Deepseek-based new and old concrete interface bonding strength judgment method

Abstract

The invention discloses a method for judging the bonding strength of a new and old concrete interface based on Deepseek, and relates to the technical field of artificial intelligence type discriminating.The method comprises the steps that information acquisition is adopted, specifically, variable information of the bonding strength of the new and old concrete interface is acquired, and the variable information further comprises influence factor information influencing the bonding performance; factors of bonding strength of new and old concrete interfaces are extracted through an artificial intelligence Deepseek detection algorithm and are analyzed, and then levels of an overall architecture, functions of each layer, training and reasoning processes from input to output, and cooperative work of different components are summarized by using a recursive model of deep learning and a neural network. And data flow, hierarchical feature extraction and optimization target development are performed, and diversified tasks are adapted through different components, so that the bonding strength of the new and old concrete interface is analyzed and judged in a comprehensive understanding and preferential manner. The judgment method has remarkable advantages, influence characteristics and influence factor information variables can be automatically and efficiently extracted, and tasks can be flexibly processed, so that the workload of a user is reduced, and the work of judging the bonding strength of the new and old concrete interface is improved.

Description

Technical Field

[0001] The present invention belongs to the cross - field of artificial intelligence and civil engineering, and particularly relates to a method for judging the bonding strength of new - old concrete interfaces in information acquisition, data processing, and prediction. Background Art

[0002] In the field of civil engineering, the combined application of new and old concrete is extremely extensive, such as building reconstruction and expansion, bridge reinforcement and repair projects, etc. The bonding strength of the new - old concrete interface is directly related to the stability, durability, and safety of the structure. Accurately judging this strength is crucial for ensuring project quality.

[0003] Currently, traditional methods for judging the bonding strength of new - old concrete interfaces mainly rely on empirical formulas and simple experimental tests. Empirical formulas are often summarized based on limited experimental data and specific engineering conditions, with poor generality. Once encountering complex engineering scenarios, such as different combinations of concrete raw materials, special construction environments, and complex structural stress conditions, these formulas are difficult to accurately reflect the bonding strength of the new - old concrete interface.

[0004] For simple experimental test methods, such as direct tensile tests, shear tests, etc., although they can obtain bonding strength data to a certain extent, there are many drawbacks. On the one hand, the experimental process is cumbersome and time - consuming, requiring a large amount of human, material, and time costs. On the other hand, these experimental methods are destructive tests, which will cause a certain degree of damage to the concrete structure and affect the subsequent use performance of the structure.

[0005] In addition, traditional methods have limited capabilities in data processing and analysis. Facing numerous complex factors affecting the bonding strength of new - old concrete interfaces, such as concrete mix ratio, age, interface treatment method, environmental temperature and humidity, etc., it is difficult to comprehensively and deeply explore the internal relationships between various factors and their comprehensive influence on the bonding strength.

[0006] With the development of civil engineering construction towards large - scale and complex directions, higher requirements are put forward for the accuracy, efficiency, and non - destructiveness of judging the bonding strength of new - old concrete interfaces. The existing traditional judgment methods can no longer meet these needs, and there is an urgent need for an innovative judgment method based on advanced technology to solve the above problems.

[0007] Therefore, it is necessary to design a method for judging the bonding strength of new - old concrete interfaces based on deepSeek to solve the above - mentioned technical problems, break through the limitations of traditional methods, and bring new solutions to the field of civil engineering. Summary of the Invention

[0008] The purpose of the present invention is to provide a new method for judging the bonding strength of new - old concrete interfaces, which is simple to use, intelligent, and highly efficient.

[0009] The technical solution to achieve the purpose of the present invention is:

[0010] Based on deepseek, the judgment method of the bond strength between the new and old concrete interfaces integrates the IoT sensor network, collects construction dynamic parameters (such as ambient temperature and humidity, vibration frequency, etc.) in real time, and performs noise reduction and normalization processing through edge computing; at the same time, a cross-project data standardization template is established, which is compatible with CSV, JSON and BIM model formats, and the data is hashed and encrypted and stored using blockchain technology to ensure that the data cannot be tampered with and is traceable. Using advanced artificial intelligence and data collection platforms, core words such as "new and old concrete" and "interface bond strength" are extracted from multi-source databases to obtain variable information on the bond strength between the new and old interfaces of concrete. At the same time, detailed and accurate information is collected for factors affecting the bonding performance, including but not limited to the raw material composition of concrete, mix ratio, construction process parameters during pouring, changes in the ambient temperature and humidity of the structure, and the age of concrete, to ensure the integrity and accuracy of the data. With the help of deep learning and recursive models of neural networks, a judgment system for the bond strength between the new and old concrete interfaces is systematically constructed. In the process of model construction, the specific role of each level from the input layer to the output layer is clarified. Use a large amount of historical data and actual engineering case data to train the recursive model, adjust the model parameters, and improve the model's adaptability and judgment accuracy to different data patterns.

[0011] Preferably, the DeepSeek detection algorithm is optimized in a targeted manner, and the transfer learning technology is used to transfer the knowledge accumulated in other related fields to the analysis of the bond strength between new and old concrete interfaces, so as to speed up the model training and improve the accuracy of feature extraction; data enhancement technology is used, and synthetic minority class oversampling technology is applied to solve the data imbalance problem, and synthetic data that meets the engineering characteristics is generated based on the conditional generative adversarial network; high-order features are constructed through feature crossover (such as the interaction term between age and temperature and humidity) and sliding window statistics (such as the average humidity change rate during the construction stage). The original data is rotated, scaled, translated, etc., to expand the training data set and improve the generalization ability of the model.

