Intelligent 3D printing concrete member size and precision calibration system

Abstract

The invention relates to the technical field of concrete, in particular to an intelligent 3D printing concrete member size and precision calibration system which comprises a sensor network module, a real-time data analysis module, a machine learning optimization module, a calibration execution module, a self-adaptive material flow control module, a comprehensive feedback module and an enhanced user interaction module. Wherein the sensor network module is used for monitoring concrete flowing, interlayer bonding and component size in the 3D printing process in real time; and the real-time data analysis module is used for receiving the data of the sensor network module, carrying out primary processing and identifying deviation or abnormal conditions. By integrating the advanced sensor network, real-time data processing, self-adaptive control and enhancing the user interaction module, the 3D printing process can be accurately monitored and adjusted, the component precision is remarkably improved, meanwhile, user operation is simplified, and therefore the defects in the aspects of printing quality control and user interface friendliness in the prior art are overcome.

Description

Technical Field

[0001] The present invention relates to the field of concrete technology, and in particular to an intelligent 3D printing concrete component size and precision calibration system. Background Art

[0002] In the modern construction industry, 3D printing technology has become an important tool for manufacturing complex geometries and optimizing the design of building components, especially in the field of concrete 3D printing. This technology provides the possibility of manufacturing large structural components, which is difficult to achieve with traditional construction methods. However, with the rapid development of technology, a series of challenges have also arisen.

[0003] First, existing 3D printing systems often have difficulty ensuring the dimensional accuracy of components and the geometric accuracy of structures when printing large concrete structures. In addition, the fluidity and interlayer adhesion of concrete materials during the printing process are difficult to monitor and control in real time, resulting in unstable printing quality. These problems limit the application of 3D printing technology in the field of precision construction, especially when it comes to complex or load-bearing structures.

[0004] Secondly, the user interaction interface in existing systems usually fails to provide sufficient information and control options, making it difficult for operators to monitor the printing status in real time and respond to possible printing problems in a timely manner. The lack of user feedback mechanism reduces operational efficiency and increases operational complexity and the risk of errors.

[0005] In summary, the existing technology fails to provide an effective solution to achieve real-time monitoring and calibration of the dimensional accuracy of 3D printed concrete components, effective control of material flow and interlayer bonding conditions, and provide an enhanced user interaction interface to optimize the management of the printing process. Therefore, it is necessary to develop a new intelligent 3D printed concrete component size and accuracy calibration system to solve the above problems. Summary of the invention

[0006] Based on the above objectives, the present invention provides an intelligent 3D printed concrete component size and accuracy calibration system.

[0007] The intelligent 3D printed concrete component size and precision calibration system includes a sensor network module, a real-time data analysis module, a machine learning optimization module, a calibration execution module, an adaptive material flow control module, a comprehensive feedback module, and an enhanced user interaction module; among them,

[0008] Sensor network module: used to monitor concrete flow, interlayer bonding and component size in real time during 3D printing;

[0009] Real-time data analysis module: receives data from the sensor network module, performs preliminary processing and identifies deviations or abnormal conditions;

[0010] Machine Learning Optimization Module: Receives the output of the real-time data analysis module and optimizes printing parameters using historical data and predictive models;

[0011] Calibration execution module: adjusts the printing path and speed according to the parameters provided by the machine learning optimization module to achieve accurate calibration;

[0012] Adaptive material flow control module: used to adjust the viscosity and flow speed of concrete materials to adapt to different printing conditions and ensure structural stability;

[0013] Comprehensive feedback module: Based on the data of each module and the calibration results, it conducts overall performance analysis and fine-tunes the system to optimize performance;

[0014] Enhanced User Interaction Module: Users can monitor system status, enter control commands, and receive real-time feedback on print status and calibration recommendations.

[0015] Furthermore, the sensor network module includes a displacement sensor unit, a fluidity sensor unit and a structural integrity sensor unit; wherein,

[0016] Displacement sensor unit: includes a group of laser displacement sensors arranged along the moving axis of the print head of the 3D printing device, which is used to monitor the thickness and flatness of the printed concrete layer in real time, and measures the distance change by emitting laser and receiving its reflected signal to calculate the size difference between layers;

[0017] Fluidity sensor unit: A viscosity sensor is installed in the concrete supply system to monitor the fluidity of concrete. The viscosity sensor calculates the viscosity of the concrete by measuring the time it takes for the concrete to pass through a standard opening, thereby calculating the fluidity of the material.

[0018] Structural integrity sensor unit: Utilizes acoustic emission sensors, deployed in the printing area, to monitor interlayer adhesion. Acoustic emission sensors can detect tiny sound waves generated by poor interlayer bonding or cracks, thereby evaluating the structural integrity of the component in real time.

