Temperature change distribution prediction method and system for concrete temperature field

Through the machine learning model based on neural network, multi-dimensional feature analysis and visual monitoring of concrete temperature field is solved, and the problem of inaccurate temperature prediction in the existing technology is realized, and the precise prediction and monitoring of concrete temperature field is achieved, ensuring the structure Safety and durability.

CN118916639AInactive Publication Date: 2024-11-08JIANGSU YUNJIN PAVEMENT ENGINEERING CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411074258.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-07
Publication Date
2024-11-08
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

In the prior art, the prediction of temperature variation distribution of concrete temperature fields depends on field measurement and empirical judgment, resulting in the prediction being inaccurate enough and unable to provide comprehensive temperature distribution information.

Method used

Using a neural network-based machine learning model, a visual temperature distribution map is generated through multi-dimensional feature collection, sensitivity analysis and dynamic feature collection, to achieve accurate prediction and monitoring of the concrete temperature field.

Benefits of technology

It improves the accuracy of concrete temperature prediction and comprehensiveness of temperature distribution information, ensuring the safety and durability of the structure.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118916639A_ABST
    Figure CN118916639A_ABST
Patent Text Reader

Abstract

The invention provides a temperature change distribution prediction method and system for a concrete temperature field, and relates to the technical field of temperature prediction, and the method comprises the steps: carrying out the multi-dimensional feature collection of a first unit block of concrete based on a predetermined dimension, and obtaining the target feature information; obtaining a first analysis result; establishing a predetermined key feature set and carrying out dynamic feature collection on the second unit block to obtain second key feature information; taking the second key feature information as input information of a temperature prediction model to obtain a second predicted temperature; and the second predicted temperature is combined with the second position of the second unit block to generate a visual temperature distribution diagram, and visual monitoring is carried out. Through the temperature prediction model, targeted temperature prediction is carried out on each block of the concrete, and a personalized prediction target of the temperatures of different positions of the concrete is realized, so that the accuracy of temperature prediction is improved, and the target of comprehensively and visually monitoring the temperatures of the positions of the concrete and the distribution condition of the temperatures is further realized.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the technical field of temperature prediction, and in particular to a method and system for predicting temperature change distribution of a concrete temperature field. Background Art

[0002] Temperature control is a crucial factor in the construction and maintenance of concrete structures. Improper temperature management may lead to concrete cracks, reduced strength, and other structural problems. Therefore, accurate prediction and monitoring of concrete temperature is essential to ensure the safety and durability of the structure. Traditional concrete temperature management methods rely on on-site measurements and empirical judgments, which are often not accurate enough and cannot provide comprehensive temperature distribution information. With the development of technology, especially the advancement of machine learning and data analysis technology, new possibilities have been provided for the prediction and monitoring of concrete temperature. Summary of the invention

[0003] The purpose of this application is to provide a method and system for predicting the temperature change distribution of a concrete temperature field, so as to solve the technical problems that the prediction of the temperature change distribution of a concrete temperature field in the prior art relies on on-site measurement and empirical judgment, and the temperature prediction is not accurate enough and cannot provide comprehensive temperature distribution information.

[0004] In view of the above problems, the present application provides a method and system for predicting temperature change distribution of concrete temperature field.

[0005] In a first aspect, the present application provides a temperature change distribution prediction method for a concrete temperature field, the method being implemented by a temperature change distribution prediction system for a concrete temperature field, wherein the method comprises: reading a predetermined dimension, and collecting multi-dimensional features of a first unit block of concrete based on the predetermined dimension to obtain target feature information; performing sensitivity analysis on the target feature information and a first real-time temperature of the first unit block to obtain a first analysis result; forming a predetermined key feature set based on the first analysis result, and performing dynamic feature collection on a second unit block based on the predetermined key feature set to obtain second key feature information; using the second key feature information as input information of a temperature prediction model to obtain a second predicted temperature, the temperature prediction model being an intelligent model obtained by machine learning of data in historical concrete temperature records based on the principle of a neural network; generating a visualized temperature distribution map based on the second predicted temperature in combination with a second position of the second unit block; and visually monitoring the temperature change of the concrete according to the visualized temperature distribution map.

