Construction quality supervision method in project supervision

Abstract

The invention provides a construction quality supervision method in engineering supervision, and the method comprises the steps: collecting multi-dimensional quality data in a construction process, and extracting key quality feature parameters; constructing a construction quality evaluation model, and performing weight distribution on the key feature parameters in combination with historical quality data; calculating based on the deviation degree of the real-time monitoring data and a standard threshold value, and generating a quality early warning index; performing correlation analysis on the early warning index and an output result of the quality evaluation model, and determining a construction quality grade; and finally, a targeted rectification suggestion is generated according to the quality grade and the deviation degree. By adopting the scheme of the invention, the dynamic and precise supervision of the construction quality can be realized, and the supervision efficiency and the quality control level are improved.

Description

Technical Field

[0001] This invention relates to the field of engineering supervision technology, and in particular to a method for supervising construction quality in engineering supervision. Background Technology

[0002] As a core component in ensuring the quality of construction projects, the efficiency of construction supervision directly impacts the project's safety performance and lifespan. With modern construction projects becoming increasingly large-scale, complex, and intelligent, traditional construction quality supervision models are gradually revealing significant limitations.

[0003] At the data collection level, traditional supervision relies on manual inspections, paper records, and post-event spot checks, which suffers from problems such as untimely data collection and limited coverage. For example, for temperature changes during concrete curing, manual data recording every 4 hours makes it difficult to capture the impact of instantaneous temperature fluctuations on strength; for verticality deviations of high-rise building formwork, fixed-point measurements with a total station are insufficient for real-time dynamic monitoring, easily leading to the omission of hidden quality hazards.

[0004] In the quality assessment phase, traditional methods rely excessively on the experience and judgment of supervisors, resulting in strong subjectivity and inconsistent standards. At the same construction stage, different supervisors may give drastically different assessment results due to differences in their professional background and years of experience. For example, regarding deviations in the thickness of the rebar protective layer, an experienced supervisor might determine it to be "acceptable," while a novice supervisor might directly demand rework, leading to project delays or wasted resources.

[0005] In terms of risk warning, traditional models mostly rely on "post-event rectification" and lack proactive prevention and control mechanisms. By the time quality problems become apparent, substantial losses have often already occurred. For example, if changes in weld stress are not monitored in a timely manner during the welding process of steel structures, structural deformation may occur after the project is completed due to accumulated loads, requiring costly reinforcement.

[0006] Furthermore, with the widespread adoption of BIM technology and IoT sensors on construction sites, the quality data generated during construction is increasing exponentially. However, traditional supervision methods lack the ability to integrate and analyze multi-source data, resulting in a large amount of valuable information (such as vibration monitoring data and material arrival inspection reports) not being effectively utilized, leading to insufficient accuracy in quality control.

[0007] Therefore, how to construct a data-driven intelligent supervision method to achieve real-time monitoring, objective evaluation and dynamic early warning of construction quality has become a technical problem that urgently needs to be solved in the field of engineering supervision. Summary of the Invention

[0008] This application provides a method for supervising construction quality in engineering supervision, aiming to solve the problems of lagging data collection, strong subjectivity in assessment, and insufficient risk warning in traditional supervision methods, and to achieve dynamic and precise control of construction quality.

[0009] To achieve the above objectives, the present invention provides a technical solution as follows: a method for supervising construction quality in engineering supervision, comprising the following steps:

[0010] Step S1: Collect real-time construction data and historical quality data from the monitoring equipment in the construction site, and extract key quality characteristic parameters during the construction process;

[0011] Step S2: Construct a construction quality assessment model based on key quality characteristic parameters, and iteratively optimize the feature weights in the model using historical quality data;

[0012] Step S3: Calculate the deviation between real-time construction data and preset standard thresholds, and generate a construction quality early warning index based on the deviation.

[0013] Step S4: Input the key quality characteristic parameters into the construction quality assessment model to obtain the quality assessment value, and determine the construction quality level through the correlation analysis between the quality assessment value and the early warning index;

[0014] Step S5: Generate rectification suggestions based on the construction quality level and deviation, and push them to the supervision terminal.

