Intelligent batch production optimization method for fabricated precast components
By standardizing the production process parameters and quality monitoring data of prefabricated components, and using a multilayer perceptron model for online detection and prediction, an optimization library was constructed. This solved the problems of untimely quality monitoring and chaotic parameter management in the production of prefabricated components, and achieved efficient quality control and production optimization.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- MODIANGOU INTELLIGENT TECH (DONGGUAN) CO LTD
- Filing Date
- 2026-03-12
- Publication Date
- 2026-06-19
AI Technical Summary
The existing prefabricated component production process suffers from chaotic management of production process parameters, untimely quality monitoring, and a lack of effective feedback mechanisms, resulting in quality problems and low efficiency. It also lacks intelligent mass production optimization methods and makes it difficult to use historical data for accurate prediction and optimization.
By standardizing the production process parameters and quality monitoring data of prefabricated components, a structured production log is generated. Combined with a multilayer perceptron model, online quality detection and prediction are performed, and a production process optimization library is built to optimize production parameters and reduce abnormal components.
It has achieved efficient quality control and improved production efficiency, significantly reduced the generation of abnormal components, lowered production costs, and promoted the intelligent and refined production of prefabricated buildings.
Smart Images
Figure CN122243276A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of production optimization technology, specifically to an intelligent batch production optimization method for prefabricated components. Background Technology
[0002] With the rapid development of industrialized construction, prefabricated buildings have been widely used due to their advantages such as high efficiency, energy saving, and environmental protection. As a core component of prefabricated buildings, the production efficiency and quality of prefabricated components directly affect the development of the entire prefabricated building industry. However, many problems exist in the traditional production process of prefabricated components that urgently need to be solved. For example: The chaotic management of production process parameters, the lack of unified management and standardized processing, makes it easy for quality problems to occur during the production process, affecting the production quality and efficiency of components. The lack of timely quality monitoring and effective feedback mechanisms, coupled with the absence of effective correlation and feedback mechanisms between existing quality monitoring data and production process parameters, makes it difficult to adjust and optimize the production process. Consequently, it is impossible to effectively reduce the number of abnormal components and promote the healthy development of the prefabricated building industry. The lack of intelligent mass production optimization methods makes it impossible to fully utilize the potential information in historical production data, making it difficult to accurately predict and optimize the production process. This affects the mass production quality and efficiency of prefabricated components and restricts the further development of prefabricated buildings. Summary of the Invention
[0003] The purpose of this invention is to overcome the shortcomings of the prior art and provide an optimized method for intelligent mass production of prefabricated components to solve the problems mentioned in the background art.
[0004] The objective of this invention can be achieved through the following technical solution: an optimized method for intelligent mass production of prefabricated components, comprising: Step 1: Collect statistics on different types of prefabricated components in the production workshop. At the same time, based on the type of prefabricated component, standardize and integrate the production process parameters and quality monitoring data corresponding to the target component to generate a production log. Step 2: Based on the quality monitoring data of the target component, conduct online production quality inspection on the target component and obtain the quality monitoring results; the quality monitoring results include normal components and abnormal components; Step 3: Based on historical production logs and quality monitoring results, construct a quality prediction model based on a multilayer perceptron. Based on the prediction results, build a production process optimization library to obtain optimized values for production parameters, thereby optimizing the subsequent batch production of target components and reducing the number of abnormal components.
[0005] As a further aspect of the present invention, when performing standardization processing in step one, the collected quality monitoring data is normalized and mapped to the [0,1] interval to eliminate the influence of different dimensions and ranges; outliers in the data are removed using the standard deviation method, and missing values are filled in by interpolation. The standardized process parameters and quality monitoring data are aligned according to time series to generate a structured production log, which includes: component type, production batch, production process parameters, quality monitoring data, and timestamps corresponding to the start and end times of the process.
