Digital real-time monitoring method and system for machining size error of steel structural member

Abstract

The invention discloses a digital real-time monitoring method and system for the machining size error of a steel structure component, and belongs to the technical field of steel structure manufacturing monitoring, and the method comprises the steps: firstly, arranging a laser range finder and a visual sensor around steel structure component machining equipment, constructing a multi-dimensional data collection network, and obtaining the size data of the component in the machining process in real time; then, the collected original data are transmitted to a data processing module, and the data are calibrated and corrected by using the established mathematical model to remove noise and interference; thirdly, inputting the processed data into an analysis module, analyzing the data by using a machine learning algorithm, and identifying the characteristics and trend of the size error; finally, a control instruction is generated according to the analysis result and transmitted to machining equipment in real time, machining parameters are adjusted, and real-time control over the size error is achieved. According to the invention, flexible, accurate and real-time monitoring of the machining dimension error of the steel structural member is realized, and the machining quality and the production efficiency of the steel structural member are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of steel structure manufacturing monitoring, and in particular relates to a digital real-time monitoring method and system for machining dimensional errors of steel structure components. Background Art

[0002] During the fabrication of steel components, dimensional accuracy is a key factor affecting their quality and engineering safety. With the widespread application of steel structures in buildings, bridges, and other fields, the requirements for machining steel components are becoming increasingly demanding. Existing technologies primarily monitor dimensional errors during steel component fabrication using manual measurement and traditional automated measurement methods.

[0003] Manual measurement relies on surveyors using tools such as tape measures and calipers to perform point-by-point measurements. This method is not only inefficient and difficult to meet the needs of large-scale industrial production, but the measurement results are also significantly affected by human factors, resulting in unstable measurement accuracy. Traditional automated measurement methods, such as the use of fixed sensor arrays, have improved measurement efficiency to a certain extent, but they also have the following problems: On the one hand, the fixed sensor layout makes it difficult to adapt to steel structures of different specifications and shapes, resulting in poor flexibility; on the other hand, existing systems mostly use simple threshold comparison methods for data processing and error analysis, which cannot accurately identify and monitor complex processing errors in real time. This results in the inability to promptly detect dimensional deviations during the processing process, which in turn affects the processing quality and production efficiency of steel structures. In addition, existing technologies lack effective integration and sharing of measurement data, making it difficult for data between different processing links to be interoperable, which is not conducive to the optimization and control of the entire processing process.

[0004] Therefore, a digital real-time monitoring method and system for the machining dimensional errors of steel structure components are needed to solve the problems of low measurement efficiency, large measurement errors, difficulty in adapting to steel structure components of different specifications and shapes, and backward data processing in the existing technology. Summary of the Invention

[0005] The purpose of the present invention is to provide a digital real-time monitoring method and system for steel structure component machining dimensional errors, so as to solve the problems raised in the above background technology.

[0006] To achieve the above-mentioned object, the present invention provides the following technical solution: a method for digitally monitoring dimensional errors in steel structure component processing in real time, comprising the following steps:

[0007] S1. Based on the specifications and shapes of the steel structure components to be processed, multiple adjustable smart sensors are flexibly arranged around the processing equipment to build a dynamically adjustable sensor network;

[0008] S2. Based on step S1, the dimension data of the component is collected in real time through the sensor network, and the collected raw data is pre-processed by the data processing center to ensure the accuracy and reliability of the data;

[0009] S3. Based on historical processing data and design standards of steel structure components, a machine learning algorithm is used to establish a steel structure component processing dimensional error prediction model;

[0010] S4. Based on the pre-processed real-time measurement data obtained in step S2, the data is input into the error prediction model established in step S3 to perform real-time error calculation and analysis. By comparing the data with the design standard dimensions, the size and direction of the dimensional error of the current processed component are determined, and the error type is identified.

