Method, system, electronic equipment and storage medium for rapid prediction of ship welding quota working hours

Abstract

The present invention proposes a method for quickly predicting the standard working hours for ship welding, which belongs to the field of shipbuilding optimization technology. The present invention first establishes an actual working hour collection system by transforming the welding machine into the Internet of Things, integrating the task dispatching system, and continuously improving the database related to the actual working hours through the system. Secondly, the xgboost algorithm is used to achieve accurate prediction of the standard working hours for ship welding. The present invention also proposes a system for quickly predicting the standard working hours for ship welding, as well as electronic equipment and storage media. Compared with the existing technology, the present invention cleans the collected welding machine parameters by setting the threshold values ​​of each parameter when the welding machine is operating, extracts the effective welding working hours, removes the interference of invalid working hours, and improves the accuracy of the actual working hour calculation. At the same time, a working hour prediction model is established based on the XGboost algorithm, and the actual working hour data collected and cleaned are iteratively trained to achieve accurate prediction of welding working hours.

Description

Technical Field

[0001] The present invention relates to the technical field of shipbuilding optimization, and in particular to a method for quickly predicting ship welding quota working hours based on welding equipment data collection. Background Art

[0002] The ship welding quota time refers to the time required to complete a specific welding workload. It serves as a crucial basis for production planning, performance evaluation, and cost accounting, helping companies implement scientific management. However, with the current upgrade of welding equipment and the development of new welding processes, the original system for calculating ship welding quota time is no longer sufficient. By iteratively optimizing the core welding time algorithm using actual time data collected from welding equipment, companies can continuously improve their ship welding quota time calculation system and increase its accuracy.

[0003] Earlier methods of setting standard working hours were mainly based on empirical estimation, that is, some enterprise managers or experienced producers estimated labor consumption based on past production experience, taking into account the company's own production equipment, production processes, physical products, and the workers' own quality levels.

[0004] Statistical analysis involves recording the hours of production, building a database based on years of accumulated production experience. The weighted average of these statistics is used to calculate the average of the quota hours. This is then used to analyze the various factors influencing the quota and determine the hourly quota.

[0005] Case-based reasoning (CBR) searches through past cases in a case library to identify the best match for a new case. Depending on the problem being solved, different similarity metrics are used as matching criteria to retrieve matching cases that meet the criteria. Of course, the data in a CBR case library is not simply a storage of actual case data. Before new case data is adopted and integrated into the case library, the actual case data must be revised and the revision results evaluated. Only cases that pass the evaluation will be retained in the case library.

[0006] Existing automated calculation methods for ship welding quota hours are mainly based on formulas. These methods are based on a quota hour calculation system. However, in reality, welding equipment is constantly upgraded, welding processes are constantly optimized, and construction specifications are constantly updated. Affected by these factors, actual welding hours are constantly changing. However, the quota hour calculation system cannot currently be independently iterated. Currently, the iterative improvement of the welding quota hour calculation system requires independent revisions by the quota standard setters. This method is easily affected by the subjective factors of the setters, resulting in large deviations between different setters. The quota hour results are inconsistent, not universal, and cannot be used as scientific and accurate data support.

[0007] The above method is affected by various factors, resulting in varying degrees of error in the calculation system, and the quota-to-actual ratio cannot reach the theoretical minimum. During the development of the quota calculation system, it is necessary to compare a large amount of quota hours with actual hours. By incorporating this data into the model algorithm library and conducting big data analysis, the quota hour calculation system can be standardized and iterated. Summary of the Invention

[0008] The technical problem to be solved by the present invention is to address the shortcomings of the above-mentioned existing technologies, overcome at least one defect of the above-mentioned existing technologies, and provide a fast and accurate method for calculating the standard working hours of ship welding. The present invention first establishes an Internet of Things transformation of the welding machine, integrates the task dispatching system, and establishes an actual working time collection system. Through this system, the association database of parameters such as processing quantity and welding process and actual working hours is continuously improved. Secondly, the XGBoost algorithm is used to achieve accurate prediction of the standard working hours of ship welding.

[0009] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0010] In a first aspect, the present invention proposes a method for quickly predicting ship welding quota working hours based on welding equipment data collection, comprising the steps of:

[0011] S1. Taking each actual weld as the target, a task dispatching system is established through IoT sensing technology. The welding tasks are dispatched based on the weld, and the welding equipment operating parameters are collected and stored in the server.

[0012] S2. Clean the collected welding equipment data to obtain the actual working hours of each weld;

[0013] S3. Correlate the obtained effective working time data with the material quantity data of the task dispatching system to form a data set of material quantity, process, and working time for each weld, and establish a welding information database;

[0014] S4. The historical data of the welding information database formed in step S3 is trained based on the extreme gradient boosting algorithm to predict the welding task operation duration. The welding operation duration predicted by the extreme gradient boosting algorithm is corrected according to the real-time welding operation duration data of the predicted welds. The model is iteratively optimized to ultimately achieve the goal of continuously reducing the quota-to-actual ratio.