[0006] In a second aspect, the present application further provides a temperature change distribution prediction system for a concrete temperature field, which is used to execute the temperature change distribution prediction method for a concrete temperature field as described in the first aspect, wherein the system includes: a feature collection module, the feature collection module is used to read a predetermined dimension, and based on the predetermined dimension, perform multi-dimensional feature collection on a first unit block of concrete to obtain target feature information; a sensitivity analysis module, the sensitivity analysis module is used to perform sensitivity analysis on the target feature information and the first real-time temperature of the first unit block to obtain a first analysis result; a key feature collection module, the key feature collection module is used to form a predetermined key feature set according to the first analysis result, and based on the predetermined key feature set, perform dynamic feature collection on a second unit block to obtain second key feature information; an intelligent prediction module, the intelligent prediction module is used to use the second key feature information as input information of a temperature prediction model to obtain a second predicted temperature, and the temperature prediction model is an intelligent model obtained by machine learning based on the principle of a neural network on data in historical concrete temperature records; a distribution map generation module, the distribution map generation module is used to generate a visualized temperature distribution map for the second predicted temperature combined with the second position of the second unit block; a visualization module, the visualization module is used to visually monitor the temperature change of the concrete according to the visualized temperature distribution map.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages: By reading the predetermined dimension, and collecting multi-dimensional features of the first unit block of concrete based on the predetermined dimension, the target feature information is obtained; the sensitivity analysis of the target feature information and the first real-time temperature of the first unit block is performed to obtain the first analysis result; a predetermined key feature set is formed according to the first analysis result, and dynamic features of the second unit block are collected based on the predetermined key feature set to obtain the second key feature information; the second key feature information is used as the input information of the temperature prediction model to obtain the second predicted temperature, and the temperature prediction model is an intelligent model obtained by machine learning of data in the historical concrete temperature record based on the principle of neural network; the second predicted temperature is combined with the second position of the second unit block to generate a visual temperature distribution map; the temperature change of the concrete is visually monitored according to the visual temperature distribution map. In other words, the temperature prediction model is used to perform targeted temperature prediction on each block of concrete, so as to achieve the personalized prediction goal of the temperature at different positions of the concrete, thereby improving the accuracy of temperature prediction, and further achieves the goal of comprehensive and intuitive monitoring of the temperature and its distribution at each position of the concrete through the visual temperature distribution map.

[0008] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented according to the contents of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are specifically cited below. It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easy to understand through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] In order to more clearly illustrate the technical solutions in the present application or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0010] Figure 1 The present invention is a flow chart of a method for predicting temperature change distribution of concrete temperature field.

[0011] Figure 2 This is a structural schematic diagram of a temperature change distribution prediction system for concrete temperature field in the present application.

[0012] Description of reference numerals: Feature collection module 11, sensitivity analysis module 12, key feature collection module 13, intelligent prediction module 14, distribution map generation module 15, visualization module 16. DETAILED DESCRIPTION

[0013] This application provides a method and system for predicting the temperature change distribution of concrete temperature field, which solves the technical problem that the prediction of the temperature change distribution of concrete temperature field in the prior art relies on on-site measurement and empirical judgment, and the temperature prediction is not accurate enough, and comprehensive temperature distribution information cannot be provided. The technical effect of improving the accuracy of temperature prediction of concrete structure and the comprehensiveness of temperature distribution information is achieved.

[0014] Below, the technical solutions in the present application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments of the present application. It should be understood that the present application is not limited to the example embodiments described herein. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application. It should also be noted that, for the convenience of description, only the parts related to the present application are shown in the accompanying drawings, rather than all of them.

[0015] Embodiment 1 Please refer to the attached Figure 1 The present application provides a method for predicting temperature change distribution of concrete temperature field, wherein the method is applied to a system for predicting temperature change distribution of concrete temperature field, and the method specifically comprises the following steps: Step 1: Read a predetermined dimension, and collect multi-dimensional features of a first unit block of concrete based on the predetermined dimension to obtain target feature information.

