Mass concrete intelligent temperature control system based on neural network and PID control

Abstract

The invention belongs to the technical field of concrete temperature regulation and control, and particularly relates to a mass concrete intelligent temperature control system based on a neural network and PID control, and the system comprises a temperature obtaining module which is used for obtaining temperature data in mass concrete; the data processing module is used for processing the temperature data; the temperature prediction module is used for predicting the temperature change trend of the mass concrete; the PID control module is used for calculating the temperature deviation between the real-time temperature data and the target temperature data according to the real-time temperature data and generating a cooling control signal through a PID algorithm; the PID control module is further used for judging whether the current PID control parameter meets a preset requirement or not according to the temperature change trend predicted by the temperature prediction module, and automatically adjusting the PID control parameter when the judgment result is that the current PID control parameter does not meet the preset requirement; and the cooling execution module is used for adjusting the cooling system to work according to the cooling control signal. According to the invention, the temperature of the large-volume concrete can be effectively controlled in the pouring process.

Description

Technical Field

[0001] The invention belongs to the technical field of concrete temperature control, and in particular relates to a large volume concrete intelligent temperature control system based on neural network and PID control. Background Art

[0002] During the pouring of large-volume concrete, the hydration reaction of cement will generate a lot of heat, causing the temperature inside the concrete to rise. Since large-volume concrete has poor thermal conductivity, temperature cracks are prone to occur when the temperature difference between the inside and outside is large, affecting the safety of the structure.

[0003] To solve this potential safety hazard, it is necessary to control the temperature of large-volume concrete. Traditional temperature control methods mainly rely on manual experience or simple PID control. However, due to the following characteristics of the pouring process of large-volume concrete, it has complex temperature change characteristics, and traditional temperature control methods are difficult to achieve ideal results.

[0004] First, the temperature change inside the concrete is not a simple linear relationship, but is affected by many factors, such as the rate of cement hydration reaction, ambient temperature, differences in thermal conductivity inside the concrete, etc.; and internal processes such as cement hydration reaction are nonlinear, which causes the temperature change of concrete to also show nonlinear characteristics.

[0005] Second, the temperature change of concrete is affected by many factors, and there are often complex interactions and coupling relationships between these factors, making it difficult to describe the temperature change with a simple mathematical model. During the pouring process, due to changes in the pouring speed, vibration method, environmental conditions (such as wind speed, solar radiation, etc.) of concrete, and the evolution of the internal microstructure of concrete, the temperature change shows great variability. The uncertainty of external factors such as environmental conditions and pouring methods makes it difficult to accurately predict and control the temperature change of concrete.

[0006] Third, due to the large size of mass concrete, the temperature changes in different parts may be significantly different, that is, the temperature distribution is uneven. This uneven temporal and spatial distribution further increases the complexity of temperature control.

[0007] Based on the above reasons, compared with conventional temperature control objects, large-volume concrete has the following problems: First, it is very difficult to control. Due to the complexity of temperature changes, conventional temperature control methods (such as manual experience or simple PID control) are difficult to achieve precise control of concrete temperature. Second, the control effect is poor; due to the difficulty in accurately predicting and responding to temperature changes, traditional control methods often lead to large temperature fluctuations, making it difficult to achieve ideal control effects. Third, the structural safety risk increases; failure or inadequacy of temperature control may cause temperature cracks inside the concrete, which in turn affects the safety and durability of the structure. Therefore, conventional temperature control methods are difficult to effectively control the temperature of large-volume concrete during the pouring process.

[0008] In summary, how to effectively control the temperature of large volumes of concrete during the pouring process has become a problem that needs to be solved urgently. Summary of the invention

[0009] In view of the above-mentioned deficiencies in the prior art, the present invention provides a large-volume concrete intelligent temperature control system based on neural network and PID control, which can effectively control the temperature of large-volume concrete during the pouring process.

[0010] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0011] The intelligent temperature control system for large volume concrete based on neural network and PID control includes:

[0012] Temperature acquisition module, used to obtain temperature data inside large volumes of concrete in real time;

[0013] A data processing module, used for processing the acquired temperature data;

[0014] The temperature prediction module is used to use historical temperature data and real-time temperature data to predict the future temperature change trend of mass concrete through a preset prediction model; the prediction model is a neural network model;

[0015] The PID control module is used to calculate the temperature deviation from the target temperature data based on the real-time temperature data, and generate a cooling control signal through the PID algorithm; the PID control module is also used to determine whether the current PID control parameters meet the preset requirements in combination with the temperature change trend predicted by the temperature prediction module, and automatically adjust the PID control parameters when the judgment result is that the preset requirements are not met;

[0016] The cooling execution module is used to adjust the operation of the cooling system according to the cooling control signal.

[0017] Compared with the prior art, the present invention has the following beneficial effects:

[0018] 1. After using this system to obtain the temperature data inside the large volume of concrete and process it, the following two coordinated operations will be performed. On the one hand, the PID control module calculates the temperature deviation from the target temperature data based on the real-time temperature data, and generates a cooling control signal through the PID algorithm, and then the cooling execution module executes the cooling control signal. In this way, the internal data of the concrete can always change towards the target temperature. On the other hand, the temperature prediction module will use the historical temperature data and the real-time temperature data to predict the future temperature change trend of the large volume of concrete. This temperature change trend is the temperature prediction trend that may appear if the PID algorithm is continued without changing the PID control parameters. After that, the PID control module will judge whether the current PID control parameters are appropriate (for example, whether the temperature change has achieved the preset effect, and whether the deviation from the target temperature is within the preset range) in combination with the future temperature change trend, and adjust the PID control parameters (such as proportional gain, integral gain and differential gain) when the judgment result is inappropriate. In this way, when the prediction result is that the temperature change trend is not ideal without changing the PID control parameters, the subsequent temperature can be made more ideal by optimizing and adjusting the PID control parameters. This system can more accurately predict the future temperature change trend of concrete and adjust the control parameters accordingly, so as to achieve precise control of concrete temperature. This helps to reduce temperature fluctuations, improve the uniformity of temperature inside concrete, and reduce the risk of temperature cracks.

[0019] By combining "using predicted temperature trends to adjust PID control parameters" with "PID algorithm for real-time temperature control", this system has better foresight and adaptability than conventional temperature control methods, and can effectively improve the accuracy of temperature control and ensure the effectiveness of internal temperature control of large volumes of concrete as much as possible.

[0020] 2. In this method, a neural network model is used as a temperature prediction model to predict the future temperature change trend of mass concrete. Neural network is an algorithm that can extract features from complex data through learning and training. It has good nonlinear mapping ability and is suitable for processing complex temperature fields. The intelligent temperature control system combined with PID control can control the temperature of mass concrete in real time and accurately in a complex and changeable environment, thereby effectively preventing the occurrence of temperature cracks.

[0021] In summary, the present invention can effectively control the temperature of a large volume of concrete during the pouring process.

[0022] Preferably, the prediction model is a neural network model based on a CNN-LSTM combination; the prediction model is trained using historical temperature data as training sample data, and the real-time temperature data is used as input of the trained prediction model to predict the future temperature change trend of the mass concrete;

[0023] The neural network model based on the CNN-LSTM combination includes a convolutional layer, a time series processing layer and a fully connected layer;

[0024] The convolutional layer is used to extract the spatial features of the temperature data to learn the association between temperature sensors at different locations;

[0025] The time series processing layer is used to process the time series information of the spatial features of the extracted temperature data, extract the time series features of the temperature data, and further analyze the change pattern of the temperature data over time;

[0026] The fully connected layer is used to integrate the spatial features of the temperature data extracted by the convolutional layer and the temporal features of the temperature data extracted by the time series processing layer, and predict the output temperature change trend prediction results.

[0027] Compared with other neural network models, the neural network model based on the CNN-LSTM combination can handle complex nonlinear relationships and adapt to different working conditions and temperature change environments. This allows the system to maintain stable and reliable prediction performance in the face of complex situations. The neural network model based on the CNN-LSTM combination can not only capture the spatial distribution characteristics of the concrete temperature field, but also handle dynamic changes in time at the same time to improve prediction accuracy. This can provide effective forward-looking information for the PID control module.