[0006] As a further aspect of the present invention, the process of performing online production quality inspection on the target component includes: S21. Set up several monitoring points on the target component on an average basis, and obtain the size data of the target component at each monitoring point; calculate the absolute difference between the size data of a single monitoring point and the size data of the same position on the design drawing of the target component to obtain the size deviation value of the monitoring point; wherein, the size deviation value of the monitoring point includes the length deviation value, the height deviation value, and the width deviation value. S22. Compare the dimensional deviation values of the monitoring points with the production error values of the corresponding dimensions. If the dimensional deviation values of the monitoring points are all greater than or equal to the production error values of the corresponding dimensions, a normal dimensional signal for the monitoring point is generated; otherwise, an abnormal dimensional signal for the monitoring point is generated.
[0007] As a further aspect of the present invention, the online production quality inspection of the target component also includes: S23. Based on the normal signal of the monitoring point size, obtain the concrete strength data of the target component at different curing times during the curing stage, and fit the concrete strength change curve using the Python data analysis library; In the concrete strength variation curve, the horizontal axis represents the curing time, and the vertical axis represents the concrete strength data; furthermore, the unit of curing time is hours or days, and the unit of concrete strength data is MPa. S24. By approximating the derivative of the concrete strength change curve, the rate of concrete strength change is obtained and judged; if the rate of concrete strength change is greater than 0, a normal concrete strength change signal is generated; otherwise, an abnormal concrete strength change signal is generated. S25. Based on the normal signal of concrete strength change, obtain the concrete strength data FZ of the target component at the end of the curing period, and compare it with the design strength standard value FB. If FZ≥FB, it means that the concrete strength of the target component meets the quality standard, and the target component is marked as a normal component; if FZ<FB, it means that the concrete strength of the target component does not meet the standard, and the target component is marked as an abnormal component and detected for scrapping.
[0008] As a further aspect of the present invention, the specific implementation steps in step three include: S31. Traverse the production logs of normal components to obtain the corresponding production process parameters and quality monitoring data, and extract features of key parameters; S32. Use the key parameters and quality monitoring data in the historical production logs as training samples for the quality prediction model, train the multilayer perceptron, and obtain the quality prediction model. S33. Based on the quality prediction model, input different combinations of key parameters to predict whether the target component to be produced is an abnormal component, obtain the corresponding key parameter combinations for which the prediction result is negative, and construct a production process tuning and optimization library.
[0009] As a further embodiment of the present invention, the specific implementation steps in step three also include: S34, identifying the target component with abnormal signals of monitoring point size and abnormal signals of concrete strength change, obtaining the corresponding production process parameters as abnormal parameters, and comparing and adjusting them with the key parameter combination in the production process optimization library to obtain the optimized value of the production parameters.
[0010] As a further aspect of the present invention, feature extraction of key parameters is performed in S31, including: The correlation coefficient between production process parameters and quality monitoring data was obtained using the Spearman correlation coefficient calculation formula. Production process parameter types with an absolute correlation coefficient greater than 0.6 were selected as key parameters.
[0011] As a further aspect of the present invention, training the multilayer perceptron in S32 includes: Key parameters are used as input features of the multilayer perceptron, and binary labels indicating whether a component is normal are used as output variables. The number of layers and nodes of the input, hidden, and output layers of the multilayer perceptron are set, and the weight coefficients and bias terms of the multilayer perceptron are continuously adjusted through the backpropagation algorithm so that the model can accurately fit the relationship between the input features and the output variables, thus obtaining a quality prediction model. The number of nodes in the input layer is equal to the number of input features; the number of layers and nodes in the hidden layer are set according to the actual number of key parameters; the number of nodes in the output layer is 1 and the sigmoid activation function is used.
[0012] Compared to existing solutions, the beneficial effects achieved by this invention are: This invention achieves efficient data management and utilization by standardizing the processing of production process parameters and quality monitoring data, and integrating them to generate production logs, thus providing a solid foundation for subsequent analysis. This invention, based on a real-time online quality inspection mechanism, can instantly detect anomalies in the production process, effectively distinguishing between normal and abnormal components, thus improving the accuracy and timeliness of quality control. Furthermore, it introduces a multi-layer perceptron to construct a quality prediction model, combining historical data with real-time monitoring results to intelligently predict and optimize production process parameters, building an optimization library. This not only significantly improves production efficiency but also substantially reduces the generation of abnormal components, lowers production costs, and enhances product quality, setting a new benchmark for intelligent and refined production in the prefabricated building industry. Attached Figure Description
[0013] The invention will now be further described with reference to the accompanying drawings.