[0015] As a further improvement to the technical solution of the present invention, step S1, extracting key quality characteristic parameters during the construction process, specifically includes:

[0016] The collected real-time construction data is preprocessed to remove outliers and noise.

[0017] Parameters strongly correlated with construction quality were selected from the preprocessed data, including material strength, structural dimensional deviation, and construction process parameters.

[0018] Principal component analysis was used to reduce the dimensionality of the screened parameters to obtain key quality characteristic parameters.

[0019] As a further improvement to the technical solution of the present invention, step S2, constructing the construction quality assessment model specifically includes:

[0020] The model is initialized with key quality feature parameters in the input layer, a backpropagation (BP) neural network structure in the hidden layer, and quality evaluation values ​​in the output layer.

[0021] A training set is constructed based on historical quality data, where the input is historical key feature parameters and the label is the quality score of the corresponding construction section.

[0022] The model loss function is minimized using the gradient descent algorithm, and the network weights are iteratively updated until the model converges.

[0023] As a further improvement to the technical solution of this invention, the gradient descent algorithm minimizes the model loss function as follows:

[0024] L(θ) = 1 / N × Σ(y_i - ŷ_i)² + λ × ||θ||2²

[0025] Where θ is the model parameter, N is the number of samples, y_i is the actual quality score, ŷ_i is the model predicted value, and λ is the regularization coefficient.

[0026] As a further improvement to the technical solution of the present invention, step S3, calculating the deviation between the real-time construction data and the preset standard threshold, specifically includes:

[0027] Preset the standard threshold ranges [μ_min, μ_max] for each key quality characteristic parameter;

[0028] Calculate the deviation rate between the real-time parameter value x and the midpoint μ0 of the threshold interval: d_i = |x - μ0| / (μ_max -μ_min);

[0029] The overall deviation is obtained by weighted summation: D = Σ(w_i × d_i), where w_i is the weight of the key feature parameter.

[0030] As a further improvement to the technical solution of the present invention, in step S3, generating the construction quality early warning index includes:

[0031] Set deviation thresholds D1 and D2; D1 < D2;

[0032] When D < D1, the warning index is 0, and a warning index of 0 means there is no warning.

[0033] When D1 ≤ D < D2, the warning index is 1, and a warning index of 1 indicates a Level 1 warning.

[0034] When D ≥ D2, the warning index is 2, and a warning index of 2 indicates a level 2 warning.

[0035] As a further improvement to the technical solution of the present invention, in step S4, determining the construction quality level includes:

[0036] Set quality assessment thresholds S1 and S2, where S1 < S2;

[0037] When the quality assessment value S ≥ S2 and the warning index is 0, the quality level is A (qualified).

[0038] When S1 ≤ S < S2 or the warning index is 1, the quality level is B (to be rectified).

[0039] When S < S1 or the warning index is 2, the quality grade is C (unqualified).

[0040] As a further improvement to the technical solution of this invention, a construction quality supervision system in engineering supervision includes:

[0041] The data acquisition module is used to collect real-time construction data and historical quality data from monitoring equipment;

[0042] The feature extraction module is used to process the collected data and extract key quality feature parameters;

[0043] The model building module is used to build and optimize construction quality assessment models.

[0044] The deviation analysis module is used to calculate the deviation degree and generate an early warning index;

[0045] The quality rating module is used to determine the construction quality level by combining the assessment value and the early warning index.

[0046] The rectification suggestion module is used to generate targeted suggestions based on the quality level and deviation.

[0047] As a further improvement to the technical solution of the present invention, a computer device includes a memory and a processor, wherein: when the processor executes a program stored in the memory, it implements a method for supervising construction quality in engineering supervision.

[0048] As a further improvement to the technical solution of the present invention, a computer-readable storage medium is provided, which stores a computer program, and when the program is executed by a processor, it implements a method for supervising construction quality in engineering supervision.