[0028] Preferably, in the process of training the prediction model, the mean square error MSE is used as the loss function of the training:

[0029]

[0030] Among them, y i is the actual temperature, y′ i is the predicted temperature, n is the number of data points; the smaller the MSE, the closer the predicted result is to the true temperature.

[0031] With this setup, the MSE loss function can accurately measure the error of the prediction model by calculating the average of the squares of the differences between the actual temperature and the predicted temperature. This error measurement is particularly important for temperature prediction, because even small temperature differences can have a significant impact on the performance of concrete.

[0032] In addition, the MSE loss function imposes a heavier penalty on outliers (i.e., predicted values ​​that are very different from the actual temperature), which helps the model avoid paying too much attention to outliers during training, thereby reducing the risk of overfitting.

[0033] Preferably, when the PID control module automatically adjusts the PID control parameters, the adjusted PID control parameters include proportional gain, integral gain and differential gain.

[0034] With this setting, the PID control module can significantly improve the response speed, stability and energy consumption performance of the temperature control system when automatically adjusting the proportional gain, integral gain and differential gain, while improving the construction quality and efficiency. This is of great significance for the temperature control system of large-volume concrete.

[0035] Preferably, when the cooling execution module adjusts the operation of the cooling system, the adjusted contents include cooling water flow, water temperature, and spray frequency.

[0036] Such a setting can, 1. optimize energy consumption and resource utilization. By adjusting the flow rate and water temperature of the cooling water at the same time, the cooling execution module can optimize energy consumption while ensuring the temperature control effect. For example, when the temperature of the concrete is high, it can be cooled quickly by increasing the flow rate and lowering the water temperature; when the temperature is close to the target value, energy consumption can be reduced by reducing the flow rate and increasing the water temperature. Spray cooling is an efficient cooling method. By adjusting the spray frequency, the cooling speed and energy consumption can be further controlled. When rapid cooling is required, the spray frequency can be increased; after the temperature stabilizes, the spray frequency can be reduced to reduce water waste and energy consumption.

[0037] 2. It can enhance the adaptability and flexibility of the system. The cooling execution module can automatically adjust the cooling parameters according to different working conditions and temperature changes. This enables the system to maintain a stable temperature control effect in various environments and enhances the adaptability of the system. In addition, various emergencies may occur during the construction process, such as weather changes, changes in material properties, etc. The cooling execution module can flexibly respond to these situations and maintain the stability of the concrete temperature by adjusting the cooling parameters.

[0038] Preferably, it also includes a user interface module for displaying temperature data, prediction results and the operating status of the cooling system in real time through a graphical interface; and for users to perform manual adjustments and interventions.

[0039] With this setting, the user interface module can display the temperature data inside the concrete in real time, including the current temperature, historical temperature change trends, etc., so that users can intuitively understand the temperature conditions of the concrete. Through the graphical interface, users can clearly see the temperature prediction results, including the temperature change trend and predicted value in the future. This helps users understand the temperature conditions in advance and make corresponding decisions. The user interface module can also display the operating status of the cooling system, including parameters such as cooling water flow, water temperature, and spray frequency, so that users can monitor the operation of the system in real time.

[0040] The user interface module provides important support and guarantee for the intelligent temperature control system of large-volume concrete by improving information transparency and visualization, enhancing user interaction and operation convenience, improving system reliability and safety, and optimizing construction efficiency and quality.

[0041] Preferably, the temperature acquisition module includes a plurality of temperature sensors, which are respectively arranged at key positions at different depths of the bulk concrete; the key positions include areas where high temperature may occur and areas with large temperature gradients.

[0042] Such a setting can, 1. accurately locate high-temperature areas. By placing temperature sensors in areas where high temperatures may occur, the system can capture temperature changes in these areas in real time, thereby ensuring accurate monitoring of high-temperature areas. This helps to promptly discover and deal with potential temperature problems and prevent quality problems such as temperature cracks. It can also fully cover the temperature gradient area. By placing multiple temperature sensors in areas with large temperature gradients, comprehensive monitoring of the temperature gradient inside the concrete can be achieved. This helps to understand the temperature distribution inside the concrete and provide data support for formulating effective temperature control strategies.

[0043] 2. It can enhance the reliability and stability of the system. Since the temperature acquisition module can capture the temperature changes inside the concrete in real time, the system can respond to abnormal temperature events more quickly. This helps reduce construction delays and quality problems caused by temperature problems. Through precise temperature monitoring and data analysis, the system can detect and solve potential temperature problems in a timely manner, thereby reducing equipment damage and maintenance costs caused by temperature problems.

[0044] Preferably, when setting the temperature sensors of the temperature acquisition module, a plurality of sensors are arranged at key locations of the bulk concrete.

[0045] Such a setting can 1. improve the accuracy and comprehensiveness of temperature monitoring. By arranging multiple sensors at key locations, accurate monitoring of the internal temperature of large volumes of concrete can be achieved. These locations are usually the areas with the most drastic or representative temperature changes, so the arrangement of sensors can capture more accurate and comprehensive temperature data. This helps to detect temperature anomalies in a timely manner and provide reliable data support for subsequent temperature control measures.

[0046] 2. It can enhance the reliability and stability of the system. Placing multiple sensors at key locations can improve the reliability and stability of the temperature monitoring system. Even if one or several sensors fail or have errors, other sensors can still provide accurate data to ensure the normal operation of the system. Redundant sensors can provide accurate alternative data, thereby improving the fault tolerance of the system. This helps to avoid problems such as temperature monitoring failure or false alarms caused by sensor failure.

[0047] Preferably, when the data processing module processes the temperature data, the processing process includes filtering, noise reduction, data correction and data integration; wherein the data integration includes spatial integration and temporal integration.

[0048] Such a setting can improve the accuracy and reliability of data and reduce the impact of errors and interference. It can also form a comprehensive temperature distribution map and time series data to provide data support for spatial analysis and time series analysis. In addition, it can also optimize the data processing process, improve data processing efficiency and automation. Moreover, it can provide more abundant and accurate data resources for subsequent data analysis and application.

[0049] Preferably, the spatial integration includes spatially integrating sensor data from different positions and depths, generating a two-dimensional or three-dimensional scene of temperature distribution through data processing, and forming an overall temperature distribution model inside the concrete; the data processing includes interpolation and smoothing processing;

[0050] The content of the time integration includes, for the temperature change data within a preset short period of time, integrating the temperature data at multiple moments through time series analysis, and extracting the key temperature change trend.

[0051] This setting can generate an overall temperature distribution model inside the concrete by spatially integrating sensor data from different locations and depths. This model can intuitively display the temperature distribution inside the concrete, and help to find abnormal temperature areas and temperature gradient changes. It can improve data accuracy and reliability. The interpolation and smoothing used in the spatial integration process can reduce errors and noise in the data, and improve data accuracy and reliability. This helps to more accurately evaluate the temperature state inside the concrete and provide reliable data support for subsequent temperature control measures.

[0052] Through time series analysis, the temperature change data within a preset short period of time can be integrated in time to extract key temperature change trends. This helps to discover the laws and trends of temperature changes and provide predictions and early warnings for subsequent temperature control measures. Time integration technology can monitor and analyze the changing trends of temperature data in real time, providing data support for the formulation and optimization of temperature control strategies. This helps to adjust temperature control measures in a timely manner to ensure that the temperature inside the concrete is always kept within a reasonable range. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to make the purpose, technical solution and advantages of the invention more clear, the present invention will be further described in detail below with reference to the accompanying drawings, in which:

[0054] Figure 1 It is a structural schematic diagram in the embodiment;

[0055] Figure 2 The temperature control measured curve of the hydration heat of large volume concrete using this system is shown in the figure. DETAILED DESCRIPTION

[0056] The following is a further detailed description through specific implementation methods:

[0057] Example:

[0058] like Figure 1 As shown, this embodiment discloses a large volume concrete intelligent temperature control system based on neural network and PID control, including a temperature acquisition module, a data processing module, a temperature prediction module, a PID control module, a cooling execution module and a user interface module.