[0011] S5. When the dimensional error exceeds the preset threshold, the system automatically issues a warning signal to prompt the operator or processing equipment to make adjustments. At the same time, based on the error analysis results, the system generates corresponding control instructions and feeds them back to the control system of the processing equipment to automatically adjust the processing parameters, realize closed-loop control of the processing process, and correct dimensional errors in a timely manner.

[0012] It should be noted in the scheme that the intelligent sensors described in step S1 include but are not limited to laser ranging sensors, visual sensors and three-dimensional scanners. Each sensor is connected to the data processing center through a wireless communication module to transmit measurement data in real time.

[0013] It is further worth mentioning that the dimension data of the component collected in step S2 includes length, width, height, and angle, and the data preprocessing includes data filtering, denoising, and normalization processing to remove interference noise during the measurement process.

[0014] It should be further explained that the historical processing data in step S3 includes processing parameters, measurement data, and corresponding dimensional error data. The processing parameters specifically include cutting speed v, welding current I, and forming pressure P; measurement data: component length L, width W, angle θ; error labels; and deviations ΔL, ΔW, and Δθ between actual dimensions and design dimensions. The support vector regression algorithm is used to establish the error prediction model for steel structure component processing dimensional errors. The specific steps are as follows:

[0015] S31, select the input feature as and the output label as [ΔL, ΔW, Δθ];

[0016] S32, using radial basis function kernel:

[0017] K(x i ,x j )=exp(-γ||x i -x j ||2 )

[0018] where x i and x j Represents two sample vectors in the data set, ||x i -x j || represents the sample vector x i and x j The Euclidean distance between them is used to measure the distance between two vectors in the feature space. γ is the bandwidth parameter of the kernel function, γ = 0.1;

[0019] S33. Divide the historical data into a training set, a validation set, and a test set according to a ratio of 7:2:1, optimize the hyperparameters through grid search, and train the error prediction model M.

[0020] As a preferred embodiment, the real-time measurement data X in step S4 is real =[v real ,I real ,P real ,L real ,W real ,θ real ] Input error prediction model M, output prediction error And calculate the absolute error between the actual measured value and the designed value:

[0021] ΔL abs =|L real -L desing |

[0022] ΔW abs =|W real -W desing |

[0023] Δθ abs =|θ real -θ desing |

[0024] By comparison and Δ abs The residual e of , determines the error type;

[0025] Among them, if |e|<δ (δ is the residual threshold), it is a systematic error; if |e|≥δ, it is a random error.

[0026] The present invention also provides the following technical solution: a digital real-time monitoring system for machining dimensional errors of steel structure components, comprising:

[0027] Intelligent sensor module, used to collect dimensional data of steel structure components in real time during processing;

[0028] The wireless communication module is responsible for transmitting the data collected by the intelligent sensor to the data processing center in real time, and receiving the control instructions issued by the data processing center;

[0029] The data processing center is used to receive data transmitted by the wireless communication module and pre-process the data, establish an error model unit and analyze the error;

[0030] The early warning and control module sends out sound and light warning signals when it detects that the dimensional error exceeds the threshold. It also generates control instructions based on the error analysis results and transmits them to the control system of the processing equipment to achieve automatic adjustment of the processing parameters.

[0031] The human-computer interaction interface provides operators with an intuitive monitoring interface, displaying real-time measurement data, dimensional error information, and processing parameters.

[0032] As a preferred embodiment, the data processing center includes a data preprocessing unit, an error model building unit and a real-time error analysis unit;

[0033] The data preprocessing unit performs filtering, denoising and normalization on the original data;

[0034] The error model establishment unit establishes an error prediction model using a machine learning algorithm;

[0035] The real-time error analysis unit compares and analyzes the real-time measurement data with the model to calculate the dimensional error.