[0049] The present invention has the following beneficial effects:

[0050] 1. Enable real-time monitoring and dynamic evaluation of construction quality, reducing manual intervention;

[0051] 2. Improve the objectivity and accuracy of quality assessment through quantitative analysis and model prediction;

[0052] 3. By combining early warning mechanisms with rectification suggestions, quality risks can be avoided in advance and rework costs can be reduced. Attached Figure Description

[0053] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0054] Figure 1 is a flowchart illustrating a method for supervising construction quality in engineering supervision according to an embodiment of the present invention;

[0055] Figure 2 is a schematic diagram of the neural network structure of the quality assessment model in an embodiment of the present invention;

[0056] Figure 3 shows the correlation matrix between the early warning index and the quality level in an embodiment of the present invention;

[0057] Figure 4 is a schematic diagram of the composition of a computing device according to an embodiment of the present invention. Detailed Implementation

[0058] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. The illustrative embodiments and descriptions of the present invention are used to explain the present invention, but are not intended to limit the present invention.

[0059] It should be noted that all directional indicators (such as up, down, left, right, front, back, upper end, lower end, top, bottom, etc.) in the embodiments of the present invention are only used to explain the relative positional relationship and movement of each component in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indicator will also change accordingly.

[0060] In this invention, unless otherwise explicitly specified and limited, the term "connection" should be interpreted broadly. For example, "connection" can be a fixed connection, a detachable connection, or an integral part; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium; it can be the internal communication of two components or the interaction between two components, unless otherwise explicitly limited. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0061] Furthermore, in this invention, descriptions involving "first," "second," etc., are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined with "first" or "second" may explicitly or implicitly include at least one of that feature. Additionally, the technical solutions of various embodiments can be combined with each other, but only on the basis of being achievable by those skilled in the art. If the combination of technical solutions is contradictory or impossible to implement, such a combination should be considered non-existent and not within the scope of protection claimed by this invention.

[0062] The present invention will be further described in detail below with reference to the accompanying drawings.

[0063] Reference Figure 1 A method for supervising construction quality in engineering supervision, comprising the following steps:

[0064] Step S1: Collect real-time construction data and historical quality data from the monitoring equipment in the construction site, and extract key quality characteristic parameters during the construction process;

[0065] Step S2: Construct a construction quality assessment model based on key quality characteristic parameters, and iteratively optimize the feature weights in the model using historical quality data;

[0066] Step S3: Calculate the deviation between real-time construction data and preset standard thresholds, and generate a construction quality early warning index based on the deviation.

[0067] Step S4: Input the key quality characteristic parameters into the construction quality assessment model to obtain the quality assessment value, and determine the construction quality level through the correlation analysis between the quality assessment value and the early warning index;

[0068] Step S5: Generate rectification suggestions based on the construction quality level and deviation, and push them to the supervision terminal.

[0069] Specifically, in this embodiment, step S1, extracting key quality characteristic parameters during the construction process, specifically includes:

[0070] The collected real-time construction data is preprocessed to remove outliers and noise.

[0071] Parameters strongly correlated with construction quality were selected from the preprocessed data, including material strength, structural dimensional deviation, and construction process parameters.

[0072] Principal component analysis was used to reduce the dimensionality of the screened parameters to obtain key quality characteristic parameters.

[0073] Specifically, in this embodiment, step S2, constructing the construction quality assessment model, specifically includes:

[0074] The model is initialized with key quality feature parameters in the input layer, a backpropagation (BP) neural network structure in the hidden layer, and quality evaluation values ​​in the output layer.

[0075] A training set is constructed based on historical quality data, where the input is historical key feature parameters and the label is the quality score of the corresponding construction section.

[0076] The model loss function is minimized using the gradient descent algorithm, and the network weights are iteratively updated until the model converges.