[0059] The temperature acquisition module is used to obtain the temperature data inside the bulk concrete in real time. The temperature acquisition module includes multiple temperature sensors, which are arranged at key positions at different depths (such as the surface, middle and bottom) of the bulk concrete; the key positions include areas where high temperatures may occur and areas with large temperature gradients (such as piers and beam connections).

[0060] Such a setting can accurately locate high-temperature areas. By arranging temperature sensors in areas where high temperatures may occur, the system can capture temperature changes in these areas in real time, thereby ensuring accurate monitoring of high-temperature areas. This helps to promptly detect and deal with potential temperature problems and prevent quality problems such as temperature cracks. It can also fully cover the temperature gradient area. By arranging multiple temperature sensors in areas with large temperature gradients, comprehensive monitoring of the temperature gradient inside the concrete can be achieved. This helps to understand the temperature distribution inside the concrete and provide data support for the formulation of effective temperature control strategies. In addition, the reliability and stability of the system can be enhanced. Since the temperature acquisition module can capture the temperature changes inside the concrete in real time, the system can respond to temperature abnormalities more quickly. This helps to reduce construction delays and quality problems caused by temperature problems. Through precise temperature monitoring and data analysis, the system can promptly detect and solve potential temperature problems, thereby reducing equipment damage and maintenance costs caused by temperature problems.

[0061] In a specific implementation, when setting the temperature sensors of the temperature acquisition module, multiple sensors are arranged at key locations of the bulk concrete.

[0062] In this way, the accuracy and comprehensiveness of temperature monitoring can be improved. By arranging multiple sensors at key key locations, accurate monitoring of the internal temperature of large volumes of concrete can be achieved. These locations are usually the areas with the most drastic or representative temperature changes, so the arrangement of sensors can capture more accurate and comprehensive temperature data. This helps to detect temperature anomalies in a timely manner and provide reliable data support for subsequent temperature control measures. It can also enhance the reliability and stability of the system. Arranging multiple sensors at key key locations can improve the reliability and stability of the temperature monitoring system. Even if one or several sensors fail or have errors, other sensors can still provide accurate data to ensure the normal operation of the system. Redundant sensors can provide accurate alternative data, thereby improving the fault tolerance of the system. This helps to avoid problems such as temperature monitoring failure or false alarms caused by sensor failures.

[0063] About data sampling and frequency

[0064] In order to capture the dynamic changes in the temperature inside the concrete, the temperature sensing module needs to set a suitable data sampling frequency. The specific sampling frequency should be adjusted according to the rate of concrete hydration reaction and the characteristics of temperature change. Generally, the temperature sampling frequency is set to collect data every two hours to balance the needs of data volume and real-time performance.

[0065] About sensor calibration and maintenance

[0066] To ensure the accuracy and reliability of temperature sensors, regular calibration and maintenance are required:

[0067] Initial calibration: Before installing the sensor, perform calibration under a standard temperature source to ensure that the temperature signal output by the sensor is consistent with the actual temperature.

[0068] Regular calibration: During the construction process, the sensor should be calibrated regularly to correct the measurement errors caused by environmental changes or sensor aging.

[0069] Fault detection and replacement: Through redundant design and data consistency check, the working status of the sensor is monitored in real time, and the faulty sensor is discovered and replaced in time to ensure the continuous and reliable operation of the system.

[0070] The data processing module is used to process the acquired temperature data. When the data processing module processes the temperature data, the processing process includes filtering, noise reduction, data correction and data integration; among them, data integration includes spatial integration and temporal integration. Such a setting can improve the accuracy and reliability of the data and reduce the impact of errors and interference. It can also form a comprehensive temperature distribution map and time series data to provide data support for spatial analysis and time series analysis. In addition, it can also optimize the data processing process and improve the efficiency and automation of data processing. In addition, it can provide more abundant and accurate data resources for subsequent data analysis and application.

[0071] In specific implementation, the content of spatial integration includes spatial integration of sensor data from different positions and depths, generating a two-dimensional or three-dimensional scene of temperature distribution through data processing, and forming an overall temperature distribution model inside the concrete; the data processing includes interpolation and smoothing. In this way, by spatially integrating sensor data from different positions and depths, an overall temperature distribution model inside the concrete can be generated. This model can intuitively display the temperature distribution inside the concrete, and help to discover abnormal temperature areas and temperature gradient changes. The data accuracy and reliability can be improved. The interpolation and smoothing used in the spatial integration process can reduce errors and noise in the data and improve the accuracy and reliability of the data. This helps to more accurately evaluate the temperature state inside the concrete and provide reliable data support for subsequent temperature control measures.

[0072] Time integration includes, for temperature change data within a preset short period of time, integrating temperature data at multiple times through time series analysis, and extracting key temperature change trends. Through time series analysis, temperature change data within a preset short period of time can be integrated to extract key temperature change trends. This helps to discover the laws and trends of temperature changes and provide predictions and early warnings for subsequent temperature control measures. Time integration technology can monitor and analyze the changing trends of temperature data in real time, providing data support for the formulation and optimization of temperature control strategies. This helps to adjust temperature control measures in a timely manner to ensure that the temperature inside the concrete is always kept within a reasonable range.

[0073] In order to ensure that the neural network can efficiently learn and predict the data, the collected temperature data needs to go through a series of preprocessing steps. These steps include filtering, noise reduction, data correction and integration, etc. The specific steps are as follows:

[0074] Data filtering

[0075] During the data collection process, the sensor may be affected by environmental noise or other interference (such as electromagnetic interference, mechanical vibration, etc.), resulting in inaccurate measurement values. In order to improve the accuracy of the data, the data preprocessing module will use filtering to remove noise.

[0076] Low-pass filter: Filter out high-frequency noise and retain low-frequency changes in temperature. Temperature changes are usually slow, and low-pass filtering can effectively eliminate sudden temperature spikes or interference. The function of filtering is to eliminate noise in the data and retain useful signals. In temperature data processing, filters can effectively remove high-frequency noise, smooth the temperature change curve, and make the data more accurate. In the temperature control system of large-volume concrete, the commonly used filtering method is low-pass filter.

[0077] How Low Pass Filters Work

[0078] A low-pass filter is a filter that allows low-frequency signals to pass through while suppressing high-frequency noise. It is suitable for removing high-frequency fluctuations and noise in temperature data. Low-pass filters smooth temperature curves and are suitable for systems that process relatively slow temperature changes, such as mass concrete temperature control.

[0079] Working principle: Frequency range division: The low-pass filter divides the signal into low-frequency and high-frequency parts based on the frequency characteristics of the signal. The low-frequency part usually represents the real temperature change, while the high-frequency part is noise. Blocking high-frequency noise: The low-pass filter allows low-frequency signals to pass through while attenuating or completely blocking high-frequency noise. Smoothing signals: Since temperature changes slowly, the temperature signal itself is a low-frequency signal. The low-pass filter smoothes the temperature change curve by weakening the high-frequency components and eliminating sudden high-frequency fluctuations. For discrete temperature data, the low-pass filter can be implemented through exponential weighted averaging or simple moving average:

[0080] y t =α·x t +(1-α);y t-1

[0081] Among them, y t is the output signal after filtering; x t is the input temperature data at the current moment; α is the smoothing coefficient, which controls the weight of the low-frequency signal; y t-1 is the filter output at the previous moment.

[0082] Data Noise Reduction

[0083] Noise identification and elimination: By statistically analyzing data fluctuations and identifying unreasonable data jumps or fluctuations, machine learning algorithms can more intelligently identify abnormal data points and handle noise through pattern recognition and training data models. Common methods include anomaly detection algorithms. Anomaly detection algorithms learn the normal distribution of data and automatically identify data points that differ greatly from the normal data distribution. They consider these points to be abnormal, thereby filtering out noise.

[0084] Distance-based anomaly detection: such as the k-nearest neighbor algorithm (KNN), which removes anomalies that are far away from most data points by calculating the distance between data points. The K-nearest neighbor algorithm determines whether a data point is an anomaly by calculating the distance between a data point and its k closest neighbors. Generally speaking, if a data point is far away from its neighbors, it may be an anomaly.