[0036] Compared with the prior art, the digital real-time monitoring method and system for machining dimensional errors of steel structure components provided by the present invention have at least the following beneficial effects:

[0037] (1) The present invention adopts a method of dynamically constructing a sensor network. According to the specifications and shapes of the steel structure components to be processed, multiple adjustable intelligent sensors can be flexibly arranged around the processing equipment, including laser ranging sensors, visual sensors and three-dimensional scanners. These sensors are equipped with adjustable mounting brackets and can be adjusted according to the characteristics of different components. At the same time, the constructed sensor network is dynamically adjustable. Unlike the existing fixed sensor array layout, it can re-plan the position and measurement range of the sensor at any time according to actual processing requirements, so that it can adapt to the processing size monitoring of steel structure components of different specifications and shapes, which greatly solves the problem of poor flexibility of the existing fixed sensor layout and significantly improves the applicability of the system. Whether it is a complex-shaped special-shaped component or a conventional standard component, this system can be efficiently adapted.

[0038] (2) The present invention uses a machine learning algorithm to establish an error prediction model. Based on historical processing data and design standards, it deeply analyzes the relationship between different processing parameters and dimensional errors, provides a more accurate error correction basis for measurement data, reduces the influence of human and environmental factors on measurement results, effectively improves the accuracy and reliability of measurement data, and achieves a significant improvement in measurement accuracy.

[0039] (3) In the present invention, during the processing of steel structure components, the sensor network collects the dimensional data of the components in real time and transmits it to the data processing center in real time through the wireless communication module. After receiving the data, the data processing center immediately inputs the pre-processed real-time measurement data into the error prediction model to perform real-time error calculation and analysis. Once it is detected that the dimensional error exceeds the preset threshold, the system can quickly and automatically issue an early warning signal. At the same time, based on the error analysis results, it immediately generates corresponding control instructions and feeds back to the control system of the processing equipment to automatically adjust the processing parameters to achieve closed-loop control of the processing process. This full-process real-time processing mechanism from data collection, analysis to feedback control can detect dimensional errors and make adjustments in a timely manner compared to the lagging monitoring and adjustment methods in the existing technology, effectively avoiding the accumulation of errors, greatly improving the processing quality and production efficiency, and ensuring that the processing process is always in a high-precision and high-efficiency operating state. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 Schematic diagram of the process of the digital real-time monitoring method for machining dimensional errors of steel structure components according to the present invention;

[0041] Figure 2 This is a schematic structural diagram of the digital real-time monitoring system for machining dimensional errors of steel structural components according to the present invention. DETAILED DESCRIPTION

[0042] The present invention will be further described below with reference to the embodiments.

[0043] See also Figure 1 The present invention provides a digital real-time monitoring method for machining dimensional errors of steel structure components, comprising the following steps:

[0044] S1. Based on the specifications and shapes of the steel structure components to be processed, multiple adjustable smart sensors are flexibly arranged around the processing equipment to build a dynamically adjustable sensor network;

[0045] The intelligent sensors include but are not limited to laser ranging sensors, visual sensors and three-dimensional scanners. Each sensor is connected to the data processing center through a wireless communication module to transmit measurement data in real time.

[0046] In some embodiments, adjustable sensor brackets are installed around key process equipment such as cutting, welding, and forming in a steel structure processing workshop according to the common specifications and shapes of the components to be processed. For example, for the processing of H-shaped steel, laser ranging sensors and visual sensors are installed on both sides and above the cutting equipment respectively. By adjusting the height and angle of the bracket, the sensors can cover the size measurement of the flange and web of the H-shaped steel. The sensors are connected to the server of the data processing center through a wireless communication module (such as Wi-Fi, Bluetooth or ZigBee) to form a flexible sensor network.

[0047] S2. Based on step S1, the dimension data of the component is collected in real time through the sensor network, and the collected raw data is pre-processed by the data processing center to ensure the accuracy and reliability of the data;

[0048] The dimension data of the acquisition component includes length, width, height, and angle. The data preprocessing includes data filtering, denoising, and normalization processing to remove interference noise during the measurement process.