[0019] Furthermore, the real-time data analysis module includes a data cleaning unit, a data synchronization unit, a deviation detection unit and an abnormal condition prediction unit; wherein,

[0020] Data cleaning unit: used to format and clean the raw data from the sensor network module to eliminate erroneous readings caused by sensor noise or environmental interference. The data cleaning unit adopts a data cleaning algorithm, specifically, an outlier detection based on the Z score, where the Z score refers to the difference between the observed value and the mean value, divided by the standard deviation. The specific outlier calculation formula is: Where X is the observed value, μ is the mean value, and σ is the standard deviation;

[0021] Data synchronization unit: used to align the time stamps of data from different sensors for synchronous analysis. The data synchronization unit uses a time series analysis method including interpolation and time stamp marking to ensure the consistency of all data points in time;

[0022] Deviation detection unit: Analyze the cleaned and synchronized data to identify deviations from the preset printing model parameters. Use real-time statistical process control charts to monitor the data flow and identify data points that exceed the control limit. The specific calculation formula is:

[0023]

[0024] in, is the sample mean, Z is the Z score of the standard normal distribution, σ is the standard deviation, and n is the sample size;

[0025] Abnormal condition prediction unit: uses support vector machine (SVM) learning algorithm to train models based on historical data to predict and identify potential abnormal conditions.

[0026] Furthermore, the training model in the abnormal condition prediction unit includes the following steps:

[0027] S1: Extract key features from historical data, including interlayer temperature of concrete, printing speed, and material flow rate, and use F = {f1, f2, …, f n} means that f i represents the i-th feature;

[0028] S2: Standardize the feature data to eliminate the impact of dimension and speed up the algorithm’s convergence. The specific standardization formula is: Among them, μ i and σ i are the mean and standard deviation of the i-th feature respectively;

[0029] S3: Select the best SVM parameters, including penalty parameter C and kernel function parameter γ, through cross-validation method, traverse parameter combinations through grid search method, and use the classification accuracy of the cross-validation set as the performance indicator to select the optimal parameter combination;

[0030] S4: Under the above selected parameters, the SVM algorithm is used to train the model. The model is trained according to the following optimization problem:

[0031]

[0032] subjecttoy i (w·φ(xi )+b)≥1-ξ i ,ξ i ≥0,

[0033] Where w is the normal vector of the hyperplane, b is the bias term, and φ(x i ) is a function that maps input data to a high-dimensional space, ξ i is the slack variable, y i is the label of the i-th data point;

[0034] S5: After the model training is completed, it is used to monitor new printing data in real time to identify abnormal patterns that deviate from the normal printing process. Specifically, for each new observation point x, the SVM model calculates the decision function as follows: f(x) = w·φ(x) + b. When the value of f(x) is less than 1, the point will be marked as abnormal.

[0035] Furthermore, the machine learning optimization module includes a historical data analysis unit, a feature engineering unit, a prediction model training unit, and a parameter optimization unit; wherein,

[0036] Historical data analysis unit: collects parameter data of the historical 3D printing process including printing speed, concrete temperature, and layer thickness, and is used to establish a training data set. Each data point is marked as d i =(x i ,y i ), where x i represents the input feature vector, y i Indicates the corresponding print quality evaluation result;

[0037] Feature Engineering Unit: Analyze and preprocess the features in the historical data set, including selecting the most influential features, normalizing, and removing noise. Feature selection is based on information gain or correlation coefficient calculation. The normalization calculation formula is:

[0038] x i ′ j =(x ij -min(x j )) / (max(x j )-min(x j ), where x ij is the jth feature of the i-th sample;

[0039] Prediction model training unit: uses processed features and historical data to train the prediction model, specifically using random forest to optimize model parameters to maximize print quality. The random forest model constructs multiple decision trees T k And take the majority vote of its pattern recognition results to make a prediction function The prediction formula is:

[0040]

[0041] Parameter Optimization Unit: Based on the output of the prediction model, an optimization algorithm is used to adjust the 3D printing parameters to obtain the best printing quality. The optimization algorithm attempts to minimize the cost function C(θ), where θ represents the set of printing parameters and the cost function is defined as the deviation from the expected printing result:

[0042]

[0043] in, is the model's predicted output for print quality under parameter θ.

[0044] Furthermore, the calibration execution module includes a path adjustment unit and a speed adjustment unit; wherein,

[0045] Path adjustment unit: responsible for adjusting the movement trajectory of the 3D printing head using an algorithm based on the output parameters of the machine learning optimization module. The adjustment algorithm uses a B-spline curve to represent the printing path and changes the printing path by modifying the control points of the B-spline curve. When P represents the control point set of the printing path, the adjusted control point set P ′ Calculated by:

[0046] P ′ =P+ΔP, where ΔP is the control point offset derived based on the optimization module parameters;

[0047] Speed ​​adjustment unit: According to the parameters provided by the optimization module, the moving speed of the print head is adjusted to ensure that the printing speed matches the material flow rate and the rapid curing requirements. The speed adjustment adopts a predictive control strategy, where the print head speed V is adjusted to:

[0048] V ′ =V+α(V opt -V),

[0049] Among them, V opt is the final speed determined by the optimization module, and α is the speed adjustment factor used to smoothly transition to the new speed setting.