[0014] Figure 1 This is a flowchart of the intelligent batch production optimization method for prefabricated components proposed in this invention. Detailed Implementation
[0015] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.
[0016] like Figure 1 As shown, this invention is an optimized method for intelligent mass production of prefabricated components, comprising the following steps: Step 1: Collect statistics on different types of prefabricated components in the production workshop. At the same time, based on the type of prefabricated component, standardize and integrate the production process parameters and quality monitoring data corresponding to the target component to generate a production log. Among them, different types of prefabricated components include, but are not limited to, prefabricated exterior wall panels, prefabricated balcony panels, prefabricated air conditioning panels, and prefabricated composite panels; the target component in the embodiments of the present invention can be any type of prefabricated component; Production process parameters are the operating parameters of the production line equipment during the production of the target component, including but not limited to mixing time, pouring speed, curing cycle, curing temperature, and curing humidity; Quality monitoring data is obtained by setting up multiple monitoring points to monitor the production process of the target component in real time. This includes the dimensional data of the target component and the concrete strength data. The dimensional data includes the length, width, and height of the precast component, which can be achieved through existing AI visual inspection technology. The specific implementation steps will not be elaborated here. The concrete strength data is based on ultrasonic rebound monitoring technology. By combining the sound velocity value of ultrasonic wave propagation and the rebound value, a two-parameter regression equation is established to obtain the concrete strength data. When implementing standardization, the collected quality monitoring data is normalized and mapped to the [0,1] interval to eliminate the influence of different units and ranges; outliers in the data are removed using the standard deviation method, and missing values are filled in by interpolation. The standardized process parameters and quality monitoring data are aligned according to time series to generate a structured production log, which includes: component type, production batch, production process parameters, quality monitoring data, and timestamps corresponding to the start and end times of the process. In this embodiment of the invention, by statistically analyzing and processing data on different types of prefabricated components in the production workshop, reliable standardized data support can be provided for the subsequent quality inspection process of the target components.
[0017] Step 2: Based on the quality monitoring data of the target component, conduct online production quality inspection on the target component and obtain the quality monitoring results; the quality monitoring results include normal components and abnormal components; The process of conducting online production quality inspection of target components includes: S21. Set up several monitoring points on the target component on an average basis, and obtain the size data of the target component at each monitoring point; calculate the absolute difference between the size data of a single monitoring point and the size data of the same position on the design drawing of the target component to obtain the size deviation value of the monitoring point; wherein, the size deviation value of the monitoring point includes the length deviation value, the height deviation value, and the width deviation value. S22. Compare the dimensional deviation values of the monitoring points with the production error values of the corresponding dimensions. If the dimensional deviation values of the monitoring points are all greater than or equal to the production error values of the corresponding dimensions, a normal dimensional signal for the monitoring point is generated; otherwise, an abnormal dimensional signal for the monitoring point is generated. It should be noted that the production error value for the corresponding dimension refers to the difference between the dimensional data of the target component and the design dimension standard value caused by various reasons during the production process. This difference value should be set within a range that does not affect the actual performance of the target component. In addition, this production error value is an absolute value, including length error value, height error value, and width error value. S23. Based on the normal signal of the monitoring point size, obtain the concrete strength data of the target component at different curing times during the curing stage, and fit the concrete strength change curve using the Python data analysis library; In the concrete strength variation curve, the horizontal axis represents the curing time, and the vertical axis represents the concrete strength data; furthermore, the unit of curing time is hours or days, and the unit of concrete strength data is MPa (megapascals). S24. By approximating the derivative of the concrete strength change curve, the rate of concrete strength change is obtained and judged; if the rate of concrete strength change is greater than 0, a normal concrete strength change signal is generated; otherwise, an abnormal concrete strength change signal is generated. It should be noted that judging the rate of change of concrete strength is used to determine whether the change of concrete strength during the curing stage of the target component is normal, not to determine whether the concrete strength of the target component meets the standard; in particular, abnormal changes in the concrete strength of the target component during the curing stage may be due to errors in the values set in the production process parameters. S25. Based on the normal signal of concrete strength change, obtain the concrete strength data FZ of the target component at the end of the curing period, and compare it with the design strength standard value FB. If FZ≥FB, it means that the concrete strength of the target component meets the quality standard, and the target component is marked