[0085] Algorithm steps:

[0086] Select parameter k: Set a positive integer k, indicating that we want to find the first k nearest neighbors of the data point.

[0087] Calculate distance: For each data point, calculate the distance between it and other data points. Commonly used distance metrics include Euclidean distance.

[0088] Euclidean distance formula:

[0089]

[0090] Among them, x i and x j are two data points; n is the dimension of the data point; m represents the mth feature dimension of the data point.

[0091] Find the k nearest neighbors: For each data point, find the k nearest neighbors to it.

[0092] Calculate anomaly score: measure the abnormality of a data point by calculating the average distance or maximum distance between it and its k nearest neighbors. Usually, points with larger distance are considered as outliers.

[0093] Anomaly score formula:

[0094]

[0095] Among them, d(x i ,x j ) is the data point x i and its neighbor x j The distance between them, k is the number of neighbors.

[0096] Set threshold: Set a threshold T. If the abnormal score S(x i ) exceeds the threshold, the point is considered to be an outlier.

[0097] If S(x i )>T,

[0098] Then x i It is an outlier that needs to be removed or corrected.

[0099] Data calibration

[0100] Sensor bias calibration: Sensors may have systematic data deviations due to temperature drift, sensor aging, etc. The system automatically calibrates the sensor's measurements by comparing with a standard temperature source or using multiple sensors for cross-checking to ensure long-term data accuracy.

[0101] Time synchronization correction: In order to ensure the comparability of data from different sensors at the same time point, the system will perform time synchronization correction on each data point to ensure that the sampling time of all sensors is consistent.

[0102] Data Integration

[0103] Spatial integration: The sensor data from different locations and depths are spatially integrated to form an overall temperature distribution model inside the concrete. The system can generate a two-dimensional or three-dimensional scene of temperature distribution through interpolation, smoothing and other processing.

[0104] Time integration: For temperature change data in a short period of time, time series analysis is used to integrate temperature data at multiple times and extract key temperature change trends.

[0105] The processed temperature data needs to be transmitted to the neural network temperature prediction module or central control system through the communication system for further analysis and processing. The data transmission process should ensure the integrity and real-time nature of the data.

[0106] The temperature prediction module is used to use historical temperature data and real-time temperature data to predict the future temperature change trend of mass concrete through a preset prediction model.

[0107] Among them, the prediction model is a neural network model. In specific implementation, the prediction model is a neural network model based on the CNN-LSTM combination. Compared with other neural network models, the neural network model based on the CNN-LSTM combination can handle complex nonlinear relationships and adapt to different working conditions and temperature change environments. This enables the system to maintain stable and reliable prediction performance in the face of complex situations. The neural network model based on the CNN-LSTM combination can not only capture the spatial distribution characteristics of the concrete temperature field, but also simultaneously handle dynamic changes in time to improve prediction accuracy. This can provide effective forward-looking information for the PID control module.

[0108] The following is the specific working principle and process of the temperature prediction module:

[0109] 1. Data input and feature extraction

[0110] The input data of the neural network temperature prediction module includes real-time temperature data from the temperature sensing module and historical temperature data collected by the system, which are used to train the neural network model and make real-time predictions.

[0111] Data input type

[0112] Real-time temperature data: Current temperature values ​​from multiple sensors to capture real-time temperature distribution inside the concrete.

[0113] Historical temperature data: Previously collected temperature data, including temperature changes over the past hours, days, or longer, is used to train the neural network and help the model learn the patterns of temperature changes.

[0114] 2. Neural Network Model Selection

[0115] In the task of predicting the temperature of large-volume concrete, common neural network models include long short-term memory networks (LSTM) and convolutional neural networks (CNN). Convolutional neural networks (CNN) can also be combined with LSTM. These models are good at processing time series data and complex nonlinear problems.

[0116] Convolutional Neural Network (CNN) combined with LSTM

[0117] For complex temperature distribution problems that require consideration of both spatial and temporal information, CNN and LSTM can be used together to form a neural network model based on the CNN-LSTM combination as a prediction model. This prediction model uses historical temperature data as training sample data for training, and uses real-time temperature data as the input of the trained prediction model to predict the future temperature change trend of mass concrete.

[0118] Specifically, the neural network model based on the CNN-LSTM combination includes a convolutional layer CNN, a time series processing layer LSTM, and a fully connected layer;

[0119] The convolutional layer (CNN) is used to extract the spatial features of the temperature data to learn the association between temperature sensors at different locations;

[0120] The time series processing layer (LSTM) is used to process the time series information of the extracted spatial features of the temperature data and extract the temporal features of the temperature data to further analyze the change pattern of the temperature data over time;

[0121] The fully connected layer is used to integrate the spatial features of the temperature data extracted by the convolutional layer and the temporal features of the temperature data extracted by the time series processing layer, and predict the output temperature change trend prediction results.

[0122] This CNN-LSTM combined model can not only capture the spatial distribution characteristics of the concrete temperature field, but also simultaneously process the dynamic changes in time to improve the prediction accuracy.

[0123] Prediction model training

[0124] In order to ensure that the neural network can accurately predict the temperature change trend of concrete, the system needs to train the neural network with a large amount of historical data. Model training is usually divided into the following steps:

[0125] 3.1 Data Preparation

[0126] Segmentation of historical temperature data: historical temperature data is divided into training set, validation set and test set for model training, tuning and testing.

[0127] Normalization: In order to improve the convergence speed and prediction accuracy of the model, temperature data is usually normalized (for example, mapping the temperature value to the range of 0 to 1).

[0128] Sliding window mechanism: In the processing of time series data, the sliding window method is used to divide the temperature time series data into multiple time periods. Each window is used as a part of the input data, and the model is trained to learn the patterns in the time series.

[0129] 3.2 Loss Function and Optimization

[0130] Loss function: In the present invention, in the process of training the prediction model, the mean square error (MSE) is used as the loss function of the training to measure the difference between the model prediction value and the actual temperature value. The smaller the MSE, the closer the prediction result is to the actual temperature.

[0131] The expression of the mean square error MSE loss function is:

[0132]

[0133] Among them, y i is the actual temperature, y′ i is the predicted temperature, n is the number of data points; the smaller the MSE, the closer the predicted result is to the true temperature.

[0134] The MSE loss function can accurately measure the error of the prediction model by calculating the average of the squares of the difference between the actual temperature and the predicted temperature. This error measurement method is particularly important for temperature prediction, because even small temperature differences can have a significant impact on the performance of concrete. In addition, the MSE loss function has a heavier penalty for outliers (i.e., predicted values ​​that are very different from the actual temperature), which helps the model avoid paying too much attention to outliers during training, thereby reducing the risk of overfitting.

[0135] Training and validation of prediction models

[0136] Forward propagation and back propagation: In model training, temperature data is forward propagated through the neural network to obtain the prediction results. Then the network weights are adjusted through the back propagation algorithm to gradually reduce the prediction error.

[0137] Validation set evaluation: During the training process, the validation set is used to evaluate the model and detect the training effect of the model in real time to prevent the model from overfitting or underfitting.

[0138] Early stopping mechanism: To prevent overfitting caused by overtraining, the system can use an early stopping mechanism to automatically stop training when the loss on the validation set no longer decreases.

[0139] Temperature forecast and real-time updates

[0140] After the neural network model is trained, the system will use the latest temperature data collected by the sensor in real time and input it into the trained neural network to predict the temperature changes in the future. The temperature results predicted by the neural network will be passed to the PID control module to guide the adjustment of the cooling system.

[0141] Real-time prediction

[0142] Real-time data input: The real-time data provided by the temperature sensor module is input into the neural network, and the predicted value of future temperature is obtained through forward propagation.

[0143] Short-term and long-term predictions: Neural networks can not only predict temperature changes in a short period of time (such as minutes and hours), but can also make long-term predictions based on historical data to help formulate long-term temperature control strategies.

[0144] Real-time updating and model adaptation

[0145] In order to improve the system's adaptability to changes in the environment and conditions, the neural network can be adaptively updated:

[0146] Data update: The system regularly collects the latest temperature data and updates the model online, so that the neural network can adapt to temperature changes during concrete pouring at any time.