[0050] Furthermore, the adaptive material flow control module includes a viscosity adjustment unit and a flow speed control unit; wherein,

[0051] Viscosity adjustment unit: The viscosity of concrete is controlled in real time by adjusting the proportion of additives in concrete materials. The viscosity adjustment unit receives material fluidity data from the real-time data analysis module and automatically adjusts the additive pumping rate according to the preset viscosity parameter range. The viscosity adjustment algorithm is executed according to the following formula:

[0052] A ′ =A+β(V target -V current ),

[0053] Where A is the current pumping rate of the additive, A ′ is the adjusted pumping rate, β is the adjustment coefficient, V target is the target temperature value, V current is the current temperature value;

[0054] Flow speed control unit: According to the adjustment of printing path and speed, the flow speed of concrete is adjusted synchronously to ensure the uniform distribution of materials and printing accuracy. The flow speed control unit uses a flow meter and an adjustable valve to control the discharge rate of concrete. The flow speed control calculation formula is:

[0055] F ′ =F+γ(F opt -F),

[0056] Where F is the current flow velocity, F ′ is the adjusted flow velocity, γ is the flow velocity adjustment coefficient, F opt It is the optimum flow speed determined by the printing speed and path.

[0057] Furthermore, the comprehensive feedback module includes a performance analysis unit and a system fine-tuning unit; wherein,

[0058] Performance analysis unit: used to integrate data from the sensor network module, real-time data analysis module, machine learning optimization module, calibration execution module and adaptive material flow control module, and use a variety of statistical analysis techniques to evaluate performance, including principal component analysis PCA and factor analysis FA, which can identify variables and potential influencing factors in the data of each module and associate them with the printing quality results. The specific performance analysis formula is expressed as:

[0059] PCA(F)=W T F,

[0060] Among them, F is the feature data matrix from each module, W is the weight matrix of the feature vector in the PCA technique, and PCA(F) is the principal component score after transformation;

[0061] System fine-tuning unit: Based on the results of the performance analysis unit, a decision tree algorithm is used to identify which parameters need to be adjusted and how to adjust them to improve the overall performance of the system. Each node in the decision tree analysis represents a parameter decision point, and the node branch represents the adjustment direction. The final leaf node represents the recommended parameter adjustment value. The specific formula for generating the decision tree is:

[0062] DT(S)=Choosebestfeaturesplit(S),

[0063] Where S is the data set provided by the performance analysis unit, and DT(S) represents the generated decision tree model, which is used to output the decision path for parameter adjustment.

[0064] Furthermore, the enhanced user interaction module includes a user interface UI unit, a user input unit and a real-time feedback unit; wherein,

[0065] User Interface UI Unit: Includes an intuitive graphical interface that allows users to monitor the 3D printing process in real time and display data from the sensor network module and the real-time data analysis module. The interface uses data visualization technology to present data as charts and real-time flow charts;

[0066] User input unit: receives user input through a touch screen or other input device, allowing the user to quickly respond to feedback provided by the system, adjust parameters or issue instructions to pause / continue printing. The user input unit processes the user input and passes the command to the relevant control module for execution;

[0067] Real-time feedback unit: Based on the analysis results of the comprehensive feedback module, it provides operational suggestions and performance optimization tips. The real-time feedback unit uses decision support algorithms to show users adjustment suggestions, including predicted results and performance improvement measures, as well as how the adjustment suggestions affect the printing process.

[0068] Beneficial effects of the present invention:

[0069] The present invention, by combining real-time data analysis, can accurately monitor key parameters in the 3D printing process, such as the flow of concrete, interlayer adhesion and component size. The implementation of this system significantly improves the monitoring capability of the printing process, thereby preventing the occurrence of printing defects and ensuring the dimensional accuracy of the printed components and the geometric accuracy of the structure.

[0070] The present invention, through the integrated machine learning optimization module and calibration execution module, can learn historical printing data and automatically adjust printing parameters, such as path and speed, to adapt to different printing conditions. This intelligent self-adjustment mechanism not only improves printing efficiency, but also optimizes the quality of components. At the same time, the adaptive material flow control module ensures the fluidity and stability of concrete materials in various printing environments, solving the challenges of material control.