[0012] Compared with the prior art, the present invention has the following significant advantages:

[0013] 1. More accurate and efficient data processing: Traditional empirical formulas have limited data sources, and manual data collection in simple experimental tests is prone to errors. This invention uses artificial intelligence technology and cooperates with DeepSeek intelligent algorithms to deeply mine key factors, greatly improving data processing accuracy and efficiency.

[0014] 2. High-efficiency and non-destructive detection: Traditional experimental tests are cumbersome, time-consuming, and damage the structure. Based on data analysis and judgment, the present invention does not require direct contact with concrete, which not only saves a large amount of time, manpower, and material resources but also avoids damage to the structure, and is particularly suitable for projects with high requirements for structural integrity.

[0015] 3. Strong ability to adapt to complex scenarios: Traditional empirical formulas are difficult to cope with complex engineering conditions. The recursive model of the present invention searches through the AutoML architecture and Bayesian optimization, combines a dynamic feedback mechanism with adversarial testing, can adaptively adjust model parameters, and still maintains high-precision prediction ability in extreme environments.

[0016] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. Description of the Drawings

[0017] Figure 1 It is a schematic flowchart of the method for judging the bonding strength of the new and old concrete interfaces based on deepseek of the present invention. Specific Embodiments

[0018] Using the deepseek database, literature retrieval is carried out with the keyword of "bonding strength of new and old concrete interfaces", and the literature closely related to this research is screened out. The data of the existing factors affecting the bonding performance of new and old concrete interfaces, such as concrete mix ratio, interface treatment method, age, environmental temperature and humidity, etc., are extracted from the literature. At the same time, the corresponding bonding strength test result data in the literature are collected and sorted into a structured data set as the basis for subsequent analysis.

[0019] The sorted data set is input into the DeepSeek algorithm, and the algorithm automatically performs feature extraction and in-depth analysis on the data. Using the visualization function of DeepSeek, the correlation and influence degree between each factor and the bonding strength are intuitively displayed.

[0020] According to the analysis results of the influencing factors, a suitable neural network architecture (such as multi-layer perceptron, recurrent neural network, etc.) is selected to construct a regression analysis calculation model for predicting the bonding strength of new and old concrete interfaces. The data set is divided into a training set and a test set according to a certain ratio (such as 70% for training and 30% for testing). The training set is used to train the model. During the training process, Bayesian optimization is used to replace the traditional grid search, the hyperparameter space is modeled based on Gaussian processes, and exploration and exploitation are dynamically balanced; the AutoML framework (such as AutoKeras) is integrated to automatically complete the model architecture search (NAS), and the generalization ability is evaluated through nested cross-validation. Early stopping is embedded during training to monitor the change of the loss function of the validation set to prevent overfitting; at the same time, the SHAP value is used to quantify the contribution degree of each input feature to the prediction result, and an interpretability report is generated to improve the fitting accuracy and generalization ability of the model.

[0021] Input the relevant data of new and old concrete collected in actual projects (including raw material information, construction process information, environmental information, etc.) into the trained model. The model processes the input data according to the learned rules and outputs the predicted value of the bond strength of the new and old concrete interface. The system analyzes and interprets the prediction results and provides a detailed analysis report. The report not only includes the predicted value of the bond strength, but also includes the contribution degree analysis of each influencing factor to the prediction result, as well as the comparative analysis with industry standards and design requirements.

[0022] When new engineering data is input, the system automatically calls the model for analysis and judgment and quickly outputs the predicted value of the bond strength of the new and old concrete interface. According to the prediction results, the system can also provide corresponding suggestions and measures. For example, in the case of insufficient bond strength, suggestions for adjusting the construction process or taking reinforcement measures are given, and whether measures to enhance the bond strength are needed, etc.

Claims

1. A method for determining the bond strength of the new-old concrete interface based on DeepSeek, characterized in that: include: Through the multi-dimensional data acquisition module, the variable information of the bond strength between the new and old concrete interfaces, as well as the key factors affecting the bonding performance including material properties, construction technology, environmental conditions, etc., can be accurately obtained. By using the proprietary artificial intelligence DeepSeek and deep learning detection algorithms, the characteristic factors closely related to the bond strength between the new and old concrete interfaces can be deeply mined and extracted to achieve efficient and accurate analysis.

2. The method for determining the bond strength between new and old concrete interfaces based on DeepSeek according to claim 1, characterized in that: Also includes: With the help of deep learning and recursive models of neural networks, the hierarchical logic of the overall architecture is systematically analyzed, and the entire process from input layer data import, to middle layer hierarchical feature extraction, model training and optimization, to output layer result presentation is clearly defined, achieving core goals such as extraction and model performance optimization, and realizing a method for determining the bond strength of the interface between new and old concrete.

3. The method for determining the bond strength between new and old concrete interfaces based on DeepSeek according to claim 1, characterized in that: Relying on intelligent algorithms, it has the ability to automatically and efficiently extract complex influencing features and massive influencing factor information variables.

Images