[0015] In a second aspect, the present invention proposes a system for quickly predicting ship welding quota working hours based on welding equipment data collection, comprising:

[0016] The parameter collection module is configured to perform the following actions: taking each actual weld as an object, establishing a task dispatching system through IoT sensing technology, dispatching welding tasks based on the weld as the object, and collecting and storing the welding equipment operating parameters in the server;

[0017] The data cleaning module is configured to perform the following actions: clean the collected welding equipment data to obtain the actual working hours of each weld;

[0018] The welding information database module is configured to perform the following actions: establish a correlation between the obtained effective working time data and the material quantity data of the task dispatching system to form a data set of material quantity, process, and working time of each weld, and establish a welding information database;

[0019] The model prediction module is configured to perform the following actions: train the historical data of the welding information database based on the extreme gradient boosting algorithm to predict the welding task operation duration, and correct the operation time predicted by the extreme gradient boosting algorithm based on the real-time welding operation time data of the predicted weld, iteratively optimize the model, and ultimately achieve the goal of continuously reducing the quota-to-actual ratio.

[0020] In a third aspect, the present invention provides an electronic device comprising: a memory and a processor, wherein the memory stores a computer program executable by the processor, and when the processor executes the computer program, the method for quickly predicting the ship welding quota working hours of the present invention is implemented.

[0021] In a fourth aspect, the present invention further proposes a storage medium having a computer program stored thereon. When the computer program is read and executed, the method for quickly predicting the ship welding quota working hours of the present invention is implemented.

[0022] The present invention adopts the above technical solution, which has the following technical effects compared with the prior art:

[0023] The present invention collects actual welding working hours through Internet of Things collection technology, integrates the task dispatching system, establishes the cross-reference relationship between actual working hours and weld welding, combines the historical welding working hours data, and establishes a welding working hour prediction model based on the XGboost machine learning algorithm to achieve accurate calculation of welding quota working hours, thereby continuously reducing the quota-to-actual ratio and ensuring the effectiveness and accuracy of the quota working hours in real time. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a processing flow chart of the present invention. DETAILED DESCRIPTION

[0025] In order to more clearly demonstrate the technical solution of the present invention, the present invention is further explained below in conjunction with the accompanying drawings and specific embodiments. It should be understood that these embodiments are only used to illustrate the present invention and are not used to limit the scope of the present invention.

[0026] This embodiment is an iterative optimization method for ship welding quota working time algorithm based on welding equipment data collection, referring to Figure 1 , the specific implementation steps are as follows:

[0027] Step 1: For each actual weld, the operating parameters of the welding equipment are collected through IoT sensing technology, including current, voltage, wire feeding time, wire feeding speed, welding speed, welding gun swing, etc. This data provides data support for subsequent calculations of welding hours and weld length;

[0028] Step 2: Clean the collected welding equipment data. By setting the thresholds of various parameters during welding operation, the effective working hours are extracted from the collected data and the invalid non-working hours are removed.

[0029] Step 3: Integrate the effective working hour data obtained through data cleaning with the task dispatching system. Taking the weld as the object, the data such as weld length, welding method, welding posture, part A material, part A thickness, part B material, part B thickness, welding current, welding voltage, welding time, etc. in the task dispatching system are formed into relational data, and the data is preprocessed. The continuous data such as weld length, part thickness, welding current, welding voltage, etc. are normalized, and the discrete data such as welding method, welding posture, part material, etc. are encoded using one-hot encoding to provide data support for subsequent working hour calculation;

[0030]

[0031] Step 4: The extreme gradient boosting (XGboost) algorithm is used to train the historical data on the amount of work, process, and working hours generated in step 3 to predict the duration of the welding task. The working hours predicted by the XGboost algorithm are corrected based on the real-time welding working hours data of the predicted welds. The model is iteratively optimized to ultimately achieve the goal of continuously reducing the quota-to-actual ratio.

[0032] As a specific embodiment of the present invention, in step 1: in the task dispatching system, welding tasks are dispatched based on welds. The dispatch order contains information such as weld name, weld length, weld type, weld form, part A material, part A thickness, part B material, part B thickness, welding process, and weld barcode. When performing welding operations, on-site construction personnel scan the barcode information in the dispatch order and then perform the welding operation. Throughout the welding process, various operating indicators of the welding machine are collected by the collection system and uploaded to the server, including parameter data such as the welding machine current, voltage, wire feed time, wire feed speed, welding speed, and welding gun swing. The more comprehensive the collected parameters, the more accurate the iterative results of the quota measurement system.