[0147] Model retraining: During the construction process, changes in external factors such as the environment and weather may affect the temperature distribution. The neural network can retrain the model based on the new data to maintain high-precision predictions.

[0148] Prediction result output and feedback

[0149] The output of the neural network prediction module is the temperature change trend over a period of time in the future, and these results will be used to guide the operation of the cooling system.

[0150] Output Format

[0151] Temperature curve: Output the temperature change curve for the future time period and display it on the user interface to help construction personnel understand the future temperature change trend.

[0152] Numerical prediction: Provides the future temperature value of each sensor location in the form of specific numerical values ​​for further analysis by the PID control module.

[0153] Feedback and model optimization

[0154] Evaluation of prediction results: The system will compare the prediction results with the actual temperature to evaluate the prediction accuracy of the model. If the error is large, the system will record the error and feed it back to the model for optimization and adjustment.

[0155] Online learning: In actual construction, neural networks can continuously adjust their prediction capabilities through online learning mechanisms to improve the system's adaptability to complex environments.

[0156] The PID control module is used to calculate the temperature deviation from the target temperature data based on the real-time temperature data, and generate a cooling control signal through the PID algorithm; the PID control module is also used to determine whether the current PID control parameters meet the preset requirements in combination with the temperature change trend predicted by the temperature prediction module, and automatically adjust the PID control parameters when the judgment result is that the preset requirements are not met.

[0157] Among them, when the PID control module automatically adjusts the PID control parameters, the adjusted PID control parameters include proportional gain, integral gain and differential gain. In this way, when the PID control module automatically adjusts the proportional gain, integral gain and differential gain, it can significantly improve the response speed, stability and energy consumption performance of the temperature control system, while improving the construction quality and efficiency. This is of great significance for the temperature control system of large-volume concrete.

[0158] The PID control module is responsible for adjusting the operating status of the cooling system in real time in the intelligent temperature control system of large-volume concrete based on neural network and PID control according to the temperature change trend predicted by the neural network and the actual measured temperature. The PID controller is based on the three adjustment mechanisms of proportion (P), integration (I) and differentiation (D). It calculates the temperature deviation and generates a control signal to achieve the precise temperature control target. The following are the specific working principles and operation steps of the PID control module:

[0159] 1. Control target setting

[0160] The basic goal of the PID control module is to ensure that the concrete temperature remains within a safe and reasonable range to avoid cracking problems caused by excessive temperature or excessive temperature gradient.

[0161] 1.1 Target temperature setting

[0162] Target temperature (SP): According to the construction requirements, the target temperature inside the mass concrete is set. The target temperature can be a fixed value or a temperature curve that changes dynamically according to the construction stage.

[0163] Allowable temperature range: In order to avoid frequent adjustment operations, a temperature tolerance range is usually set. As long as the temperature deviation is within this range, the PID controller will keep the existing cooling state unchanged.

[0164] 1.2 Input Signal

[0165] Real-time temperature signal: Real-time temperature data from the temperature sensor (also called process variable, PV).

[0166] Neural network prediction results: The temperature change trend predicted by the neural network in the future provides predictive information for PID control and helps the system adjust cooling measures.

[0167] 2. Temperature deviation calculation

[0168] The PID control module first calculates the temperature deviation (e), which is the difference between the currently measured temperature (PV) and the set target temperature (SP):

[0169] e(t)=SP-PV(t)

[0170] If the temperature deviation is positive (e>0), it means that the current temperature is lower than the target temperature and the system may need to reduce cooling.

[0171] If the temperature deviation is negative (e<0), it means that the current temperature is higher than the target temperature and the system needs to increase cooling efforts.

[0172] 3. The core working principle of PID controller

[0173] The PID controller generates a control signal based on the temperature deviation through three mechanisms: proportional control (P), integral control (I), and derivative control (D), which work together to ensure that the system reaches the target temperature quickly and accurately.

[0174] 3.1 Proportional control (P)

[0175] Proportional control adjusts the cooling system output according to the current temperature deviation. The role of proportional control is to quickly respond to the current deviation by linearly amplifying the temperature deviation value:

[0176] P=K i e(t)

[0177] Among them, K p is the proportional gain, which determines the response strength of the proportional controller to temperature deviation, and e(t) is the current temperature deviation.

[0178] 3.2 Integral Control (I)

[0179] Integral control eliminates long-term deviations by accumulating historical temperature deviations to ensure that the system eventually reaches the set target temperature. The role of integral control is to gradually correct long-term steady-state errors by accumulating errors when the system has long-term steady-state errors:

[0180]

[0181] K i It is the integral gain, which determines the response strength of the integral controller to the accumulated deviation. The integral term accumulates all past deviation values. When the system error persists, the integral control will gradually increase the control amount and eventually eliminate the error.

[0182] 3.3 Differential Control (D)

[0183] Derivative control adjusts based on the rate of change of the temperature deviation. It predicts the future trend of the deviation and adjusts the cooling system to prevent the temperature from changing too quickly or too slowly. The output of the derivative controller is proportional to the rate of change of the temperature deviation:

[0184]

[0185] K d is the differential gain, which determines the response strength of the differential controller to the rate of change of the error. The differential term can help the system reduce temperature fluctuations and enhance the stability of the system. The differential control is very effective in suppressing rapid temperature changes in the system (such as sudden temperature rise).

[0186] 3.4 Total control output

[0187] The total output of the PID controller is the superposition of the three control quantities of proportion, integration and differentiation:

[0188]

[0189] Among them, u(t) is the output of the controller, which is used to control the parameters of the cooling system, such as cooling water flow rate.

[0190] According to the specification, the temperature difference between the inside and outside of large-volume concrete shall not exceed 25°C. When the temperature difference between the inside and outside is greater than the set value, an alert will be generated, and the corresponding module will intervene to respond and generate a control signal (such as adjusting the water flow). Figure 2 As shown in the figure, this is the measured temperature control curve of the hydration heat of large volume concrete using this system, and the overall temperature control effect is good.

[0191] 4. Control output and cooling system execution

[0192] The output signal u(t) of the PID control module is transmitted to the cooling system execution module to adjust the operating parameters of the cooling equipment in real time.

[0193] 4.1 Control cooling water system

[0194] Cooling water flow regulation: PID control signal controls the water flow of the cooling water system, increases or decreases the cooling water flow, and quickly reduces or maintains the concrete temperature.

[0195] Cooling water temperature adjustment: According to specific needs, PID control can also adjust the temperature of cooling water to ensure a more precise cooling effect.

[0196] Spray frequency control: If a spray system is used, a PID control signal can adjust the frequency and intensity of the spray to provide precise cooling measures.

[0197] 5. Adaptive PID control

[0198] In order to cope with the complexity of concrete temperature changes during construction, the PID control module can use an adaptive PID control algorithm to dynamically adjust the PID parameters (K p , K i , K d ) to adapt to changes in environment and working conditions.

[0199] 5.1 Parameter online adjustment

[0200] Adaptive PID control dynamically adjusts the proportional, integral and differential parameters by analyzing the temperature change characteristics of the system in real time.

[0201] For example:

[0202] When the temperature changes rapidly, increase the differential gain K d , suppress temperature fluctuations in advance.

[0203] When the temperature deviation persists, increase the integral gain K. i , quickly eliminate steady-state errors.

[0204] 5.2 Intelligent Adjustment

[0205] The PID control module can combine the future temperature change trend predicted by the neural network to adjust the control parameters (such as proportional gain, integral gain and differential gain) to avoid temperature fluctuations. This combination makes the system more forward-looking and adaptable, and further improves the accuracy of temperature control.

[0206] The cooling execution module is used to adjust the operation of the cooling system according to the cooling control signal. In specific implementation, when the cooling execution module adjusts the operation of the cooling system, the adjustment content includes cooling water flow, water temperature, and spray frequency.