[0016] Step 2: Perform sensitivity analysis on the target characteristic information and the first real-time temperature of the first unit block to obtain a first analysis result.

[0017] Step three: forming a predetermined key feature set according to the first analysis result, and performing dynamic feature collection on the second unit block based on the predetermined key feature set to obtain second key feature information.

[0018] Step 4: Use the second key feature information as input information of a temperature prediction model to obtain a second predicted temperature. The temperature prediction model is an intelligent model obtained by machine learning based on the neural network principle on data in historical concrete temperature records.

[0019] Step 5: The second predicted temperature is combined with the second position of the second unit block to generate a visual temperature distribution map.

[0020] Step six: Visually monitor the temperature change of the concrete according to the visualized temperature distribution diagram.

[0021] Specifically, the predetermined dimensions are first read, and based on these dimensions, multi-dimensional features are collected for the first unit block of concrete to obtain target feature information. Then, a sensitivity analysis is performed on the collected target feature information and the real-time temperature of the first unit block to obtain the first analysis result. Next, a predetermined key feature set is formed based on the first analysis result, and then dynamic feature collection is performed on the second unit block based on this feature set to obtain the second key feature information, which is then used as the input information of the temperature prediction model. Among them, the temperature prediction model is an intelligent model obtained by machine learning of the data recorded by historical concrete temperature based on the principle of neural network. The output result of the temperature prediction model is the predicted temperature of the second unit block, that is, the second predicted temperature, which is combined with the position information of the second unit block to generate a visual temperature distribution map. Finally, according to this visual temperature distribution map, the temperature change of concrete can be monitored. By using machine learning and data analysis technology to predict and monitor the temperature change of concrete, the temperature of concrete can be managed more effectively to ensure the safety and durability of the structure. Achieving the goal of accurately predicting and monitoring the temperature change of concrete will help improve the efficiency and quality of concrete structure construction and maintenance.

[0022] Further, the method comprises the following steps: The predetermined dimensions include concrete dimensions, environmental condition dimensions and construction parameter dimensions, wherein the concrete dimensions include material components and material ratios, the environmental condition dimensions include air temperature, humidity and wind speed, and the construction parameter dimensions include pouring speed and curing conditions.

[0023] Specifically, the concrete construction process includes three main dimensions: concrete dimension, environmental condition dimension and construction parameter dimension. The concrete dimension focuses on the quality of concrete itself, including material components and material ratio. Material components refer to the types of various raw materials in concrete, such as cement, sand, gravel, etc., while material ratio refers to the specific proportion of these raw materials. These two factors directly affect the basic properties of concrete, such as strength and durability. The environmental condition dimension involves the impact of the construction environment on the performance of concrete, mainly including temperature, humidity and wind speed. Too high or too low temperature will affect the hardening speed and quality of concrete. Excessive humidity may cause excessive moisture on the surface of concrete, affecting strength, while excessive wind speed will accelerate the evaporation of moisture on the surface of concrete, causing cracking. The construction parameter dimension is related to the specific operation of concrete construction, including pouring speed and curing conditions. Pouring speed refers to the speed at which concrete is poured into the mold. Too fast may cause too many bubbles inside the concrete, affecting strength, while too slow may cause concrete stratification and affect uniformity. Curing conditions refer to environmental control during the hardening process of concrete, such as temperature and humidity control. Good curing conditions can ensure that concrete hardens evenly and achieves the expected performance.

[0024] In actual projects, it is necessary to comprehensively consider these three dimensions and determine various parameters through rigorous testing and calculation to ensure the safety and efficiency of construction. The material components and proportions of concrete need to be designed according to project requirements, and environmental conditions and construction parameters also need to be adjusted according to actual conditions to ensure that the quality of concrete meets project requirements.

[0025] Further, the method comprises the following steps: Taking multiple feature information of multiple features in the target feature information as independent variables; taking the first real-time temperature as a dependent variable; and analyzing a scatter plot drawn based on a mapping relationship between the independent variable and the dependent variable to obtain the first analysis result.