[0049] In some embodiments, when a steel structure component enters the processing process, the sensor begins to collect data in real time, the laser ranging sensor measures the length and distance parameters of the component, the visual sensor obtains the two-dimensional image information of the component, and the three-dimensional scanner obtains the three-dimensional point cloud data of the component. The pre-processing unit of the data processing center uses median filtering, Gaussian filtering and other algorithms to denoise the original data to remove noise caused by factors such as equipment vibration and ambient light changes. At the same time, the data is normalized to convert the measurement data of different sensors into a unified coordinate system and unit for subsequent analysis and processing.

[0050] S3. Based on historical processing data and design standards of steel structure components, a machine learning algorithm is used to establish a steel structure component processing dimensional error prediction model;

[0051] The historical processing data includes processing parameters, measurement data, and corresponding dimensional error data. The processing parameters specifically include cutting speed v, welding current I, and forming pressure P; measurement data includes component length L, width W, and angle θ; error labels include deviations ΔL, ΔW, and Δθ between actual dimensions and design dimensions. The error prediction model for machining dimensional errors of steel structure components is established using a support vector regression algorithm. The specific steps are as follows:

[0052] S31, select the input feature as and the output label as [ΔL, ΔW, Δθ];

[0053] S32, using radial basis function kernel:

[0054] K(x i ,x j)=exp(-γ||x i -x j || 2 )

[0055] where x i and x j Represents two sample vectors in the data set, ||x i -x j || represents the sample vector x i and x j The Euclidean distance between them is used to measure the distance between two vectors in the feature space. γ is the bandwidth parameter of the kernel function, γ = 0.1;

[0056] S33. Divide the historical data into a training set, a validation set, and a test set according to a ratio of 7:2:1, optimize the hyperparameters through grid search, and train the error prediction model M.

[0057] In some embodiments, a large amount of data generated during historical processing is collected, including processing parameters (such as cutting speed, welding current, forming pressure, temperature, etc.), measurement data and corresponding dimensional error data, and these data are trained using deep learning algorithms (such as neural networks, random forests, etc.) to establish a mapping relationship model between processing parameters and dimensional errors. For example, in the welding process, a welding error prediction model is established by analyzing the relationship between welding current, welding speed and weld dimensional error.

[0058] S4. Based on the pre-processed real-time measurement data obtained in step S2, the data is input into the error prediction model established in step S3 to perform real-time error calculation and analysis. By comparing the data with the design standard dimensions, the size and direction of the dimensional error of the current processed component are determined, and the error type is identified.

[0059] The real-time measurement data X real =[v real ,I real ,P real ,L real ,W real ,θ real ] Input error prediction model M, output prediction error And calculate the absolute error between the actual measured value and the designed value:

[0060] ΔL abs =|L real -L desing |

[0061] ΔW abs =|W real -W desing |

[0062] Δθ abs =|θ real -θ desing |

[0063] By comparison and Δ abs The residual e of , determines the error type;

[0064] Among them, if |e|<δ (δ is the residual threshold), it is a systematic error; if |e|≥δ, it is a random error.

[0065] In some embodiments, the data processing center inputs the pre-processed real-time measurement data into an established error prediction model. The model calculates the dimensional error of the component based on the current processing parameters and measurement data. At the same time, the real-time measurement data is compared with the standard dimensions in the design drawings of the component to calculate the deviation value of each dimensional parameter. By analyzing the size, direction and change trend of the deviation value, the error type and possible causes of the error are determined, such as tool wear of the cutting equipment, positioning accuracy deviation of the welding robot, etc.

[0066] S5. When the dimensional error exceeds the preset threshold, the system automatically issues a warning signal to prompt the operator or processing equipment to make adjustments. At the same time, based on the error analysis results, the system generates corresponding control instructions and feeds them back to the control system of the processing equipment to automatically adjust the processing parameters, realize closed-loop control of the processing process, and correct dimensional errors in a timely manner.