[0033] As a specific embodiment of the present invention, in step 2: since the collected welding machine data contains valid operating parameters and invalid operating parameters, these collected data cannot be directly used for the definition of actual working hours. It is necessary to set the threshold values ​​of various parameters during welding machine operation, clean the collected data, extract the effective operating time of the welding machine, and finally obtain the actual working hours of each weld to provide support for subsequent data processing.

[0034] As a specific embodiment of the present invention, in step 3: the operation data obtained in step 2 is associated with the material quantity data in the dispatching system. Taking the weld as the object, the weld length, welding form, welding posture, part A material, part A thickness, part B material, and part B thickness in the dispatching system are matched with the actual working hours in step 2. The data is pre-processed to form a data set of the material quantity, process, and working hours of each weld, and a welding information database is established. As the database is continuously improved, it provides comprehensive data support for subsequent working hour measurement model training.

[0035] Among them, the maximum and minimum normalization formula is as follows:

[0036]

[0037] As a specific embodiment of the present invention, in step 4: the first step of establishing a welding time prediction model based on the XGboost algorithm is to construct an objective function:

[0038]

[0039] Where l is the loss function; y i is the true value of the i-th sample; is the predicted value of the i-th sample after t-1 iterations; f t (X i ) is a new function added in the iterative process; is the regularization term of the objective function; C is the constant term.

[0040] The objective function includes a loss function and a regularization term, wherein the regularization term is used to adjust the complexity of the model to prevent the model from overfitting during training; the second step uses the Taylor series method to expand the objective function and extract and separate some constants to simplify the calculation; the third step is to parameterize the tree structure and bring it into the objective equation; the fourth step is to find a tree with the optimal structure, and a greedy algorithm is used in the search process; finally, the data collected in step 3 is used as input data and grid search and K-fold cross validation are used to optimize the model parameter values ​​of the preset default Xgboost model. The model parameters include the maximum depth of the subtree, the weight sum of the minimum leaf node samples, the proportion of training data samples, the learning rate, the number of iterations, the feature sampling ratio, the leaf node value of the partition tree, and the random seed, to obtain the optimized Xgboost model to ensure the model prediction performance. After completing the training of the prediction model and parameter optimization, the characteristic attributes of the weld to be predicted, such as weld name, weld length, welding form, welding posture, part A material, part A thickness, part B material, part B thickness, etc., are used as input parameters of the model, and the working hours are used as output to obtain the predicted working hours of the weld to be predicted.

[0041] This embodiment also proposes an emergency short-term forecast device based on multi-source data fusion, including:

[0042] The ship welding quota man-hour rapid prediction system based on welding equipment data collection includes:

[0043] The parameter collection module is configured to perform the following actions: taking each actual weld as an object, establishing a task dispatching system through IoT sensing technology, dispatching welding tasks based on the weld as the object, and collecting and storing the welding equipment operating parameters in the server;

[0044] The data cleaning module is configured to perform the following actions: clean the collected welding equipment data to obtain the actual working hours of each weld;

[0045] The welding information database module is configured to perform the following actions: establish a correlation between the obtained effective working time data and the material quantity data of the task dispatching system to form a data set of material quantity, process, and working time of each weld, and establish a welding information database;

[0046] The model prediction module is configured to perform the following actions: train the historical data of the welding information database based on the extreme gradient boosting algorithm to predict the welding task operation duration, and correct the operation time predicted by the extreme gradient boosting algorithm based on the real-time welding operation time data of the predicted weld, iteratively optimize the model, and ultimately achieve the goal of continuously reducing the quota-to-actual ratio.

[0047] It should be noted that the description of the system device in the embodiment of the present application is similar to the description of the above-mentioned method embodiment, and has similar beneficial effects as the method embodiment, so it will not be repeated.

[0048] This embodiment further provides an electronic device, including: a memory and a processor, wherein the memory stores a computer program executable by the processor, and when the processor executes the computer program, the method for quickly predicting the ship welding quota working hours according to the embodiment of the present invention is implemented.

[0049] Finally, this embodiment provides a storage medium having a computer program stored thereon. When the computer program is read and executed, the method for quickly predicting the ship welding quota working hours of the present invention is implemented.

[0050] The program code for implementing the methods of the present application can be written in any combination of one or more programming languages. Such program code can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the program code is executed by the processor or controller, the functions / operations specified in the flow charts and / or block diagrams are implemented. The program code can be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0051] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), and the Internet.

[0052] A computer system may include a client and a server. The client and server are generally remote from each other and typically interact through a communication network. The client-server relationship arises through computer programs running on the respective computers and having a client-server relationship with each other. The server may be a cloud server, a server in a distributed system, or a server integrated with a blockchain.