as a normal component; if FZ<FB, it means that the concrete strength of the target component does not meet the standard, and the target component is marked as an abnormal component and detected for scrapping. Among them, the design strength standard value refers to the minimum strength value that concrete should reach, determined in the engineering design stage of the target component based on relevant specifications, engineering requirements, and factors such as the safety and durability of the concrete structure. Step 3: Based on historical production logs and quality monitoring results, construct a quality prediction model based on a multilayer perceptron. Based on the prediction results, build a production process optimization library to obtain optimized values for production parameters, thereby optimizing the subsequent batch production of target components and reducing the number of abnormal components. The specific process includes: S31. Traverse the production logs of normal components to obtain the corresponding production process parameters and quality monitoring data, and perform feature extraction of key parameters, including: using the Spearman correlation coefficient calculation formula to obtain the correlation coefficient between production process parameters and quality monitoring data, and selecting production process parameter types with an absolute value of correlation coefficient greater than 0.6 as key parameters; among which, the Spearman correlation coefficient calculation formula is an existing formula and will not be elaborated further. S32. Using key parameters and quality monitoring data from historical production logs as training samples for the quality prediction model, the multilayer perceptron is trained, including: using key parameters as input features of the multilayer perceptron, and binary classification labels indicating whether a component is normal as output variables; setting the number of layers and nodes of the input, hidden, and output layers of the multilayer perceptron, and continuously adjusting the weight coefficients and bias terms of the multilayer perceptron through the backpropagation algorithm, so that the model can accurately fit the relationship between the input features and the output variables, thus obtaining the quality prediction model; The number of nodes in the input layer is equal to the number of input features; the number of layers and nodes in the hidden layer are set according to the actual number of key parameters; the output layer has 1 node and uses the Sigmoid activation function. It should be noted that the multilayer perceptron is an existing model. During the training process, cross-validation is used to evaluate and optimize the model's performance to prevent overfitting and improve the model's generalization ability. The specific training process will not be described in detail. S33. Based on the quality prediction model, input different combinations of key parameters to predict whether the target component to be produced is an abnormal component, obtain the corresponding key parameter combination if the prediction result is negative, and build a production process tuning and optimization library. S34. Identify the target component with abnormal signals of monitoring point size and abnormal signals of concrete strength change, obtain the corresponding production process parameters as abnormal parameters, and compare and adjust them with the key parameter combination in the production process optimization library to obtain the optimized value of production parameters. In this embodiment of the invention, the optimized production parameter values are applied to the subsequent batch production of target components, thereby optimizing the batch production of target components, reducing the number of abnormal components, and improving the production efficiency and quality of prefabricated components.
[0018] In the several embodiments provided by this invention, it should be understood that the disclosed system can be implemented in other ways. For example, the embodiments of the invention described above are merely illustrative; for example, the division of modules is only a logical functional division, and there may be other division methods in actual implementation.
[0019] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.
[0020] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.
[0021] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.
[0022] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.
Claims
1. An optimized method for intelligent mass production of prefabricated components, characterized in that, include: Step 1: Collect statistics on different types of prefabricated components in the production workshop. At the same time, based on the type of prefabricated component, standardize and integrate the production process parameters and quality monitoring data corresponding to the target component to generate a production log. Step 2: Based on the quality monitoring data of the target component, conduct online production quality inspection on the target component and obtain the quality monitoring results; the quality monitoring results include normal components and abnormal components; Step 3: Based on historical production logs and quality monitoring results, construct a quality prediction model based on a multilayer perceptron. Based on the prediction results, build a production process optimization library to obtain optimized values for production parameters, thereby optimizing the subsequent batch production of target components and reducing the number of abnormal components.
2. The intelligent batch production optimization method for prefabricated components according to claim 1, characterized in that, In step one, when performing standardization, the collected quality monitoring data is normalized and mapped to the [0,1] interval to eliminate the influence of different dimensions and ranges. Outliers in the data are removed using the standard deviation method, and missing values are filled in using interpolation. The standardized process parameters and quality monitoring data are aligned according to time series to generate a structured production log, which includes: component type, production batch, production process parameters, quality monitoring data, and timestamps corresponding to the start and end times of the process.