[0071] The present invention provides an intuitive interface and real-time feedback system for the operator through an enhanced user interaction module, which not only enables the operator to effectively monitor and adjust the printing process, but also improves the user's understanding and application of system functions through the educational auxiliary unit. Therefore, the present invention greatly simplifies the operation process, reduces the difficulty of operation, makes 3D printing technology more user-friendly, and provides higher flexibility and controllability for printing operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] In order to more clearly illustrate the technical solutions in the present invention 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 for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0073] Figure 1 Schematic diagram of a concrete component size and accuracy calibration system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0074] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments.

[0075] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0076] like Figure 1As shown, the intelligent 3D printing concrete component size and accuracy calibration system includes a sensor network module, a real-time data analysis module, a machine learning optimization module, a calibration execution module, an adaptive material flow control module, a comprehensive feedback module and an enhanced user interaction module; wherein,

[0077] Sensor network module: used to monitor concrete flow, interlayer bonding and component size in real time during 3D printing;

[0078] Real-time data analysis module: receives data from the sensor network module, performs preliminary processing and identifies deviations or abnormal conditions;

[0079] Machine Learning Optimization Module: Receives the output of the real-time data analysis module and optimizes printing parameters using historical data and predictive models;

[0080] Calibration execution module: adjusts the printing path and speed according to the parameters provided by the machine learning optimization module to achieve accurate calibration;

[0081] Adaptive material flow control module: used to adjust the viscosity and flow speed of concrete materials to adapt to different printing conditions and ensure structural stability;

[0082] Comprehensive feedback module: Based on the data of each module and the calibration results, it conducts overall performance analysis and fine-tunes the system to optimize performance;

[0083] Enhanced User Interaction Module: Users can monitor system status, enter control commands, and receive real-time feedback on print status and calibration recommendations.

[0084] The sensor network module includes a displacement sensor unit, a fluidity sensor unit and a structural integrity sensor unit; wherein,

[0085] Displacement sensor unit: includes a group of laser displacement sensors arranged along the moving axis of the print head of the 3D printing device, which is used to monitor the thickness and flatness of the printed concrete layer in real time, and measures the distance change by emitting laser and receiving its reflected signal to calculate the size difference between layers;

[0086] Fluidity sensor unit: A viscosity sensor is installed in the concrete supply system to monitor the fluidity of concrete. The viscosity sensor calculates the viscosity of the concrete by measuring the time it takes for the concrete to pass through a standard opening, thereby calculating the fluidity of the material.

[0087] Structural integrity sensor unit: uses acoustic emission sensors, deployed in the printing area, to monitor inter-layer bonding. Acoustic emission sensors can detect tiny sound waves generated by poor inter-layer bonding or cracks, thereby evaluating the structural integrity of the component in real time;

[0088] Through specific sensor configuration, the sensor network module is able to monitor key physical parameters in the 3D printing process in real time. This monitoring capability can significantly improve printing quality and prevent printing defects caused by insufficient monitoring.

[0089] The real-time data analysis module includes a data cleaning unit, a data synchronization unit, a deviation detection unit, and an abnormal condition prediction unit; wherein,

[0090] Data cleaning unit: used to format and clean the raw data from the sensor network module to eliminate erroneous readings caused by sensor noise or environmental interference. The data cleaning unit adopts a data cleaning algorithm, specifically, an outlier detection based on the Z score, where the Z score refers to the difference between the observed value and the mean value, divided by the standard deviation. The specific outlier calculation formula is: Where X is the observed value, μ is the mean value, and σ is the standard deviation;

[0091] Data synchronization unit: used to align the timestamps of data from different sensors for synchronous analysis. The data synchronization unit uses time series analysis methods including interpolation and timestamp marking to ensure the consistency of all data points in time;

[0092] Deviation detection unit: Analyze the data after cleaning and synchronization to identify the deviation from the preset printing model parameters. Use real-time statistical process control (SPC) charts, such as control chart algorithms, to monitor the data flow and identify data points that exceed the control limits. The specific calculation formula is:

[0093]

[0094] in, is the sample mean, Z is the Z score of the standard normal distribution, σ is the standard deviation, and n is the sample size;

[0095] Abnormal condition prediction unit: uses support vector machine (SVM) learning algorithm to train models based on historical data to predict and identify potential abnormal conditions.