[0067] In some embodiments, different error thresholds are set, such as a slight error threshold, a moderate error threshold, and a severe error threshold. When the dimensional error exceeds the slight error threshold, the system issues a light warning signal to alert the operator; when it exceeds the moderate error threshold, a sound warning signal is issued, and the specific location and cause of the error are displayed on the human-computer interaction interface; when it exceeds the severe error threshold, the system automatically pauses the processing equipment, issues an emergency warning signal, and waits for the operator to handle it. At the same time, for automatically adjustable processing equipment, such as CNC cutting machines, welding robots, etc., the system generates corresponding control instructions based on the error analysis results, and automatically adjusts processing parameters, such as cutting speed, welding current, tool position, etc., to achieve real-time correction of the processing process.

[0068] See also Figure 2 The present invention provides a digital real-time monitoring system for machining dimensional errors of steel structure components, comprising:

[0069] Intelligent sensor module, used to collect dimensional data of steel structure components in real time during processing;

[0070] The wireless communication module is responsible for transmitting the data collected by the intelligent sensor to the data processing center in real time, and receiving the control instructions issued by the data processing center;

[0071] A data processing center is used to receive data transmitted by the wireless communication module and pre-process the data, establish an error model unit and analyze the error; the data processing center includes a data pre-processing unit, an error model establishment unit and a real-time error analysis unit;

[0072] The data preprocessing unit performs filtering, denoising and normalization on the original data;

[0073] The error model establishment unit establishes an error prediction model using a machine learning algorithm;

[0074] The real-time error analysis unit compares and analyzes the real-time measurement data with the model to calculate the dimensional error;

[0075] The early warning and control module sends out sound and light warning signals when it detects that the dimensional error exceeds the threshold. It also generates control instructions based on the error analysis results and transmits them to the control system of the processing equipment to achieve automatic adjustment of the processing parameters.

[0076] The human-computer interaction interface provides operators with an intuitive monitoring interface, displaying real-time measurement data, dimensional error information, and processing parameters.

[0077] This solution has the following working process: first, laser rangefinders and visual sensors are arranged around the steel structure component processing equipment to build a multi-dimensional data acquisition network to obtain dimensional data during the component processing process in real time; then, the collected raw data is transmitted to the data processing module, and the established mathematical model is used to calibrate and correct the data to remove noise and interference; then, the processed data is input into the analysis module, and the machine learning algorithm is used to analyze the data to identify the characteristics and trends of dimensional errors; finally, control instructions are generated based on the analysis results and transmitted to the processing equipment in real time to adjust the processing parameters and achieve real-time control of dimensional errors.

[0078] In summary: Through the adjustable intelligent sensor network, it is possible to adapt to the processing dimension monitoring of steel structure components of different specifications and shapes, solve the problem of poor flexibility of the existing fixed sensor layout, and improve the applicability of the system; use multiple types of sensors for data acquisition, and through data preprocessing and error model establishment, effectively improve the accuracy and reliability of the measurement data, and reduce the influence of human factors and environmental factors on the measurement results; realize real-time data collection, analysis and feedback control of the steel structure component processing process, can timely detect dimensional errors and make adjustments, avoid error accumulation, and improve processing quality and production efficiency; use machine learning algorithms to establish an error prediction model, which can automatically analyze the relationship between processing parameters and dimensional errors, realize accurate identification and prediction of complex errors, and provide strong support for the optimization of the processing process.

[0079] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0080] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any 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. A digital real-time monitoring method for machining dimensional errors of steel structure components, characterized by: The following steps are involved: S1. Based on the specifications and shapes of the steel structure components to be processed, multiple adjustable smart sensors are flexibly arranged around the processing equipment to build a dynamically adjustable sensor network; S2. Based on step S1, the dimension data of the component is collected in real time through the sensor network, and the collected raw data is pre-processed by the data processing center to ensure the accuracy and reliability of the data; S3. Based on historical processing data and design standards of steel structure components, a machine learning algorithm is used to establish a steel structure component processing dimensional error prediction model; S4. Based on the pre-processed real-time measurement data obtained in step S2, the data is input into the error prediction model established in step S3 to perform real-time error calculation and analysis. By comparing the data with the design standard dimensions, the size and direction of the dimensional error of the current processed component are determined, and the error type is identified. S5. When the dimensional error exceeds the preset threshold, the system automatically issues a warning signal to prompt the operator or processing equipment to make adjustments. At the same time, based on the error analysis results, the system generates corresponding control instructions and feeds them back to the control system of the processing equipment to automatically adjust the processing parameters, realize closed-loop control of the processing process, and correct dimensional errors in a timely manner.