[0207] In this way, energy consumption and resource utilization can be optimized. By adjusting the flow rate and water temperature of cooling water at the same time, the cooling execution module can optimize energy consumption while ensuring the temperature control effect. For example, when the concrete temperature is high, the temperature can be quickly reduced by increasing the flow rate and lowering the water temperature; when the temperature is close to the target value, the energy consumption can be reduced by reducing the flow rate and increasing the water temperature. Spray cooling is an efficient cooling method. By adjusting the spray frequency, the cooling speed and energy consumption can be further controlled. When rapid cooling is required, the spray frequency can be increased; after the temperature is stable, the spray frequency can be reduced to reduce the waste of water resources and energy consumption. In addition, the adaptability and flexibility of the system can be enhanced. The cooling execution module can automatically adjust the cooling parameters according to different working conditions and temperature changes. This enables the system to maintain a stable temperature control effect in various environments and enhances the adaptability of the system. In addition, various emergencies may occur during the construction process, such as weather changes, changes in material properties, etc. The cooling execution module can flexibly respond to these situations and maintain the stability of the concrete temperature by adjusting the cooling parameters.

[0208] The cooling execution module is the execution unit in the intelligent temperature control system of large-volume concrete. It is responsible for actually operating the cooling equipment to adjust the temperature of the concrete according to the control signal sent by the PID control module. Its main task is to ensure that the temperature inside the concrete is maintained within a safe range by adjusting the cooling water system and the spray cooling system, thereby preventing cracks or other structural problems caused by excessive temperature or excessive temperature gradient. The following is the specific working principle and operation steps of the cooling execution module:

[0209] 1. Control signal input

[0210] The cooling execution module receives control signals (i.e., cooling control signals) from the PID control module. These signals are generated based on real-time temperature data and temperature deviations and are used to adjust the parameters of different cooling devices.

[0211] 1.1 Control signal content

[0212] Cooling water flow control signal: used to adjust the water flow in the cooling water system to adjust the cooling water input.

[0213] Cooling water temperature control signal: In some systems, the temperature of the cooling water can be controlled to further accurately control the cooling effect of the concrete.

[0214] Spray frequency and intensity control signal: In the spray cooling system, adjust the frequency and intensity of the spray to adjust the cooling rate.

[0215] 2. Composition and working mechanism of cooling system

[0216] The cooling execution module can operate a variety of cooling equipment, including cooling water systems and spray systems, and flexibly select or combine these equipment for temperature control according to actual conditions.

[0217] 2.1 Cooling water system

[0218] The cooling water system is a commonly used cooling method for temperature control of large-volume concrete. Cooling water is transported through cooling pipes arranged in the concrete to absorb the heat inside the concrete.

[0219] Working principle:

[0220] Cooling pipe layout: During the concrete pouring process, cooling water pipes are pre-arranged inside the concrete, usually in a spiral or grid layout to ensure that the cooling water can cover the key areas of the concrete.

[0221] Cooling water circulation: Cooling water flows in the pipes through a circulation pump, absorbs the heat generated inside the concrete, and takes the heat away through a heat exchanger or cooling tower.

[0222] Cooling water temperature adjustment: The PID control module can adjust the temperature of the cooling water according to the temperature change of the concrete. If the concrete temperature is too high, the temperature of the cooling water can be lowered to increase the cooling effect.

[0223] Control mechanism:

[0224] Flow regulating valve: The cooling water flow is controlled by the flow regulating valve. The PID control module controls the cooling water flow by adjusting the valve opening, thereby affecting the cooling effect.

[0225] Increase flow rate: When the concrete temperature is high, the PID controller increases the flow rate to enhance the cooling effect.

[0226] Reduce flow rate: When the concrete temperature is close to or below the target temperature, the flow rate is reduced to avoid overcooling.

[0227] Cooling water circulation pump: By adjusting the power or speed of the circulation pump, the flow rate of cooling water can be controlled to further improve the cooling efficiency.

[0228] 2.3 Spray cooling system

[0229] The spray cooling system cools the concrete surface by spraying water mist. The water mist absorbs heat during evaporation, thereby reducing the surface temperature of the concrete. The spray system is suitable for auxiliary cooling in high temperature environments to prevent cracking of the concrete surface.

[0230] Working principle:

[0231] Water mist cooling: The spray system sprays water mist evenly onto the concrete surface, and the evaporation of the water mist takes away the heat, achieving a rapid cooling effect.

[0232] Nozzle arrangement: Nozzles are usually arranged above the concrete surface, covering the concrete surface and ensuring uniform distribution of water mist.

[0233] Control mechanism:

[0234] Spray frequency and intensity adjustment: PID control signal adjusts the spray frequency and intensity. When more cooling is needed, the system increases the spray frequency and water mist volume; when the temperature tends to stabilize, the spray intensity will automatically decrease or stop.

[0235] Water pressure regulation: By adjusting the water pressure in the spray system, the density and coverage of the water mist can be controlled to further adjust the cooling effect.

[0236] 3. Control execution feedback mechanism

[0237] The cooling execution module not only performs cooling operations, but also ensures the accuracy and reliability of the cooling effect through a real-time feedback mechanism.

[0238] 3.1 Real-time monitoring and feedback

[0239] Sensor feedback: The cooling execution module will regularly obtain real-time temperature data from the temperature sensing module to monitor the actual temperature changes after the cooling equipment is operated.

[0240] Feedback adjustment: If the temperature change after cooling is executed does not achieve the expected effect, the cooling execution module will immediately feedback to the PID control module for secondary adjustment. For example, if the cooling effect is insufficient, the system will automatically increase the cooling intensity.

[0241] 3.2 Redundancy and Fault Handling

[0242] Redundant design: The cooling execution module is usually designed with redundant cooling equipment to ensure that when the main cooling equipment fails, the backup cooling system can be automatically started to prevent temperature out of control.

[0243] Fault detection: If a cooling device (such as a cooling water pump or sprayer) fails, the cooling execution module will immediately report the fault information and automatically switch to the backup system based on the feedback mechanism to ensure uninterrupted temperature control operation.

[0244] 4. Automatic and manual control modes

[0245] The cooling execution module supports automatic control mode and manual control mode, ensuring that the cooling equipment can be flexibly controlled in different construction stages and complex environments.

[0246] 4.1 Automatic control mode

[0247] In automatic control mode, the cooling execution module operates completely according to the signal sent by the PID control module without manual intervention.

[0248] The system adjusts the cooling water flow and spray frequency in real time to ensure accurate temperature control and fast response.

[0249] 4.2 Manual control mode

[0250] In manual mode, users can manually set the operating parameters of the cooling equipment through the control terminal interface. This mode is usually used for cooling equipment control in construction sites where special operations or emergency situations are required.

[0251] Users can directly adjust the cooling water flow and the operating status of the spray system to ensure that the temperature can still be effectively controlled in emergency situations.

[0252] 5. System optimization and intelligent control

[0253] Based on the basic operating process, the cooling execution module can also be optimized through intelligent control algorithms and historical data analysis.

[0254] 5.1 Run historical data analysis

[0255] The cooling execution module can record the operating history data of the cooling equipment, including operating parameters such as cooling water flow rate and spray frequency.

[0256] Through historical data analysis, the system can find the best cooling parameter settings and optimize future cooling execution strategies.

[0257] 5.2 Predictive Adjustment

[0258] Combined with the future temperature change trend provided by the neural network temperature prediction module, the cooling execution module can adjust the parameters of the cooling equipment to avoid excessively high or low temperatures.

[0259] For example, when it is predicted that the temperature will rise in the future, the cooling execution module can increase the cooling water flow rate to prevent the temperature from rising sharply.

[0260] 6. Overall control process

[0261] The following is the overall workflow of the cooling execution module:

[0262] Receive control signals: Receive control signals from the PID control module to determine the cooling equipment parameters that need to be adjusted (such as cooling water flow, etc.).

[0263] Perform cooling operations: adjust the operating status of the cooling system according to the control signal, including increasing or decreasing the cooling water flow rate and the operating frequency of the spray system.

[0264] Real-time feedback: Real-time monitoring of temperature changes to ensure that the cooling effect reaches the expected target. If the cooling effect does not meet expectations, the system will provide feedback and automatically adjust the cooling parameters.

[0265] Fault handling: If a cooling device fails, the system automatically switches to the backup cooling device to ensure uninterrupted temperature control.