[0026] Specifically, the target feature information contains multiple features, each of which has its own specific information. These feature information are used as independent variables, which means that the target feature information is the variable that is attempted to be understood and controlled in the analysis. For example, these features may be various factors that affect a system or process, such as time, pressure, humidity, etc. Secondly, the first real-time temperature is a dependent variable, and the focus is on the result or output. In the process of data analysis of the target feature information and the first real-time temperature, it is necessary to analyze how the real-time temperature changes with the change of the target feature information. Then, the analysis is performed by drawing a scatter plot based on the mapping relationship between the independent variable and the dependent variable. A scatter plot is a chart used to show the relationship between two variables. In this case, each scatter point represents a data point, the horizontal axis is one or more feature information in the target feature information, and the vertical axis is the real-time temperature. By observing the scatter plot, it can be seen how the real-time temperature changes with the change of the target feature information. Finally, the first analysis result can be obtained by analyzing the scatter plot. This result may reveal a certain pattern or relationship between the target feature information and the real-time temperature, such as positive correlation, negative correlation, or no relationship. These findings may be very useful for understanding the behavior of the system or predicting future temperature changes.

[0027] In general, the plurality of characteristic information is used as independent variables, the first real-time temperature is used as the dependent variable, and the relationship between the plurality of characteristic information and the first real-time temperature is analyzed by drawing a scatter plot, thereby obtaining valuable analysis results.

[0028] Further, the method comprises the following steps: Extract a first feature from the multiple features, and traverse the scatter plot to obtain a first scatter plot of the first feature; read a first grid strategy, and grid the first scatter plot based on the first grid strategy to obtain a first grid result; introduce a grid mutual information evaluation function to analyze and evaluate the first grid result to obtain a first mutual information value; adjust the first grid strategy to obtain a second grid strategy, and grid the first scatter plot based on the second grid strategy to obtain a second grid result; obtain a second mutual information value of the second grid result according to the grid mutual information evaluation function; when the second mutual information value is greater than the first mutual information value, use the second mutual information value as the first target mutual information value of the first scatter plot; descend the multiple target mutual information values ​​of the multiple features formed based on the first target mutual information value to obtain a descending list of target mutual information values; use the descending list of target mutual information values ​​as the first analysis result.

[0029] Specifically, the scatter plot is gridded and the processing results are analyzed and evaluated using the grid mutual information evaluation function, and finally a descending list of target mutual information values ​​is obtained.

[0030] First, the first feature is extracted from multiple features, and the first scatter plot of the first feature is obtained by traversing the scatter plot. This step is to analyze the impact of each feature on the real-time temperature separately. Then the first grid strategy is read, and the first scatter plot is gridded based on the strategy to obtain the first gridding result. Gridding is to divide the scatter plot into a series of grids to better analyze the relationship between data distribution and features. Next, the grid mutual information evaluation function is introduced to analyze and evaluate the first gridding result to obtain the first mutual information value. Mutual information is an indicator to measure the mutual dependence between two variables, which is used here to evaluate the degree of correlation between features and real-time temperature. Then, the first grid strategy is adjusted to obtain the second grid strategy, and the first scatter plot is gridded based on the strategy to obtain the second gridding result. This step is to find a better grid strategy to better capture the relationship between features and real-time temperature. Next, the second mutual information value of the second gridding result is obtained according to the grid mutual information evaluation function. If the second mutual information value is greater than the first mutual information value, it means that the new grid strategy better captures the relationship between features and real-time temperature, so the second mutual information value is used as the first target mutual information value of the first scatter plot. Finally, the multiple target mutual information values ​​of the multiple features formed based on the first target mutual information value are arranged in descending order to obtain a descending list of target mutual information values. This list reflects the degree of correlation between each feature and the real-time temperature, which can be used as the first analysis result to help understand which features have a greater impact on the real-time temperature.

[0031] In general, by extracting features, gridding scatter plots, evaluating mutual information values, adjusting grid strategies, and constructing a descending list of target mutual information values, valuable analysis results can be obtained.