[0096] The training model in the abnormal condition prediction unit includes the following steps:

[0097] S1: Extract key features from historical data, including interlayer temperature of concrete, printing speed, and material flow rate, and use F = {f1, f2, …, f n} means that f i represents the i-th feature;

[0098] S2: Standardize the feature data to eliminate the impact of dimension and speed up the algorithm’s convergence. The specific standardization formula is: Among them, μi and σ i are the mean and standard deviation of the i-th feature respectively;

[0099] S3: Select the best SVM parameters, including penalty parameter C and kernel function parameter γ, through the cross-validation method, traverse the parameter combination through the grid search method, and use the classification accuracy of the cross-validation set as the performance indicator to select the optimal parameter combination;

[0100] S4: Under the above selected parameters, the SVM algorithm is used to train the model. The model is trained according to the following optimization problem:

[0101]

[0102] subjecttoy i (w·φ(x i )+b)≥1-ξ i ,ξ i ≥0,

[0103] Where w is the normal vector of the hyperplane, b is the bias term, and φ(x i ) is a function that maps input data to a high-dimensional space, ξ i is the slack variable, y i is the label of the i-th data point;

[0104] S5: After the model training is completed, it is used to monitor new printing data in real time to identify abnormal patterns that deviate from the normal printing process. Specifically, for each new observation point x, the SVM model calculates the decision function as: f(x) = w·φ(x) + b. When the value of f(x) is less than 1, the point will be marked as abnormal.

[0105] The real-time data analysis module can effectively process and analyze sensor data and promptly identify deviations or abnormal conditions. This instant feedback mechanism helps prevent possible printing errors and improve the stability and reliability of the overall printing process.

[0106] The machine learning optimization module includes a historical data analysis unit, a feature engineering unit, a prediction model training unit, and a parameter optimization unit;

[0107] Historical data analysis unit: collects parameter data of the historical 3D printing process including printing speed, concrete temperature, and layer thickness, which are used to establish a training data set. Each data point is marked as d i =(x i ,y i ), where x i represents the input feature vector, y i Indicates the corresponding print quality evaluation result;

[0108] Feature Engineering Unit: Analyze and preprocess the features in the historical data set, including selecting the most influential features, normalizing, and removing noise. Feature selection is based on information gain or correlation coefficient calculation. The normalization calculation formula is:

[0109] x i ′ j =(x ij -min(x j )) / (max(x j )-min(x j ), where x ij is the jth feature of the i-th sample;

[0110] Prediction model training unit: Use the processed features and historical data to train the prediction model. Specifically, random forest is used to optimize the model parameters to maximize the printing quality. The random forest model constructs multiple decision trees T k And take the majority vote of its pattern recognition results to make a prediction function The prediction formula is:

[0111]

[0112] Parameter Optimization Unit: Based on the output of the prediction model, an optimization algorithm is used to adjust the 3D printing parameters to obtain the best printing quality. The optimization algorithm attempts to minimize the cost function C(θ), where θ represents the set of printing parameters and the cost function is defined as the deviation from the expected printing result:

[0113]

[0114] in, is the model's predicted output for the print quality under parameter θ;

[0115] The machine learning optimization module can learn historical printing data and optimize printing parameters, which makes the printing process more intelligent and automatically adjusts to changing printing conditions, thereby improving printing efficiency and component quality.

[0116] The calibration execution module includes a path adjustment unit and a speed adjustment unit; wherein,

[0117] Path adjustment unit: responsible for adjusting the movement trajectory of the 3D printing head using an algorithm based on the output parameters of the machine learning optimization module. The adjustment algorithm uses a B-spline curve to represent the printing path and changes the printing path by modifying the control points of the B-spline curve. When P represents the control point set of the printing path, the adjusted control point set P ′ Calculated by:

[0118] P ′ =P+ΔP, where ΔP is the control point offset derived based on the optimization module parameters;

[0119] Speed ​​adjustment unit: According to the parameters provided by the optimization module, the moving speed of the print head is adjusted to ensure that the printing speed matches the material flow rate and the rapid curing requirements. The speed adjustment adopts a predictive control strategy, where the print head speed V is adjusted to:

[0120] V ′ =V+α(V opt -V),

[0121] Among them, V opt is the post speed determined by the optimization module, α is the speed adjustment factor used to smoothly transition to the new speed setting;

[0122] The Calibration Execution module ensures precise print path and speed adjustments based on the predictions of the machine learning model; this fine control helps reduce material waste, optimize the printing process, and improve the accuracy and quality of the final product.

[0123] The adaptive material flow control module includes a viscosity adjustment unit and a flow velocity control unit; wherein,

[0124] Viscosity adjustment unit: The viscosity of concrete is controlled in real time by adjusting the proportion of additives in concrete materials. The viscosity adjustment unit receives material fluidity data from the real-time data analysis module and automatically adjusts the additive pumping rate according to the preset viscosity parameter range. The viscosity adjustment algorithm is executed according to the following formula:

[0125] A ′ =A+β(V target -V current ),

[0126] Where A is the current pumping rate of the additive, A ′ is the adjusted pumping rate, β is the adjustment coefficient, V target is the target temperature value, V current is the current temperature value;

[0127] Flow speed control unit: According to the adjustment of printing path and speed, the flow speed of concrete is adjusted synchronously to ensure the uniform distribution of materials and printing accuracy. The flow speed control unit uses a flow meter and an adjustable valve to control the discharge rate of concrete. The flow speed control calculation formula is:

[0128] F ′ =F+γ(F opt -F),

[0129] Where F is the current flow velocity, F ′ is the adjusted flow velocity, γ is the flow velocity adjustment coefficient, F opt It is the optimal flow speed determined by the printing speed and path;

[0130] The adaptive material flow control module provides precise viscosity and flow speed adjustment for concrete materials. This adaptive adjustment function ensures that structural stability can be maintained under different printing conditions and solves material flow problems that may occur during the printing process.

