the location identifier to construct a quality time-series data sequence for each location to be measured. The system calls a preset anomaly detection rule library, which stores the mapping relationship between the quality indicators and threshold conditions. The system matches the quality time series data sequence with the threshold conditions of the corresponding quality indicators, marks the test points with data anomalies as quality deviation points, and records the point anomaly event information corresponding to the quality deviation points. The point anomaly event information includes the deviation quality indicator, the deviation value, and the time of anomaly occurrence. The quality index deviation of the quality deviation point is calculated based on the abnormal event information of the point, and the quality index deviation is input into a pre-established multi-layer weighted directed graph model. The multi-layer weighted directed graph model determines the propagation path of the influence of the quality index deviation on the quality index of at least one related process and the propagation influence value corresponding to the propagation path. The quality index deviation of all test points under the same construction process is summarized, along with the influence propagation path and propagation influence value of all quality deviation points, to generate a process-level comprehensive quality deviation index. Based on the process-level quality comprehensive deviation index, a preset evaluation level mapping table is queried to determine the corresponding quality level and handling suggestions. The quality level, the handling suggestions, and the deviation quality indicators, deviation values, anomaly occurrence time, and location identifiers in the location anomaly event information are combined to generate construction quality comprehensive evaluation information containing text descriptions and structured data fields.

2. The automatic construction quality information evaluation method according to claim 1, characterized in that, The multi-source sensing device includes vibrating wire strain sensors deployed at the test points of the concrete structure, ultrasonic stress sensors deployed at the test points of the steel structure, temperature and humidity sensors deployed within a set range around the test points, and a data acquisition terminal. The data acquisition terminal is used to collect the structural mechanical parameters collected by the vibrating wire strain sensor or ultrasonic stress sensor corresponding to the same test point and the environmental parameters collected by the temperature and humidity sensor, and to add corresponding timestamps and point identifiers to the collected multi-dimensional quality sensing data.

3. The automatic construction quality information evaluation method according to claim 2, characterized in that, The material state parameters in the multidimensional quality perception data include the internal hydration heat temperature of concrete, the surface moisture content of concrete, and the thickness of the steel reinforcement protective layer. The environmental parameters include ambient temperature, ambient humidity, and wind speed. The structural mechanical parameters include concrete strain values, steel structure stress values, and support axial force values.

4. The automatic construction quality information evaluation method according to claim 1, characterized in that, In the step of performing time-series alignment processing on the multidimensional quality sensing data according to the timestamp and the location identifier, the time-series alignment processing includes interpolating the missing quality sensing data between two adjacent timestamps for the same location to be tested. The interpolation method adopts a weighted moving average interpolation algorithm based on the historical quality time-series data sequence change trend of the same location to be tested.

5. The automatic construction quality information evaluation method according to claim 1, characterized in that, The mapping relationship between the quality indicators and threshold conditions stored in the anomaly detection rule base includes a static threshold mapping relationship and a dynamic threshold mapping relationship. The threshold conditions in the static threshold mapping relationship are fixed upper or lower limits that do not change with environmental parameters. The threshold conditions in the dynamic threshold mapping relationship are floating threshold ranges that are adjusted according to changes in environmental parameters based on a preset function relationship.

6. The automatic assessment method for construction quality information according to claim 1, characterized in that, The influence propagation weights corresponding to the directed edges in the multi-layer weighted directed graph model are predetermined in the following ways: Samples of quality index deviations for different construction processes are collected from historical construction quality data. The deviation quality index nodes are used as independent variables, and the actual deviation values ​​of the associated process quality index nodes are used as dependent variables. Multiple linear regression is used to fit the regression coefficients of each directed edge as influence propagation weights. Alternatively, multiple engineering experts score the influence degree of each directed edge, and the arithmetic mean of the scores is normalized and used as the influence propagation weight.

7. The automatic construction quality information evaluation method according to claim 6, characterized in that, In the step of determining the influence propagation path of the quality index deviation on at least one related process quality index and the propagation influence value corresponding to the influence propagation path by the multi-layer weighted directed graph model, the quality index deviation is used as the initial perturbation value of the corresponding index layer node in the multi-layer weighted directed graph model. The related process quality index nodes are traversed layer by layer along the direction of the directed edge. The node sequence passed during the traversal is recorded as the influence propagation path. The initial perturbation value is multiplied by the influence propagation weight corresponding to each directed edge passed through to obtain the propagation influence value of the related process quality index node.