[0266] Automatic and manual control: supports automatic adjustment and manual intervention. Users can switch control modes as needed to ensure flexible response to different construction needs.

[0267] The user interface module is used to display temperature data, prediction results and the operating status of the cooling system in real time through a graphical interface; it is also used for users to make manual adjustments and interventions.

[0268] With this setting, the user interface module can display the temperature data inside the concrete in real time, including the current temperature, historical temperature change trends, etc., so that users can intuitively understand the temperature conditions of the concrete. Through the graphical interface, users can clearly see the temperature prediction results, including the temperature change trends and predicted values ​​in the future. This helps users understand the temperature conditions in advance and make corresponding decisions. The user interface module can also display the operating status of the cooling system, including parameters such as cooling water flow, water temperature, and spray frequency, so that users can monitor the operation of the system in real time. In addition, the user interface module provides important support and guarantee for the intelligent temperature control system of large-volume concrete by improving information transparency and visualization, enhancing user interaction and operation convenience, improving system reliability and safety, and optimizing construction efficiency and quality.

[0269] The user interface module is the key link between operators and control systems in the intelligent temperature control system for mass concrete. It displays the system operating status, real-time temperature data, control signals and the working status of cooling equipment through a friendly and intuitive interface, helping users monitor, analyze and adjust the operation of the system. The user interface module also supports manual adjustment functions, allowing users to intervene and control the system according to actual needs. The following is the specific working principle and operation process of the user interface module:

[0270] a. Data display and real-time monitoring

[0271] The primary function of the user interface module is to display real-time data of system operation, including sensor data, temperature prediction, cooling equipment status, PID controller output, etc. Through real-time monitoring, users can fully understand the temperature changes of concrete and the working status of the cooling system.

[0272] a.1 Real-time temperature data display

[0273] Temperature distribution diagram: The interface displays the real-time temperature data collected by various sensors inside the concrete, and displays these data as a two-dimensional or three-dimensional temperature distribution diagram through visualization tools. Users can intuitively see the temperature conditions in different areas of the concrete.

[0274] Temperature change trend chart: The interface displays the temperature change trend over a period of time and displays the temperature history in the form of a line chart or a bar chart, making it easier for users to track the temperature change process.

[0275] Temperature alarm: When the temperature at certain sensor locations exceeds the preset threshold, the system will highlight the area on the interface and trigger an alarm to inform the user of the temperature abnormality.

[0276] a.2 Cooling system status display

[0277] Cooling water system: The user interface displays the current operating status of the cooling water system, including cooling water flow, circulating pump status, cooling water temperature, etc. By comparing the inlet and outlet water temperatures, users can evaluate the actual cooling effect of the cooling water.

[0278] Spray system: If spray cooling is used, the interface will display the frequency, intensity and coverage area of ​​the spray to help users understand the actual operation of the spray cooling system.

[0279] a.3 Control parameters and signal display

[0280] PID controller output: The interface displays the output signal of the current PID controller, including the real-time adjustment values ​​of proportional control, integral control and differential control. Users can understand the regulation intensity of the PID controller on the cooling system.

[0281] Control target: Displays the current system temperature setting value and the allowed temperature range to help users understand the temperature control target.

[0282] System operation mode: Displays whether the current system is in automatic control mode or manual control mode. Users can switch modes at any time.

[0283] b. User manual control and adjustment

[0284] The user interface module not only displays system data, but also allows users to manually intervene in the cooling system and control parameters. Especially in special circumstances (such as equipment failure or environmental changes), users can manually adjust the operating status of the cooling system through the interface.

[0285] b.1 Manual mode switching

[0286] Manual and automatic mode switching: users can easily switch control modes through the interface. In automatic mode, the system automatically adjusts the cooling system according to the signal of the PID controller; in manual mode, users can directly control the parameters of the cooling system.

[0287] Operation confirmation prompt: When the user switches to manual mode or modifies important parameters, the system will issue a confirmation prompt to avoid misoperation.

[0288] b.2 Manual adjustment of cooling equipment

[0289] Cooling water system adjustment: Users can manually adjust the cooling water flow, circulating pump operation status and cooling water temperature. For example, when the concrete temperature changes suddenly, users can quickly increase the cooling water flow or reduce the cooling water temperature to accelerate cooling.

[0290] Spray system adjustment: If spray cooling is used, the user can adjust the frequency and intensity of the spray to reduce the concrete surface temperature in time.

[0291] b.3 Manual adjustment of PID parameters

[0292] Proportional gain (Kp), integral gain (Ki), differential gain (Kd): Users can manually adjust the parameters of the PID controller through the interface to cope with special temperature changes. For example, when the temperature fluctuates too much, the user can increase the differential gain (Kd) to suppress the fluctuation, or increase the integral gain (Ki) to eliminate the steady-state error.

[0293] b.4 Manually adjust history

[0294] Operation log: The interface will record all manual adjustment history, including user-modified parameters, device status changes, etc., for subsequent query and analysis.

[0295] c. Alarm and troubleshooting

[0296] The user interface module integrates the system's alarm and fault handling mechanism. When an abnormal situation occurs during system operation (such as overtemperature, equipment failure, etc.), the interface will issue an alarm prompt and guide the user to take appropriate measures.

[0297] c.1 Temperature alarm

[0298] High temperature alarm: When the temperature detected by certain sensors exceeds the safety threshold, the system will issue an alarm in the user interface. The high temperature area will be highlighted, and the alarm message will clearly indicate the problem location and sensor number.

[0299] Abnormal temperature trend alarm: If the neural network prediction module predicts that the temperature will rise abnormally in a certain period of time in the future, the system will issue an alarm in advance to prompt the user to take measures.

[0300] c.2 Cooling system fault alarm

[0301] Equipment failure alarm: If cooling equipment such as cooling water pumps and spray systems fails during operation (such as water pump failure, etc.), the system will issue an alarm in the interface and provide detailed fault information.

[0302] Insufficient cooling effect alarm: If the cooling effect fails to meet expectations (such as the temperature drops slowly after cooling), the system will prompt in the interface that the cooling efficiency is insufficient and recommend that the user make manual adjustments.

[0303] c.3 Alarm processing

[0304] Suggestions and troubleshooting: When the system alarm occurs, the user interface automatically provides suggestions, including increasing cooling flow, replacing equipment, or checking the operation of cooling equipment. The system can also provide troubleshooting steps to help users quickly locate the problem and repair it.

[0305] Alarm records and storage: All alarm information will be automatically stored in the system. Users can view historical alarm records at any time through the interface to understand the operation status of the system at different stages.

[0306] d. System setup and configuration management

[0307] The user interface module also allows users to perform advanced settings and configuration management of the system to ensure that the system can be flexibly applied in different environments and optimize the operating experience.

[0308] d.1 System Configuration

[0309] Cooling system configuration: Users can configure the operating parameters of the cooling water system and spray system (such as flow range, etc.) according to specific construction conditions to ensure the compatibility of the system with on-site equipment.

[0310] Sensor configuration: Users can configure the sampling frequency, data refresh rate and other parameters of the temperature sensor in the interface to optimize data acquisition performance.

[0311] d.2 Temperature control strategy setting

[0312] Target temperature curve setting: According to the needs of different construction stages, users can set different temperature control strategies and target temperature curves in the interface. For example, in the early hardening stage of concrete, a lower target temperature can be set; in the later stage, the temperature setting value can be increased to promote curing.

[0313] Temperature safety threshold setting: Users can adjust the temperature alarm thresholds of different areas according to project requirements to adapt to special construction conditions (such as high temperature weather, night construction, etc.).

[0314] d.3 User Rights Management

[0315] Permission control: The system supports multi-level user permission management, and the administrator can assign different operating permissions to different operators. For example, ordinary operators can only view temperature data and cooling system status, while administrators can modify control parameters, adjust system configuration, etc.

[0316] User log: The system will record each user's operation behavior to facilitate subsequent inquiries and responsibility tracking.

[0317] e. Data storage and historical analysis

[0318] The user interface module also integrates data storage and analysis functions, allowing users to view the system's historical operating data and perform trend analysis and operation evaluation.