[0077] Specifically, in this embodiment, the gradient descent algorithm minimizes the model loss function as follows:

[0078] L(θ) = 1 / N × Σ(y_i - ŷ_i)² + λ × ||θ||2²

[0079] Where θ is the model parameter, N is the number of samples, y_i is the actual quality score, ŷ_i is the model predicted value, and λ is the regularization coefficient.

[0080] Specifically, in this embodiment, step S3, calculating the deviation between real-time construction data and a preset standard threshold, specifically includes:

[0081] Preset the standard threshold ranges [μ_min, μ_max] for each key quality characteristic parameter;

[0082] Calculate the deviation rate between the real-time parameter value x and the midpoint μ0 of the threshold interval: d_i = |x - μ0| / (μ_max -μ_min);

[0083] The overall deviation is obtained by weighted summation: D = Σ(w_i × d_i), where w_i is the weight of the key feature parameter.

[0084] Specifically, in this embodiment, step S3, generating the construction quality early warning index, includes:

[0085] Set deviation thresholds D1 and D2; D1 < D2;

[0086] When D < D1, the warning index is 0, and a warning index of 0 means there is no warning.

[0087] When D1 ≤ D < D2, the warning index is 1, and a warning index of 1 indicates a Level 1 warning.

[0088] When D ≥ D2, the warning index is 2, and a warning index of 2 indicates a level 2 warning.

[0089] Specifically, in this embodiment, step S4, determining the construction quality level includes:

[0090] Set quality assessment thresholds S1 and S2, where S1 < S2;

[0091] When the quality assessment value S ≥ S2 and the warning index is 0, the quality level is A (qualified).

[0092] When S1 ≤ S < S2 or the warning index is 1, the quality level is B (to be rectified).

[0093] When S < S1 or the warning index is 2, the quality grade is C (unqualified).

[0094] Specifically, in this embodiment, a construction quality supervision system for engineering supervision includes:

[0095] The data acquisition module is used to collect real-time construction data and historical quality data from monitoring equipment;

[0096] The feature extraction module is used to process the collected data and extract key quality feature parameters;

[0097] The model building module is used to build and optimize construction quality assessment models.

[0098] The deviation analysis module is used to calculate the deviation degree and generate an early warning index;

[0099] The quality rating module is used to determine the construction quality level by combining the assessment value and the early warning index.

[0100] The rectification suggestion module is used to generate targeted suggestions based on the quality level and deviation.

[0101] Specifically, in this embodiment, a computer device includes a memory and a processor, wherein: when the processor executes a program stored in the memory, it implements a method for supervising construction quality in engineering supervision.

[0102] Specifically, in this embodiment, a computer-readable storage medium stores a computer program, which, when executed by a processor, implements a method for supervising construction quality in engineering supervision.

[0103] Specifically, in this embodiment, a computer-readable storage medium stores a computer program, which, when executed by a processor, implements a method for supervising construction quality in engineering supervision.

[0104] Reference Figures 1 to 3 Furthermore, it should be noted that the construction quality supervision method in the engineering supervision of this application specifically includes the following steps:

[0105] Multi-source data acquisition and key feature extraction

[0106] Real-time construction process data (covering material performance parameters, structural geometric parameters, and process execution parameters) is collected from IoT monitoring devices (such as stress sensors, infrared thermometers, and 3D laser scanners), high-definition monitoring systems, and engineering management platforms deployed at the construction site. Simultaneously, historical quality acceptance data and accident case data from similar projects are retrieved. After cleaning (removing outliers and filling missing values) and standardizing the collected raw data, principal component analysis (PCA) is used to reduce the dimensionality of the high-dimensional data, extracting key characteristic parameters that significantly affect construction quality, such as concrete curing strength, rebar cover thickness, and settlement of the formwork support system.