8. The automatic assessment method for construction quality information according to claim 1, characterized in that, In the step of calculating the quality index deviation of the quality deviation point based on the point anomaly event information, the quality index deviation is calculated according to the following formula: ； In the formula, Indicates the first The number of test points corresponds to the first... The deviation of a quality indicator; This indicates that the corresponding number in the multidimensional quality perception data is... Real-time collected values ​​of quality indicators; This indicates that the corresponding anomaly detection rule in the anomaly detection rule base is... The benchmark threshold for each quality indicator; Indicates the first The real-time rate of change of a quality indicator between the current sampling time and the previous sampling time; Indicates the first The maximum allowable rate of change threshold for a quality indicator under the same sampling interval; and Let be the weight coefficient, and satisfy... ; It is a preset minimum positive number; Indicates the first The process sensitivity weighting factor for each quality indicator is preset based on the structural safety level of the current construction process.

9. The automatic assessment method for construction quality information according to claim 7, characterized in that, In the step of summarizing the quality index deviations of all test points under the same construction process, the influence propagation paths and propagation influence values ​​corresponding to all quality deviation points, and generating a process-level comprehensive quality deviation index, the process-level comprehensive quality deviation index is calculated according to the following formula: ； In the formula, This represents the overall deviation index of process-level quality. Indicates the first The number of test points corresponds to the first... The deviation of a quality indicator; Indicates the first The number of test points corresponds to the first... The positional weight coefficient of each quality indicator in the construction process; This represents the total number of test points under the aforementioned construction procedure; Indicates the first The total number of quality indicators corresponding to each test point; Indicates the penalty factor for cross-process impact; Indicates the first The propagation impact value corresponding to each propagation path; Indicates the propagation distance attenuation coefficient; Indicates the first The number of nodes that an influence propagation path passes through in the multi-layer weighted directed graph model; This represents the total number of all activated influence propagation paths under the same construction procedure.

10. An automatic construction quality information evaluation system, characterized in that, The system includes: The acquisition module is used to acquire the quality control items corresponding to the construction process, determine at least one test point and at least one quality indicator corresponding to each test point based on the quality control items, and establish the correspondence between the test point and the quality indicator. The receiving module is used to receive multi-dimensional quality sensing data collected in real time by multi-source sensing devices deployed at the test point. The multi-dimensional quality sensing data includes at least structural mechanical parameters, material state parameters and environmental parameters, and records the timestamp and point identifier corresponding to the multi-dimensional quality sensing data. The construction module is used to perform time-series alignment processing on the multidimensional quality sensing data according to the timestamp and the location identifier based on the correspondence, and construct a quality time-series data sequence for each of the locations to be measured; The matching module is used to call a preset anomaly detection rule library. The anomaly detection rule library stores the mapping relationship between the quality indicators and threshold conditions. It matches the quality time series data sequence with the threshold conditions of the corresponding quality indicators, marks the test points with data anomalies as quality deviation points, and records the point anomaly event information corresponding to the quality deviation points. The point anomaly event information includes the deviation quality indicator, the deviation value, and the time of anomaly occurrence. The generation module is used to calculate the quality index deviation of the quality deviation point based on the point abnormal event information, and input the quality index deviation into a pre-established multi-layer weighted directed graph model. The multi-layer weighted directed graph model determines the influence propagation path of the quality index deviation on the quality index of at least one related process and the propagation influence value corresponding to the influence propagation path. The summary module is used to summarize the quality index deviation of all the test points under the same construction process, the influence propagation path and propagation influence value corresponding to all the quality deviation points, and generate a process-level comprehensive quality deviation index. The evaluation module is used to query a preset evaluation level mapping table based on the comprehensive deviation index of the process-level quality, determine the corresponding quality level and handling suggestions, and combine the quality level, the handling suggestions and the deviation quality index, deviation value, abnormal occurrence time and location identifier in the location abnormal event information to generate comprehensive construction quality evaluation information containing text description and structured data fields.

Images