[0131] The comprehensive feedback module includes a performance analysis unit and a system fine-tuning unit;

[0132] Performance analysis unit: used to integrate data from the sensor network module, real-time data analysis module, machine learning optimization module, calibration execution module and adaptive material flow control module, and use a variety of statistical analysis techniques to evaluate performance. Statistical analysis techniques include principal component analysis PCA and factor analysis FA. This technology can identify variables and potential influencing factors in the data of each module and associate them with the print quality results. The specific performance analysis formula is expressed as:

[0133] PCA(F)=W T F,

[0134] Among them, F is the feature data matrix from each module, W is the weight matrix of the feature vector in the PCA technique, and PCA(F) is the principal component score after transformation;

[0135] System fine-tuning unit: Based on the results of the performance analysis unit, a decision tree algorithm is used to identify which parameters need to be adjusted and how to adjust them to improve the overall performance of the system. Each node in the decision tree analysis represents a parameter decision point, and the node branch represents the adjustment direction. The final leaf node represents the recommended parameter adjustment value. The formula generated by the decision tree is as follows:

[0136] DT(S)=Choosebestfeaturesplit(S),

[0137] Where S is the data set provided by the performance analysis unit, DT(S) represents the generated decision tree model, which is used to output the decision path for parameter adjustment;

[0138] The comprehensive feedback module provides system performance evaluation and fine-tuning based on comprehensive data analysis. This comprehensive evaluation helps the system self-optimize, solves performance bottlenecks that may be missed by single data point analysis, and ensures long-term efficiency and reliability.

[0139] The enhanced user interaction module includes a user interface UI unit, a user input unit and a real-time feedback unit; wherein,

[0140] User Interface UI Unit: Includes an intuitive graphical interface that allows users to monitor the 3D printing process in real time and displays data from the sensor network module and the real-time data analysis module. The interface uses data visualization technology to present data as charts and real-time flow charts, allowing users to intuitively identify trends and potential problems in the printing process;

[0141] User input unit: receives user input through a touch screen or other input device, allowing the user to quickly respond to feedback provided by the system, adjust parameters or issue instructions to pause / continue printing. The user input unit processes the user input and passes the command to the relevant control module for execution;

[0142] Real-time feedback unit: Provides operational suggestions and performance optimization tips based on the analysis results of the comprehensive feedback module. The real-time feedback unit uses decision support algorithms to show users adjustment suggestions, including predicted results and performance improvement measures, as well as how the adjustment suggestions affect the printing process;

[0143] The enhanced user interaction module improves the user's operating experience and system control capabilities. Through an intuitive user interface and real-time feedback, users can manage the printing process more effectively, solve the obstacles that users may encounter in complex system operations, and enhance user participation and the intuitiveness of operations.

[0144] The present invention is intended to cover all such substitutions, modifications and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. Intelligent 3D printing concrete component size and accuracy calibration system, characterized by: It includes sensor network module, real-time data analysis module, machine learning optimization module, calibration execution module, adaptive material flow control module, comprehensive feedback module and enhanced user interaction module; among them, Sensor network module: used to monitor concrete flow, interlayer bonding and component size in real time during 3D printing; Real-time data analysis module: receives data from the sensor network module, performs preliminary processing and identifies deviations or abnormal conditions; Machine Learning Optimization Module: Receives the output of the real-time data analysis module and optimizes printing parameters using historical data and predictive models; Calibration execution module: adjusts the printing path and speed according to the parameters provided by the machine learning optimization module to achieve accurate calibration; Adaptive material flow control module: used to adjust the viscosity and flow speed of concrete materials to adapt to different printing conditions and ensure structural stability; Comprehensive feedback module: Based on the data of each module and the calibration results, it conducts overall performance analysis and fine-tunes the system to optimize performance; Enhanced User Interaction Module: Users can monitor system status, enter control commands, and receive real-time feedback on print status and calibration recommendations.