[0107] Construction and Optimization of Construction Quality Assessment Model

[0108] A construction quality assessment model based on an improved BP neural network was constructed. The input layer of the model consists of extracted key feature parameters, and the hidden layers are set to two layers (the number of nodes is 1.5 times and 1 times the number of feature parameters, respectively). The output layer is the quality assessment value from 0 to 100. A training set was constructed using historical quality data (the input is the historical key feature parameters, and the label is the actual quality score of the corresponding construction section). The model parameters were optimized by minimizing the loss function using the gradient descent algorithm. The loss function formula is as follows:

[0109] L(θ) = 1 / N × Σ(y_i - ŷ_i)² + λ × ||θ||2²

[0110] Where θ is the model weight parameter, N is the number of training samples, y_i is the actual quality score of the i-th sample, ŷ_i is the model predicted score, and λ is the regularization coefficient (ranging from 0.02 to 0.08), used to suppress overfitting and improve generalization ability. During model training, a 5-fold cross-validation method is used to adjust the learning rate in real time (initial value set to 0.01, decaying by 10% every 100 iterations) until the model's accuracy on the validation set stabilizes above 90%.

[0111] Deviation Calculation and Early Warning Index Generation

[0112] For each key characteristic parameter, a preset threshold range [μ_min, μ_max] specified by industry standards or design documents is established (e.g., the standard range for the compressive strength of concrete cubes is 30MPa-35MPa). The deviation rate between the real-time monitored parameter value x and the midpoint μ0 of the threshold range is calculated.

[0113] d_i = |x - μ0| / (μ_max - μ_min)

[0114] Where d_i is the individual deviation rate of the i-th key feature parameter. The weights w_i of each key feature parameter are determined using the analytic hierarchy process (AHP) (the sum of weights is 1), and the overall deviation is calculated using a weighted summation formula:

[0115] D = Σ(w_i × d_i)

[0116] Based on the importance of the project, deviation thresholds D1 (0.2) and D2 (0.5) are set to generate a three-level early warning index: when D < D1, the early warning index is 0 (no early warning); when D1 ≤ D < D2, the early warning index is 1 (level 1 early warning, indicating attention); when D ≥ D2, the early warning index is 2 (level 2 early warning, requiring immediate action).

[0117] Construction quality grade determination

[0118] The key feature parameters extracted in real time are input into the trained quality assessment model to obtain the quality assessment value S of the target construction section. Combining the assessment value with the early warning index, the construction quality is divided into three levels:

[0119] Grade A (Qualified): S ≥ 85 points and warning index is 0, indicating that the construction quality meets the design requirements and no rectification is required;

[0120] Grade B (To be rectified): 70 points ≤ S < 85 points or a warning index of 1 indicates a slight quality deviation, which needs to be rectified within a specified period and re-inspected.

[0121] Grade C (Unqualified): S < 70 points or warning index of 2 indicates serious quality hazards, requiring immediate work stoppage and rectification. Work can only resume after passing inspection.

[0122] Targeted rectification suggestions generation and push

[0123] Based on the quality grade, overall deviation rate, and individual deviation rate, the system automatically generates differentiated rectification suggestions. For example, when the quality is classified as Grade B and the deviation mainly comes from the rebar spacing (d_i=0.35, weight w_i=0.25), it is recommended to increase the number of positioning rebars, use a laser positioning device to calibrate the spacing, and retest every 2 hours and upload the data. When the quality is classified as Grade C and the concrete strength deviation rate d_i=0.6, it is recommended to immediately stop using this batch of concrete, test the cement grade and mix proportion, remix and send for testing. Rectification suggestions are pushed to on-site supervisors in real time through the supervision terminal (such as tablet computers, smart safety helmet displays) and are simultaneously updated to the project management platform.

[0124] To make the technical solution of this application clearer, the following detailed explanation is provided in the context of supervision during the construction of frame structures in building engineering:

[0125] Implementation Case: Quality Supervision of Concrete Construction for Frame Columns

[0126] Data Acquisition and Feature Extraction

[0127] Temperature sensors (monitoring concrete curing temperature, sampling once every 30 minutes), rebound hammers (testing concrete surface strength, once every 24 hours), and rebar scanners (testing rebar spacing and cover thickness, once per construction section) were deployed in the frame column construction area. The pouring process was recorded using a high-definition camera (one frame captured every 5 minutes). The mix design report for this batch of concrete, the cement arrival inspection report, and quality acceptance data for frame columns from three previous similar projects were retrieved from the project management platform. After preprocessing the collected data, five key characteristic parameters were extracted using PCA: 7-day concrete curing strength (X1), rebar spacing deviation (X2), cover thickness (X3), pouring height deviation (X4), and formwork verticality (X5).