[0032] Further, the method comprises the following steps: The expression of the grid mutual information evaluation function is as follows: ;in, Characterizes the first of the multiple features The characteristics of Mutual information value, the The mutual information value refers to the The feature corresponding to Feature information With the first real-time temperature The mutual information of Characterize the KL divergence, Characterizing the Feature information With the first real-time temperature The joint probability distribution of Characterizing the Feature information With the first real-time temperature The product of the marginal probability distributions of A first strategy reliability characterizing the first grid strategy.

[0033] Furthermore, the method includes the following steps: reading predetermined strategy characteristics, and collecting characteristics of the first grid strategy based on the predetermined strategy characteristics to obtain first grid characteristic parameters; weighting the normalized first grid characteristic parameters to obtain the first strategy reliability; wherein the predetermined strategy characteristics include the total number of strategy grids and the average grid granularity of the strategy.

[0034] Specifically, the expression of the grid mutual information evaluation function is as follows: ;in, Characterizes the first of the multiple features The characteristics of Mutual information value, the The mutual information value refers to the The feature corresponding to Feature information With the first real-time temperature The mutual information of Characterize the KL divergence, Characterizing the Feature information With the first real-time temperature The joint probability distribution of Characterizing the Feature information With the first real-time temperature The product of the marginal probability distributions of A first strategy reliability characterizing the first grid strategy.

[0035] Further, based on the predetermined strategy features, the first grid strategy is feature collected and evaluated to obtain the reliability of the first strategy. First, the predetermined strategy features are read, which include the total number of strategy grids and the average grid granularity of the strategy. These two features are used to describe the basic properties of the grid strategy. The total number of grids reflects the complexity of the strategy, while the average grid granularity reflects the fineness of the strategy. Then, based on the predetermined strategy features, the first grid strategy is feature collected to obtain the first grid feature parameters. This step is to extract parameters related to the predetermined strategy features from the first grid strategy for subsequent evaluation. Next, the normalized first grid feature parameters are weighted. Normalization is to eliminate the dimensional influence between different features so that different features can be compared at the same scale. Weighting is to give different weights to different features according to the importance of each feature to obtain a more accurate evaluation result. Finally, the reliability of the first strategy is calculated based on the weighted first grid feature parameters. Strategy reliability is an indicator used to measure the pros and cons of the first grid strategy, that is, the effectiveness of the strategy in capturing the relationship between features and real-time temperature. In general, by reading the predetermined strategy features, collecting features of the grid strategy based on these features, normalizing and weighting the collected feature parameters, and finally obtaining the first strategy reliability, the effectiveness of the grid strategy is evaluated.

[0036] Further, the method comprises the following steps: Extract a first temperature position to be inserted in the visualized temperature distribution map; obtain a first search neighborhood of the first temperature position to be inserted based on a predetermined neighborhood scheme, the first search neighborhood including multiple neighborhood positions with predicted temperature identifiers; obtain a first neighborhood predicted temperature of a first neighborhood position among the multiple neighborhood positions with predicted temperature identifiers; use a first spatial distance between the first neighborhood position and the first temperature position to be inserted as a feedback adjustment coefficient to weightedly adjust the first neighborhood predicted temperature to obtain a first temperature to be inserted; render the first temperature to be inserted to the first temperature position to be inserted in the visualized temperature distribution map.

[0037] Specifically, first, the first temperature position to be inserted is extracted from the visualized temperature distribution map. This position is a specific point in the map, and its temperature value needs to be determined by prediction. Next, the first search neighborhood of the first temperature position to be inserted is obtained based on a predetermined neighborhood scheme. The neighborhood scheme defines how to select points adjacent to the temperature position to be inserted, and these points have predicted temperature identifiers, which means that a determined temperature value has been obtained. Then, from these neighborhood positions with predicted temperature identifiers, the first neighborhood predicted temperature of the first neighborhood position is obtained. This temperature value is predicted based on the actual temperature data of the neighborhood position. Next, the first neighborhood predicted temperature is weightedly adjusted with the first spatial distance between the first neighborhood position and the first temperature position to be inserted as the feedback adjustment coefficient. The closer the spatial distance, the greater the influence of the predicted temperature of the neighborhood position on the temperature to be inserted. In this way, the temperature value of the temperature position to be inserted can be predicted more accurately. Finally, the obtained first temperature to be inserted is rendered to the first temperature position to be inserted in the visualized temperature distribution map. Rendering is to display the calculated temperature value in a visual form for easy observation and analysis.