2. The method for digital real-time monitoring of steel structure component machining dimensional errors according to claim 1, characterized in that: The smart sensors in step S1 include but are not limited to laser ranging sensors, visual sensors and three-dimensional scanners. Each sensor is connected to the data processing center via a wireless communication module to transmit measurement data in real time.

3. The digital real-time monitoring method for machining dimensional errors of steel structural components according to claim 2 is characterized by: The dimension data of the component collected in step S2 include length, width, height, and angle. The data preprocessing includes data filtering, denoising, and normalization to remove interference noise during the measurement process.

4. The method for digital real-time monitoring of steel structure component machining dimensional errors according to claim 3, characterized in that: The historical processing data in step S3 includes processing parameters, measurement data and corresponding dimensional error data. The processing parameters specifically include cutting speed v, welding current I, and forming pressure P; measurement data: component length L, width W, angle θ; error labels; The deviations between the actual size and the design size are ΔL, ΔW, and Δθ. The error prediction model for machining size errors of steel structure components is established by using a support vector regression algorithm. The specific steps are as follows: S31, select the input feature as and the output label as [ΔL, ΔW, Δθ]; S32, using radial basis function kernel: K(x i ,x j )=exp(-γ||x i -x j || 2 ) where x i and x j Represents two sample vectors in the data set, ||x i -x j || represents the sample vector x i and x j The Euclidean distance between them is used to measure the distance between two vectors in the feature space. γ is the bandwidth parameter of the kernel function, γ = 0.1; S33. Divide the historical data into a training set, a validation set, and a test set according to a ratio of 7:2:1, optimize the hyperparameters through grid search, and train the error prediction model M.

5. The digital real-time monitoring method for machining dimensional errors of steel structural components according to claim 4 is characterized by: The real-time measurement data X in step S4 real =[v real ,I real ,P real ,L real ,W real ,θ real ] Input error prediction model M, output prediction error And calculate the absolute error between the actual measured value and the designed value: By comparison and Δ abs The residual e of , determines the error type; Among them, if |e|<δ (δ is the residual threshold), it is a systematic error; if |e|≥δ, it is a random error.

6. A digital real-time monitoring system for dimensional errors in steel structure component processing, characterized by: include: Intelligent sensor module, used to collect dimensional data of steel structure components in real time during processing; The wireless communication module is responsible for transmitting the data collected by the intelligent sensor to the data processing center in real time, and receiving the control instructions issued by the data processing center; The data processing center is used to receive data transmitted by the wireless communication module and pre-process the data, establish an error model unit and analyze the error; The early warning and control module sends out sound and light warning signals when it detects that the dimensional error exceeds the threshold. It also generates control instructions based on the error analysis results and transmits them to the control system of the processing equipment to achieve automatic adjustment of the processing parameters. The human-computer interaction interface provides operators with an intuitive monitoring interface, displaying real-time measurement data, dimensional error information, and processing parameters.

7. The digital real-time monitoring system for machining dimensional errors of steel structural components according to claim 6 is characterized by: The data processing center includes a data preprocessing unit, an error model building unit and a real-time error analysis unit; The data preprocessing unit performs filtering, denoising and normalization on the original data; The error model establishment unit establishes an error prediction model using a machine learning algorithm; The real-time error analysis unit compares and analyzes the real-time measurement data with the model to calculate the dimensional error.

Images