[0319] e.1 Historical data storage

[0320] Temperature data storage: The system will regularly save the historical data collected by the temperature sensor, including temperature distribution, change trend and other information. Users can view past temperature data at any time in the interface.

[0321] Control parameter and operation record storage: Every operation of the user, including the adjustment of cooling equipment, modification of control parameters, etc., will be recorded, and the user can view the operation history through the interface.

[0322] e.2 Data analysis and report generation

[0323] Data trend analysis: The user interface provides a temperature trend analysis function to help users analyze the temperature change characteristics of different construction stages and predict possible temperature change trends in the future.

[0324] Report generation: The system can automatically generate temperature control operation reports according to user needs. The reports include temperature data, cooling equipment operating status, alarm records, etc., which are convenient for subsequent construction quality evaluation and optimization.

[0325] f.Multi-terminal support and remote monitoring

[0326] The user interface module supports multiple terminal access methods. Users can monitor the system's operating status, perform operations and make adjustments through local terminals or remote devices.

[0327] f.1Multi-terminal support

[0328] PC terminal: Users can log in to the interface through a PC or industrial computer to view system operation status and control cooling equipment in real time.

[0329] Mobile device support: Users can access the system anytime and anywhere through mobile devices such as mobile phones and tablets to understand the temperature control status and perform remote operations.

[0330] Remote control: In an environment that supports remote monitoring, users can remotely control cooling equipment, modify control parameters and view alarm information through a secure network connection.

[0331] f.2 Data Synchronization

[0332] Real-time synchronization: Data between different terminals is synchronized in real time, ensuring that no matter which device the user uses to access the system, they can obtain the latest system status and data.

[0333] Cloud storage: The user interface module can be integrated with the cloud platform, and all historical data and alarm records can be automatically synchronized to the cloud to ensure data security and persistent storage.

[0334] After obtaining and processing the temperature data inside the bulk concrete using this system, the following two coordinated operations will be performed. On the one hand, the PID control module calculates the temperature deviation from the target temperature data based on the real-time temperature data, and generates a cooling control signal through the PID algorithm, and then the cooling execution module executes the cooling control signal. In this way, the internal data of the concrete can always change towards the target temperature. On the other hand, the temperature prediction module will use the historical temperature data and the real-time temperature data to predict the future temperature change trend of the bulk concrete. This temperature change trend is the temperature prediction trend that may appear if the PID algorithm is continued without changing the PID control parameters. Afterwards, the PID control module will judge whether the current PID control parameters are appropriate (for example, whether the temperature change has achieved the preset effect, and whether the deviation from the target temperature is within the preset range) in combination with the future temperature change trend, and adjust the PID control parameters (such as proportional gain, integral gain and differential gain) when the judgment result is inappropriate. In this way, when the prediction result is that the temperature change trend is not ideal without changing the PID control parameters, the subsequent temperature can be made more ideal by optimizing and adjusting the PID control parameters. This system can more accurately predict the future temperature change trend of concrete and adjust the control parameters accordingly, so as to achieve precise control of concrete temperature. This helps to reduce temperature fluctuations, improve the uniformity of temperature inside concrete, and reduce the risk of temperature cracks.

[0335] By combining "using predicted temperature trends to adjust PID control parameters" with "PID algorithm for real-time temperature control", this system has better foresight and adaptability than conventional temperature control methods, and can effectively improve the accuracy of temperature control, and ensure the effectiveness of internal temperature control of large-volume concrete as much as possible. In addition, in this method, a neural network model is used as a temperature prediction model to predict the future temperature change trend of large-volume concrete. A neural network is an algorithm that can extract features from complex data through learning and training, has good nonlinear mapping capabilities, and is suitable for processing complex temperature fields. The intelligent temperature control system combined with PID control can perform real-time and precise control of the temperature of large-volume concrete in a complex and changeable environment, thereby effectively preventing the occurrence of temperature cracks. The present invention can effectively control the temperature of large-volume concrete during the pouring process.

[0336] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit the technical solution. Those skilled in the art should understand that those modifications or equivalent substitutions of the technical solution of the present invention that do not depart from the purpose and scope of the technical solution should be included in the scope of the claims of the present invention.

Claims

1. Mass concrete intelligent temperature control system based on neural network and PID control, characterized by: include: Temperature acquisition module, used to obtain temperature data inside large volumes of concrete in real time; A data processing module, used for processing the acquired temperature data; The temperature prediction module is used to use historical temperature data and real-time temperature data to predict the future temperature change trend of mass concrete through a preset prediction model; the prediction model is a neural network model; The PID control module is used to calculate the temperature deviation from the target temperature data based on the real-time temperature data, and generate a cooling control signal through the PID algorithm; the PID control module is also used to determine whether the current PID control parameters meet the preset requirements in combination with the temperature change trend predicted by the temperature prediction module, and automatically adjust the PID control parameters when the judgment result is that the preset requirements are not met; The cooling execution module is used to adjust the operation of the cooling system according to the cooling control signal.

2. The intelligent temperature control system for mass concrete based on neural network and PID control as claimed in claim 1, characterized in that: The prediction model is a neural network model based on a CNN-LSTM combination; the prediction model uses historical temperature data as training sample data for training, and uses real-time temperature data as input of the trained prediction model to predict the future temperature change trend of mass concrete; The neural network model based on the CNN-LSTM combination includes a convolutional layer, a time series processing layer and a fully connected layer; The convolutional layer is used to extract the spatial features of the temperature data to learn the association between temperature sensors at different locations; The time series processing layer is used to process the time series information of the spatial features of the extracted temperature data, extract the time series features of the temperature data, and further analyze the change pattern of the temperature data over time; The fully connected layer is used to integrate the spatial features of the temperature data extracted by the convolutional layer and the temporal features of the temperature data extracted by the time series processing layer, and predict the output temperature change trend prediction results.

3. The intelligent temperature control system for mass concrete based on neural network and PID control as claimed in claim 2, characterized in that: In the process of training the prediction model, the mean square error MSE is used as the loss function of the training: Among them, y i is the actual temperature, y′ i is the predicted temperature, n is the number of data points; the smaller the MSE, the closer the predicted result is to the true temperature.

4. The intelligent temperature control system for mass concrete based on neural network and PID control as claimed in claim 1, characterized in that: When the PID control module automatically adjusts the PID control parameters, the adjusted PID control parameters include proportional gain, integral gain and differential gain.

5. The intelligent temperature control system for mass concrete based on neural network and PID control as claimed in claim 1, characterized in that: When the cooling execution module adjusts the operation of the cooling system, the adjustments include cooling water flow, water temperature, and spray frequency.

6. The intelligent temperature control system for mass concrete based on neural network and PID control as claimed in claim 1, characterized in that: It also includes a user interface module for displaying temperature data, prediction results and the operating status of the cooling system in real time through a graphical interface; and for users to perform manual adjustments and interventions.

7. The intelligent temperature control system for mass concrete based on neural network and PID control as claimed in claim 1, characterized in that: The temperature acquisition module includes a plurality of temperature sensors which are respectively arranged at key positions at different depths of the bulk concrete; the key positions include areas where high temperature may occur and areas with large temperature gradients.

8. The intelligent temperature control system for mass concrete based on neural network and PID control as claimed in claim 7, characterized in that: When setting the temperature sensors of the temperature acquisition module, multiple sensors are arranged at key locations of the bulk concrete.

9. The intelligent temperature control system for mass concrete based on neural network and PID control as claimed in claim 1, characterized in that: When the data processing module processes the temperature data, the processing process includes filtering, noise reduction, data correction and data integration; among which, data integration includes spatial integration and temporal integration.

10. The intelligent temperature control system for mass concrete based on neural network and PID control as claimed in claim 9, characterized in that: The spatial integration includes spatially integrating sensor data from different positions and depths, generating a two-dimensional or three-dimensional scene of temperature distribution through data processing, and forming an overall temperature distribution model inside the concrete; the data processing includes interpolation and smoothing processing; The content of the time integration includes, for the temperature change data within a preset short period of time, integrating the temperature data at multiple moments through time series analysis, and extracting the key temperature change trend.

Images