2. The intelligent 3D printing concrete component size and accuracy calibration system according to claim 1 is characterized in that: The sensor network module includes a displacement sensor unit, a fluidity sensor unit and a structural integrity sensor unit; wherein, Displacement sensor unit: includes a group of laser displacement sensors arranged along the moving axis of the print head of the 3D printing device, which is used to monitor the thickness and flatness of the printed concrete layer in real time, and measures the distance change by emitting laser and receiving its reflected signal to calculate the size difference between layers; Fluidity sensor unit: A viscosity sensor is installed in the concrete supply system to monitor the fluidity of concrete. The viscosity sensor calculates the viscosity of the concrete by measuring the time it takes for the concrete to pass through a standard opening, thereby calculating the fluidity of the material. Structural integrity sensor unit: Utilizes acoustic emission sensors, deployed in the printing area, to monitor interlayer adhesion. Acoustic emission sensors can detect tiny sound waves generated by poor interlayer bonding or cracks, thereby evaluating the structural integrity of the component in real time.

3. The intelligent 3D printing concrete component size and accuracy calibration system according to claim 2 is characterized in that: The real-time data analysis module includes a data cleaning unit, a data synchronization unit, a deviation detection unit and an abnormal condition prediction unit; wherein, Data cleaning unit: used to format and clean the raw data from the sensor network module to eliminate erroneous readings caused by sensor noise or environmental interference. The data cleaning unit adopts a data cleaning algorithm, specifically, an outlier detection based on the Z score, where the Z score refers to the difference between the observed value and the mean value, divided by the standard deviation. The specific outlier calculation formula is: Where X is the observed value, μ is the mean value, and σ is the standard deviation; Data synchronization unit: used to align the time stamps of data from different sensors for synchronous analysis. The data synchronization unit uses a time series analysis method including interpolation and time stamp marking to ensure the consistency of all data points in time; Deviation detection unit: Analyze the cleaned and synchronized data to identify deviations from the preset printing model parameters. Use real-time statistical process control charts to monitor the data flow and identify data points that exceed the control limit. The specific calculation formula is: in, is the sample mean, Z is the Z score of the standard normal distribution, σ is the standard deviation, and n is the sample size; Abnormal condition prediction unit: uses support vector machine (SVM) learning algorithm to train models based on historical data to predict and identify potential abnormal conditions.

4. The intelligent 3D printing concrete component size and accuracy calibration system according to claim 3 is characterized in that: The training model in the abnormal condition prediction unit includes the following steps: S1: Extract key features from historical data, including interlayer temperature of concrete, printing speed, and material flow rate, and use F = {f1, f2, …, f n } means that f i represents the i-th feature; S2: Standardize the feature data to eliminate the impact of dimension and speed up the algorithm’s convergence. The specific standardization formula is: Among them, μ i and σ i are the mean and standard deviation of the i-th feature respectively; S3: Select the best SVM parameters, including penalty parameter C and kernel function parameter γ, through cross-validation method, traverse parameter combinations through grid search method, and use the classification accuracy of the cross-validation set as the performance indicator to select the optimal parameter combination; S4: Under the above selected parameters, the SVM algorithm is used to train the model. The model is trained according to the following optimization problem: subjecttoy i (w·φ(x i )+b)≥1-ξ i ,x i ≥0, Where w is the normal vector of the hyperplane, b is the bias term, and φ(x i ) is a function that maps input data to a high-dimensional space, ξ i is the slack variable, y i is the label of the i-th data point; S5: After the model training is completed, it is used to monitor new printing data in real time to identify abnormal patterns that deviate from the normal printing process. Specifically, for each new observation point x, the SVM model calculates the decision function as follows: f(x) = w·φ(x) + b. When the value of f(x) is less than 1, the point will be marked as abnormal.

5. The intelligent 3D printing concrete component size and accuracy calibration system according to claim 4 is characterized in that: The machine learning optimization module includes a historical data analysis unit, a feature engineering unit, a prediction model training unit, and a parameter optimization unit; wherein, Historical data analysis unit: collects parameter data of the historical 3D printing process including printing speed, concrete temperature, and layer thickness, and is used to establish a training data set. Each data point is marked as d i =(x i ,y i ), where x i represents the input feature vector, y i Indicates the corresponding print quality evaluation result; Feature Engineering Unit: Analyze and preprocess the features in the historical data set, including selecting the most influential features, normalizing, and removing noise. Feature selection is based on information gain or correlation coefficient calculation. The normalization calculation formula is: x i ′ j =(x ij -min(x j )) / (max(x j )-min(x j ), where x ij is the jth feature of the i-th sample; Prediction model training unit: uses processed features and historical data to train the prediction model, specifically using random forest to optimize model parameters to maximize print quality. The random forest model constructs multiple decision trees T k And take the majority vote of its pattern recognition results to make a prediction y. The prediction formula of function y is: Parameter Optimization Unit: Based on the output of the prediction model, an optimization algorithm is used to adjust the 3D printing parameters to obtain the best printing quality. The optimization algorithm attempts to minimize the cost function C(θ), where θ represents the set of printing parameters and the cost function is defined as the deviation from the expected printing result: in, is the model's predicted output for print quality under parameter θ.