[0038] In general, by extracting the temperature position to be inserted, obtaining the neighborhood predicted temperature based on the neighborhood scheme, and making weighted adjustments based on the spatial distance, the predicted temperature value is finally rendered into a visual temperature distribution map.

[0039] In summary, the temperature change distribution prediction method for concrete temperature field provided by the present application has the following technical effects: By reading the predetermined dimension, and collecting multi-dimensional features of the first unit block of concrete based on the predetermined dimension, the target feature information is obtained; the sensitivity analysis of the target feature information and the first real-time temperature of the first unit block is performed to obtain the first analysis result; a predetermined key feature set is formed according to the first analysis result, and dynamic features of the second unit block are collected based on the predetermined key feature set to obtain the second key feature information; the second key feature information is used as the input information of the temperature prediction model to obtain the second predicted temperature, and the temperature prediction model is an intelligent model obtained by machine learning of data in the historical concrete temperature record based on the principle of neural network; the second predicted temperature is combined with the second position of the second unit block to generate a visual temperature distribution map; the temperature change of the concrete is visually monitored according to the visual temperature distribution map. In other words, the temperature prediction model is used to perform targeted temperature prediction on each block of concrete, so as to achieve the personalized prediction goal of the temperature at different positions of the concrete, thereby improving the accuracy of temperature prediction, and further achieves the goal of comprehensive and intuitive monitoring of the temperature and its distribution at each position of the concrete through the visual temperature distribution map.

[0040] Embodiment 2 Based on the same inventive concept as the temperature change distribution prediction method for concrete temperature field in the aforementioned embodiment, the present application also provides a temperature change distribution prediction system for concrete temperature field, see the attached Figure 2 , the system comprising: The feature collection module 11 is used to read a predetermined dimension and collect multi-dimensional features of a first unit block of concrete based on the predetermined dimension to obtain target feature information.

[0041] The sensitivity analysis module 12 is used to perform a sensitivity analysis on the target feature information and the first real-time temperature of the first unit block to obtain a first analysis result.

[0042] The key feature collection module 13 is used to form a predetermined key feature set according to the first analysis result, and to perform dynamic feature collection on the second unit block based on the predetermined key feature set to obtain second key feature information.

[0043] The intelligent prediction module 14 is used to use the second key feature information as input information of a temperature prediction model to obtain a second predicted temperature. The temperature prediction model is an intelligent model obtained by machine learning of data in historical concrete temperature records based on the principle of neural networks.

[0044] The distribution map generating module 15 is used to generate a visualized temperature distribution map based on the second predicted temperature and the second position of the second unit block.

[0045] A visualization module 16 is used to visually monitor the temperature change of the concrete according to the visualized temperature distribution diagram.

[0046] Furthermore, the feature collection module 11 in the system is also used for: The predetermined dimensions include concrete dimensions, environmental condition dimensions and construction parameter dimensions, wherein the concrete dimensions include material components and material ratios, the environmental condition dimensions include air temperature, humidity and wind speed, and the construction parameter dimensions include pouring speed and curing conditions.

[0047] Furthermore, the sensitivity analysis module 12 in the system is also used for: Taking multiple feature information of multiple features in the target feature information as independent variables; taking the first real-time temperature as a dependent variable; and analyzing a scatter plot drawn based on a mapping relationship between the independent variable and the dependent variable to obtain the first analysis result.

[0048] Furthermore, the sensitivity analysis module 12 in the system is also used for: Extract a first feature from the multiple features, and traverse the scatter plot to obtain a first scatter plot of the first feature; read a first grid strategy, and grid the first scatter plot based on the first grid strategy to obtain a first grid result; introduce a grid mutual information evaluation function to analyze and evaluate the first grid result to obtain a first mutual information value; adjust the first grid strategy to obtain a second grid strategy, and grid the first scatter plot based on the second grid strategy to obtain a second grid result; obtain a second mutual information value of the second grid result according to the grid mutual information evaluation function; when the second mutual information value is greater than the first mutual information value, use the second mutual information value as the first target mutual information value of the first scatter plot; descend the multiple target mutual information values ​​of the multiple features formed based on the first target mutual information value to obtain a descending list of target mutual information values; use the descending list of target mutual information values ​​as the first analysis result.