[0128] Model training and evaluation

[0129] A quality assessment model was trained using 1000 sets of historical frame column construction data (700 training sets and 300 test sets). The input was [X1, X2, X3, X4, X5], and the output was a quality score (0-100 points). The loss function was set to λ=0.05. After 800 iterations, the model converged, and the average error on the test set was controlled within ±3 points.

[0130] Real-time monitoring and deviation calculation

[0131] During the construction of a certain frame column, the real-time monitoring data are as follows: X1 = 32 MPa (standard range 28 MPa - 32 MPa, μ0 = 30 MPa), X2 = +8 mm (standard range -5 mm to +5 mm, μ0 = 0 mm), X3 = 25 mm (standard range 20 mm - 30 mm, μ0 = 25 mm), X4 = +20 mm (standard range -10 mm to +10 mm, μ0 = 0 mm), X5 = 0.8‰ (standard range ≤ 1‰, μ0 = 0.5‰). The calculation yields:

[0132] d1=|32-30| / (32-28)=0.5; d2=|8-0| / (5+5)=0.8; d3=0; d4=|20-0| / (10+10)=1.0; d5=|0.8-0.5| / (1-0)=0.3

[0133] The weights [w1=0.3, w2=0.25, w3=0.2, w4=0.15, w5=0.1] are determined using the analytic hierarchy process. The overall deviation is calculated as follows: D = 0.3×0.5 + 0.25×0.8 + 0.2×0 + 0.15×1.0 + 0.1×0.3 = 0.15 + 0.2 + 0 + 0.15 + 0.03 = 0.53. Since D ≥ D² (0.5), the warning index is 2.

[0134] Quality level assessment and rectification suggestions

[0135] The model outputs a quality assessment value of S=68 points. Combined with the warning index of 2, the construction quality of the frame column is judged to be Grade C (unqualified). The system generates rectification suggestions: immediately stop subsequent pouring, and retest the strength of the poured section using the rebound method; adjust the spacing of the reinforcing bars to the standard range, and add positioning stirrups (spacing ≤500mm); recalibrate the formwork support system, and use a total station to monitor the verticality every hour until three consecutive data meet the standard (≤0.5‰).

[0136] This embodiment achieves accurate assessment and dynamic control of the construction quality of frame columns through quantitative analysis and intelligent models. Compared with traditional supervision methods, it advances the discovery time of quality problems by 12 hours and improves the rectification efficiency by 40%.

[0137] The above embodiments are merely preferred examples of this application. In practical applications, the key feature parameters, weights, and thresholds can be adjusted according to the type of project (such as bridges, tunnels, and high-rise buildings) to adapt to the supervision needs of different scenarios.

[0138] In addition, this application also provides a computer device, referring to Figure 4 The computer device includes a memory and a processor. The memory stores code, and the processor is configured to retrieve the code and execute the construction quality supervision method described above in the engineering supervision.

[0139] In some embodiments, the construction quality supervision method in the above-described engineering supervision can be implemented by a computer device, which includes at least one processor, a communication bus, a memory, and at least one communication interface.

[0140] A processor can be a general-purpose central processing unit (CPU) or an application-specific integrated circuit (ASIC).

[0141] A communication bus can be used to transmit information between the aforementioned components.

[0142] The memory can be read-only memory (ROM) or other types of static storage devices capable of storing static information and instructions, random access memory (RAM) or other types of dynamic storage devices capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, universal optical discs, Blu-ray discs, etc.), magnetic disks or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited to these. The memory can exist independently and be connected to the processor via a communication bus. The memory can also be integrated with the processor.