6. The intelligent 3D printing concrete component size and accuracy calibration system according to claim 5 is characterized in that: The calibration execution module includes a path adjustment unit and a speed adjustment unit; wherein, Path adjustment unit: responsible for adjusting the movement trajectory of the 3D printing head using an algorithm based on the output parameters of the machine learning optimization module. The adjustment algorithm uses a B-spline curve to represent the printing path and changes the printing path by modifying the control points of the B-spline curve. When P represents the control point set of the printing path, the adjusted control point set P ′ Calculated by: P ′ =P+ΔP, where ΔP is the control point offset derived based on the optimization module parameters; Speed ​​adjustment unit: According to the parameters provided by the optimization module, the moving speed of the print head is adjusted to ensure that the printing speed matches the material flow rate and the rapid curing requirements. The speed adjustment adopts a predictive control strategy, where the print head speed V is adjusted to: V ′ =V+α(V opt -V), Among them, V opt is the final speed determined by the optimization module, and α is the speed adjustment factor used to smoothly transition to the new speed setting.

7. The intelligent 3D printing concrete component size and accuracy calibration system according to claim 6 is characterized in that: The adaptive material flow control module includes a viscosity adjustment unit and a flow speed control unit; wherein, Viscosity adjustment unit: The viscosity of concrete is controlled in real time by adjusting the proportion of additives in concrete materials. The viscosity adjustment unit receives material fluidity data from the real-time data analysis module and automatically adjusts the additive pumping rate according to the preset viscosity parameter range. The viscosity adjustment algorithm is executed according to the following formula: AND ′ =A+β(V target -IN current ), Where A is the current pumping rate of the additive, A ′ is the adjusted pumping rate, β is the adjustment coefficient, V target is the target temperature value, V current is the current temperature value; Flow speed control unit: According to the adjustment of printing path and speed, the flow speed of concrete is adjusted synchronously to ensure the uniform distribution of materials and printing accuracy. The flow speed control unit uses a flow meter and an adjustable valve to control the discharge rate of concrete. The flow speed control calculation formula is: F ′ =F+γ(F opt -F), Where F is the current flow velocity, F ′ is the adjusted flow velocity, γ is the flow velocity adjustment coefficient, F opt It is the optimum flow speed determined by the printing speed and path.

8. The intelligent 3D printing concrete component size and accuracy calibration system according to claim 7, characterized in that: The comprehensive feedback module includes a performance analysis unit and a system fine-tuning unit; wherein, Performance analysis unit: used to integrate data from the sensor network module, real-time data analysis module, machine learning optimization module, calibration execution module and adaptive material flow control module, and use a variety of statistical analysis techniques to evaluate performance, including principal component analysis PCA and factor analysis FA, which can identify variables and potential influencing factors in the data of each module and associate them with the printing quality results. The specific performance analysis formula is expressed as: PCA(F)=W T F, Among them, F is the feature data matrix from each module, W is the weight matrix of the feature vector in the PCA technique, and PCA(F) is the principal component score after transformation; System fine-tuning unit: Based on the results of the performance analysis unit, a decision tree algorithm is used to identify which parameters need to be adjusted and how to adjust them to improve the overall performance of the system. Each node in the decision tree analysis represents a parameter decision point, and the node branch represents the adjustment direction. The final leaf node represents the recommended parameter adjustment value. The specific formula for generating the decision tree is: DT(S)=Choosebestfeaturesplit(S), Where S is the data set provided by the performance analysis unit, and DT(S) represents the generated decision tree model, which is used to output the decision path for parameter adjustment.

9. The intelligent 3D printing concrete component size and accuracy calibration system according to claim 8, characterized in that: The enhanced user interaction module includes a user interface UI unit, a user input unit and a real-time feedback unit; wherein, User Interface UI Unit: Includes an intuitive graphical interface that allows users to monitor the 3D printing process in real time and display data from the sensor network module and the real-time data analysis module. The interface uses data visualization technology to present data as charts and real-time flow charts; User input unit: receives user input through a touch screen or other input device, allowing the user to quickly respond to feedback provided by the system, adjust parameters or issue instructions to pause / continue printing. The user input unit processes the user input and passes the command to the relevant control module for execution; Real-time feedback unit: Based on the analysis results of the comprehensive feedback module, it provides operational suggestions and performance optimization tips. The real-time feedback unit uses decision support algorithms to show users adjustment suggestions, including predicted results and performance improvement measures, as well as how the adjustment suggestions affect the printing process.

Images