[0049] Furthermore, the sensitivity analysis module 12 in the system is also used for: The expression of the grid mutual information evaluation function is as follows: ;in, Characterizes the first of the multiple features The characteristics of Mutual information value, the The mutual information value refers to the The feature corresponding to Feature information With the first real-time temperature The mutual information of Characterize the KL divergence, Characterizing the Feature information With the first real-time temperature The joint probability distribution of Characterizing the Feature information With the first real-time temperature The product of the marginal probability distributions of A first strategy reliability characterizing the first grid strategy.

[0050] Furthermore, the sensitivity analysis module 12 in the system is also used for: Read the predetermined strategy characteristics, and collect the characteristics of the first grid strategy based on the predetermined strategy characteristics to obtain the first grid characteristic parameters; weight the normalized first grid characteristic parameters to obtain the first strategy reliability; wherein the predetermined strategy characteristics include the total number of strategy grids and the average strategy grid granularity.

[0051] Furthermore, the system further comprises a distribution map adjustment module, wherein the distribution map adjustment module is used to: Extract a first temperature position to be inserted in the visualized temperature distribution map; obtain a first search neighborhood of the first temperature position to be inserted based on a predetermined neighborhood scheme, the first search neighborhood including multiple neighborhood positions with predicted temperature identifiers; obtain a first neighborhood predicted temperature of a first neighborhood position among the multiple neighborhood positions with predicted temperature identifiers; use a first spatial distance between the first neighborhood position and the first temperature position to be inserted as a feedback adjustment coefficient to weightedly adjust the first neighborhood predicted temperature to obtain a first temperature to be inserted; render the first temperature to be inserted to the first temperature position to be inserted in the visualized temperature distribution map.

[0052] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. Figure 1 The temperature change distribution prediction method and specific examples for concrete temperature field in the first embodiment are also applicable to the temperature change distribution prediction system for concrete temperature field in the present embodiment. Through the above detailed description of the temperature change distribution prediction method for concrete temperature field, those skilled in the art can clearly know the temperature change distribution prediction system for concrete temperature field in the present embodiment, so for the sake of brevity of the specification, it will not be described in detail here. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description.

[0053] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

[0054] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application belong to the scope of the present application and its equivalent technology, the present application is also intended to include these modifications and variations.

Claims

1. A method for predicting temperature change distribution of concrete temperature field, characterized in that: include: Reading a predetermined dimension, and collecting multi-dimensional features of a first unit block of concrete based on the predetermined dimension to obtain target feature information; Performing a sensitivity analysis on the target characteristic information and a first real-time temperature of the first unit block to obtain a first analysis result; According to the first analysis result, a predetermined key feature set is formed, and based on the predetermined key feature set, dynamic feature collection is performed on the second unit block to obtain second key feature information; The second key feature information is used as input information of a temperature prediction model to obtain a second predicted temperature, wherein the temperature prediction model is an intelligent model obtained by machine learning based on the neural network principle on data in historical concrete temperature records; The second predicted temperature is combined with the second position of the second unit block to generate a visual temperature distribution map; The temperature change of the concrete is visually monitored according to the visual temperature distribution diagram.

2. The temperature change distribution prediction method for concrete temperature field according to claim 1, characterized in that: The predetermined dimensions include concrete dimensions, environmental condition dimensions and construction parameter dimensions, wherein the concrete dimensions include material components and material ratios, the environmental condition dimensions include air temperature, humidity and wind speed, and the construction parameter dimensions include pouring speed and curing conditions.

3. The temperature change distribution prediction method for concrete temperature field according to claim 1, characterized in that: include: Taking multiple feature information of multiple features in the target feature information as independent variables; Taking the first real-time temperature as a dependent variable; The first analysis result is obtained by analyzing a scatter plot drawn based on a mapping relationship between the independent variable and the dependent variable.