[0143] The memory stores program code for executing the solution of this application, and its execution is controlled by a processor. The processor executes the program code stored in the memory. The program code may include one or more software modules. In the above embodiments, the supervision method for construction quality in engineering supervision can be implemented by a processor and one or more software modules in the program code in the memory.

[0144] A communication interface is a device that uses any transceiver or similar device to communicate with other devices or communication networks, such as Ethernet, radio access network (RAN), wireless local area networks (WLAN), etc.

[0145] In a specific implementation, as one example, a computer device may include multiple processors, each of which may be a single-core (single-CPU) processor or a multi-core (multi-CPU) processor. Here, a processor may refer to one or more devices, circuits, and / or processing cores used to process data (e.g., computer program instructions).

[0146] The aforementioned computer device can be a general-purpose computer device or a special-purpose computer device. In specific implementations, the computer device can be a desktop computer, a portable computer, a network server, a handheld digital assistant (PDA), a mobile phone, a tablet computer, a wireless terminal device, a communication device, or an embedded device. This application does not limit the type of computer device.

[0147] In addition, this application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-mentioned method for supervising construction quality in engineering supervision.

[0148] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects:

[0149] In this embodiment, comprehensive construction data is acquired from IoT monitoring devices, monitoring systems, and management platforms through multi-source data acquisition and key feature extraction. Key feature parameters are obtained through preprocessing and principal component analysis, providing a precise data foundation for quality assessment. A construction quality assessment model based on an improved BP neural network is constructed, and parameters are optimized using the loss function L(θ) = 1 / N × Σ(y_i - ŷ_i)² + λ × ||θ||²². Combined with historical data, an objective and quantitative quality assessment is achieved. Through deviation calculation and early warning index generation, parameter deviations are monitored in real time based on the formulas d_i = |x - μ0| / (μ_max - μ_min) and D = Σ(w_i × d_i), providing early warnings of quality risks. Quality levels are classified according to the quality assessment value and early warning index, and targeted rectification suggestions are generated to achieve dynamic control of construction quality.

[0150] Specifically, firstly, the combination of multi-dimensional data collection and key feature extraction breaks through the limitations of traditional manual data collection. This not only enables real-time and comprehensive acquisition of various quality parameters during construction but also focuses on core influencing factors through principal component analysis, making quality assessment more targeted and scientific, and avoiding interference from redundant data. Secondly, the assessment model based on an improved BP neural network effectively enhances the model's generalization ability by introducing a regularization term into the loss function, reducing overfitting and making quality assessment values ​​closer to reality. This eliminates the subjectivity and inconsistency caused by excessive reliance on human experience in traditional supervision, achieving unified and objective assessment standards. Thirdly, the establishment of a deviation calculation and early warning index mechanism can issue timely warnings at the nascent stage of quality problems, transforming "post-event rectification" into "pre-event prevention," greatly reducing the probability of quality accidents and minimizing economic losses and project delays caused by rework and reinforcement. Fourthly, the differentiated rectification suggestions generated by combining quality level and deviation provide clear action guidelines for on-site construction and supervision personnel, improving the efficiency and targeting of rectification work and ensuring that quality problems are resolved quickly and effectively.

[0151] In summary, this application solution, through data-driven intelligent means, enables real-time monitoring, accurate assessment, early warning, and efficient rectification of construction quality in engineering supervision, comprehensively improving the intelligence level and work efficiency of construction quality supervision, and providing strong technical support for ensuring the quality of building projects.