3. The intelligent batch production optimization method for prefabricated components according to claim 2, characterized in that, The process of conducting online production quality inspection of target components includes: S21. Set up several monitoring points on the target component on an average basis, and obtain the size data of the target component at each monitoring point; calculate the absolute difference between the size data of a single monitoring point and the size data of the same position on the design drawing of the target component to obtain the size deviation value of the monitoring point; wherein, the size deviation value of the monitoring point includes the length deviation value, the height deviation value, and the width deviation value. S22. Compare the dimensional deviation values of the monitoring points with the production error values of the corresponding dimensions. If the dimensional deviation values of the monitoring points are all greater than or equal to the production error values of the corresponding dimensions, a normal dimensional signal for the monitoring point is generated; otherwise, an abnormal dimensional signal for the monitoring point is generated.
4. The intelligent batch production optimization method for prefabricated components according to claim 3, characterized in that, Online production quality inspection of target components also includes: S23. Based on the normal signal of the monitoring point size, obtain the concrete strength data of the target component at different curing times during the curing stage, and fit the concrete strength change curve using the Python data analysis library; In the concrete strength variation curve, the horizontal axis represents the curing time, and the vertical axis represents the concrete strength data; furthermore, the unit of curing time is hours or days, and the unit of concrete strength data is MPa. S24. By approximating the derivative of the concrete strength change curve, the rate of concrete strength change is obtained and judged; if the rate of concrete strength change is greater than 0, a normal concrete strength change signal is generated; otherwise, an abnormal concrete strength change signal is generated. S25. Based on the normal signal of concrete strength change, obtain the concrete strength data FZ of the target component at the end of the curing period, and compare it with the design strength standard value FB. If FZ≥FB, it means that the concrete strength of the target component meets the quality standard, and the target component is marked as a normal component; if FZ<FB, it means that the concrete strength of the target component does not meet the standard, and the target component is marked as an abnormal component and detected for scrapping.
5. The intelligent batch production optimization method for prefabricated components according to claim 1, characterized in that, The specific implementation steps in step three include: S31. Traverse the production logs of normal components to obtain the corresponding production process parameters and quality monitoring data, and extract features of key parameters; S32. Use the key parameters and quality monitoring data in the historical production logs as training samples for the quality prediction model, train the multilayer perceptron, and obtain the quality prediction model. S33. Based on the quality prediction model, input different combinations of key parameters to predict whether the target component to be produced is an abnormal component, obtain the corresponding key parameter combinations for which the prediction result is negative, and construct a production process tuning and optimization library.
6. The intelligent batch production optimization method for prefabricated components according to claim 5, characterized in that, The specific implementation steps in step three also include: S34, identifying the target component with abnormal signals of monitoring point size and abnormal signals of concrete strength change, obtaining the corresponding production process parameters as abnormal parameters, and comparing and adjusting them with the key parameter combination in the production process optimization library to obtain the optimized value of the production parameters.
7. The intelligent batch production optimization method for prefabricated components according to claim 5, characterized in that, Feature extraction of key parameters is performed in S31, including: The correlation coefficient between production process parameters and quality monitoring data was obtained using the Spearman correlation coefficient calculation formula. Production process parameter types with an absolute correlation coefficient greater than 0.6 were selected as key parameters.
8. The intelligent batch production optimization method for prefabricated components according to claim 5, characterized in that, Training the multilayer perceptron in S32 includes: Key parameters are used as input features of the multilayer perceptron, and binary labels indicating whether a component is normal are used as output variables. The number of layers and nodes of the input, hidden, and output layers of the multilayer perceptron are set, and the weight coefficients and bias terms of the multilayer perceptron are continuously adjusted through the backpropagation algorithm so that the model can accurately fit the relationship between the input features and the output variables, thus obtaining a quality prediction model. The number of nodes in the input layer is equal to the number of input features; the number of layers and nodes in the hidden layer are set according to the actual number of key parameters; the number of nodes in the output layer is 1 and the sigmoid activation function is used.