4. The temperature change distribution prediction method for concrete temperature field according to claim 3 is characterized in that: include: Extracting a first feature from the multiple features, and traversing the scatter plot to obtain a first scatter plot of the first feature; Reading a first grid strategy, and performing gridding processing on the first scatter plot based on the first grid strategy to obtain a first gridding result; Introducing a grid mutual information evaluation function to analyze and evaluate the first gridding result to obtain a first mutual information value; Adjusting the first grid strategy to obtain a second grid strategy, and performing gridding processing on the first scatter plot based on the second grid strategy to obtain a second gridding result; Obtaining a second mutual information value of the second gridding result according to the grid mutual information evaluation function; When the second mutual information value is greater than the first mutual information value, using the second mutual information value as a first target mutual information value of the first scatter plot; Descending the target mutual information values ​​of the multiple features formed based on the first target mutual information value to obtain a descending list of target mutual information values; The target mutual information values ​​are listed in descending order as the first analysis result.

5. The temperature change distribution prediction method for concrete temperature field according to claim 4 is characterized in that: The expression of the grid mutual information evaluation function is as follows: ; in, Characterizes the first of the multiple features The characteristics of Mutual information value, the The mutual information value refers to the The feature corresponding to Feature information With the first real-time temperature The mutual information of Characterize the KL divergence, Characterizing the Feature information With the first real-time temperature The joint probability distribution of Characterizing the Feature information With the first real-time temperature The product of the marginal probability distributions of A first strategy reliability characterizing the first grid strategy.

6. The temperature change distribution prediction method for concrete temperature field according to claim 5, characterized in that: include: Reading a predetermined strategy feature, and collecting features of the first grid strategy based on the predetermined strategy feature to obtain a first grid feature parameter; Weighting the normalized first grid feature parameters to obtain the first strategy reliability; The predetermined strategy features include the total number of strategy grids and the average strategy grid granularity.

7. The temperature change distribution prediction method for concrete temperature field according to claim 1, characterized in that: Also includes: Extracting a first temperature position to be inserted from the visualized temperature distribution map; Acquire a first search neighborhood of the first temperature position to be inserted based on a predetermined neighborhood scheme, wherein the first search neighborhood includes a plurality of neighborhood positions having predicted temperature identifiers; Obtaining a first neighborhood predicted temperature of a first neighborhood position among the plurality of neighborhood positions having predicted temperature identifiers; Taking a first spatial distance between the first neighborhood position and the first temperature position to be inserted as a feedback adjustment coefficient, weighted adjustment is performed on the first neighborhood predicted temperature to obtain a first temperature to be inserted; The first temperature to be inserted is rendered to the first temperature to be inserted position in the visualized temperature distribution map.

8. A temperature change distribution prediction system for concrete temperature field, characterized in that: The temperature change distribution prediction system for concrete temperature field is used to perform the steps of the temperature change distribution prediction method for concrete temperature field according to any one of claims 1 to 7, and the temperature change distribution prediction system for concrete temperature field comprises: A feature collection module, the feature collection module is used to read a predetermined dimension, and based on the predetermined dimension, perform multi-dimensional feature collection on the first unit block of concrete to obtain target feature information; A sensitivity analysis module, the sensitivity analysis module is used to perform a sensitivity analysis on the target characteristic information and the first real-time temperature of the first unit block to obtain a first analysis result; A key feature collection module, the key feature collection module is used to form a predetermined key feature set according to the first analysis result, and perform dynamic feature collection on the second unit block based on the predetermined key feature set to obtain second key feature information; An intelligent prediction module, the intelligent prediction module is used to use the second key feature information as input information of a temperature prediction model to obtain a second predicted temperature, the temperature prediction model is an intelligent model obtained by machine learning of data in historical concrete temperature records based on the principle of a neural network; A distribution map generating module, the distribution map generating module is used to generate a visual temperature distribution map in combination with the second predicted temperature and the second position of the second unit block; A visualization module is used to visually monitor the temperature change of the concrete according to the visualized temperature distribution diagram.