[0152] The technical solutions provided by the embodiments of the present invention have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of the embodiments of the present invention. The descriptions of the embodiments above are only for helping to understand the principles of the embodiments of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the embodiments of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for supervising construction quality in engineering supervision, characterized in that, Includes the following steps: Step S1: Collect real-time construction data and historical quality data from the monitoring equipment in the construction site, and extract key quality characteristic parameters during the construction process; Step S2: Construct a construction quality assessment model based on key quality characteristic parameters, and iteratively optimize the feature weights in the model using historical quality data; Step S3: Calculate the deviation between real-time construction data and preset standard thresholds, and generate a construction quality early warning index based on the deviation. Step S4: Input the key quality characteristic parameters into the construction quality assessment model to obtain the quality assessment value, and determine the construction quality level through the correlation analysis between the quality assessment value and the early warning index; Step S5: Generate rectification suggestions based on the construction quality level and deviation, and push them to the supervision terminal.

2. The method for supervising construction quality in engineering supervision according to claim 1, characterized in that: In step S1, the extraction of key quality characteristic parameters during the construction process includes: The collected real-time construction data is preprocessed to remove outliers and noise. Parameters strongly correlated with construction quality were selected from the preprocessed data, including material strength, structural dimensional deviation, and construction process parameters. Principal component analysis was used to reduce the dimensionality of the screened parameters to obtain key quality characteristic parameters.

3. The method for supervising construction quality in engineering supervision according to claim 1, characterized in that: In step S2, constructing the construction quality assessment model includes: The model is initialized with key quality feature parameters in the input layer, a backpropagation (BP) neural network structure in the hidden layer, and quality evaluation values ​​in the output layer. A training set is constructed based on historical quality data, where the input is historical key feature parameters and the label is the quality score of the corresponding construction section. The model loss function is minimized using the gradient descent algorithm, and the network weights are iteratively updated until the model converges.

4. The method for supervising construction quality in engineering supervision according to claim 3, characterized in that: The gradient descent algorithm minimizes the model loss function as follows: L(θ) = 1 / N × Σ(y_i - ŷ_i)² + λ × ||θ||2² Where θ is the model parameter, N is the number of samples, y_i is the actual quality score, ŷ_i is the model predicted value, and λ is the regularization coefficient.

5. The method for supervising construction quality in engineering supervision according to claim 1, characterized in that: In step S3, calculating the deviation between the real-time construction data and the preset standard threshold includes: Preset the standard threshold ranges [μ_min, μ_max] for each key quality characteristic parameter; Calculate the deviation rate between the real-time parameter value x and the midpoint μ0 of the threshold interval: d_i = |x - μ0| / (μ_max - μ_min); The overall deviation is obtained by weighted summation: D = Σ(w_i × d_i), where w_i is the weight of the key feature parameter.

6. The method for supervising construction quality in engineering supervision according to claim 1, characterized in that: In step S3, generating the construction quality early warning index includes: Set deviation thresholds D1 and D2; D1 < D2; When D < D1, the warning index is 0, and a warning index of 0 means there is no warning. When D1 ≤ D < D2, the warning index is 1, and a warning index of 1 indicates a Level 1 warning. When D ≥ D2, the warning index is 2, and a warning index of 2 indicates a level 2 warning.

7. The method for supervising construction quality in engineering supervision according to claim 1, characterized in that: In step S4, determining the construction quality level includes: Set quality assessment thresholds S1 and S2, where S1 < S2; When the quality assessment value S ≥ S2 and the warning index is 0, the quality level is A. When S1 ≤ S < S2 or the warning index is 1, the quality level is B. When S < S1 or the warning index is 2, the quality level is C.

8. A construction quality supervision system for engineering supervision, implementing the method described in any one of claims 1-7, characterized in that, include: The data acquisition module is used to collect real-time construction data and historical quality data from monitoring equipment; The feature extraction module is used to process the collected data and extract key quality feature parameters; The model building module is used to build and optimize construction quality assessment models. The deviation analysis module is used to calculate the deviation degree and generate an early warning index; The quality rating module is used to determine the construction quality level by combining the assessment value and the early warning index. The rectification suggestion module is used to generate targeted suggestions based on the quality level and deviation.

9. A computer device, comprising a memory and a processor, characterized in that, When the processor executes a program stored in the memory, it implements the method as described in any one of claims 1-7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the program is executed by the processor, it implements the method as described in any one of claims 1-7.

Images