[0053] The above descriptions are only partial embodiments of the present invention. It should be pointed out that ordinary technicians in this technical field can make several improvements and modifications without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A method for quickly predicting ship welding quota working hours based on welding equipment data collection, characterized in that: Including steps: S1. Taking each actual weld as the target, a task dispatching system is established through IoT sensing technology. The welding tasks are dispatched based on the weld, and the welding equipment operating parameters are collected and stored in the server. S2. Clean the collected welding equipment data to obtain the actual working hours of each weld; S3. Correlate the obtained actual working time data with the material quantity data of the task dispatching system to form a data set of material quantity, process, and working time for each weld, and establish a welding information database; S4. The historical data of the welding information database formed in step S3 is trained based on the extreme gradient boosting algorithm to predict the welding task duration. The welding task duration predicted by the extreme gradient boosting algorithm is corrected based on the real-time welding task duration data of the predicted welds. The model is iteratively optimized to ultimately achieve the goal of continuously reducing the quota-to-actual ratio. In step S4, the steps of establishing the welding time prediction XGboost model based on the extreme gradient boosting algorithm are as follows: (1) Constructing the objective function Among them, l is the loss function; y i is the true value of the i-th sample; is the predicted value of the i-th sample after t-1 iterations; f t (X i ) is a new function added in the iterative process; is the regularization term of the objective function; C is the constant term; The objective function includes a loss function and a regularization term. The regularization term is used to adjust the complexity of the model to prevent the model from overfitting during training. (2) Use Taylor series method to expand the objective function: in, is the first-order derivative of the loss function; is the second-order derivative of the loss function; It is related to the previous t-1 trees, so when adding the t-th tree, it is a constant term and has no effect on the optimization of the function. Therefore, removing all constant terms, the objective function is: (3) Parameterize the tree structure and bring it into the target equation; (4) Find a tree with the best structure and use a greedy algorithm in the search process; (5) The data collected in step S3 is used as input data and grid search and K-fold cross validation are used to optimize the model parameter values ​​of the preset default Xgboost model. The model parameters include the maximum depth of the subtree, the weight sum of the minimum leaf node samples, the proportion of training data samples, the learning rate, the number of iterations, the feature sampling ratio, the leaf node value of the partition tree, and the random seed, to obtain the optimized Xgboost model to ensure the model prediction performance.

2. The method for rapid prediction of ship welding quota working hours based on welding equipment data collection according to claim 1 is characterized in that: In step S1, welding tasks are assigned based on welds. The quantity data in the work order includes weld name, weld length, weld type, weld form, part A material, part A thickness, part B material, part B thickness, welding process, and weld barcode. When performing welding operations, on-site construction personnel scan the barcode information in the work order and then perform welding operations.

3. The method for rapid prediction of ship welding quota working hours based on welding equipment data collection according to claim 1 is characterized in that: In step S1, the welding equipment operating parameters are collected, which means that during the entire welding process, various operating indicators of the welding machine are collected by the collection system to the server, including the welding machine's current, voltage, wire feeding time, wire feeding speed, welding speed, and welding gun swing.

4. The method for rapid prediction of ship welding quota working hours based on welding equipment data collection according to claim 1 is characterized in that: In said S2, the collected data is cleaned by setting the threshold values ​​of various parameters when the welding machine is operating, extracting the effective working hours from the collected data, removing the invalid non-working hours, and obtaining the actual working hours of each weld.

5. The method for rapid prediction of ship welding quota man-hours based on welding equipment data collection according to claim 1 is characterized in that: In step S3, taking the weld as the object, the weld length, welding type, welding posture, part A material, part A thickness, part B material, and part B thickness in the work order are matched with the actual working hours in step S2, and the data is preprocessed to form a data set of the material quantity, process, and working hours of each weld, and a welding information database is established.

6. The method for rapid prediction of ship welding quota man-hours based on welding equipment data collection according to claim 1 is characterized in that: After completing the training of the prediction model and parameter optimization, the characteristic attributes of the weld to be predicted are used as the input parameters of the model, and the working hours are used as the output, so as to obtain the predicted working hours of the weld to be predicted.

7. A rapid prediction system for ship welding quota working hours based on welding equipment data collection, characterized in that: include: The parameter collection module is configured to perform the following actions: taking each actual weld as an object, establishing a task dispatching system through IoT sensing technology, dispatching welding tasks based on the weld as the object, and collecting and storing the welding equipment operating parameters in the server; The data cleaning module is configured to perform the following actions: clean the collected welding equipment data to obtain the actual working hours of each weld; The welding information database module is configured to perform the following actions: establish a correlation between the obtained actual working time data and the material quantity data of the task dispatching system to form a data set of material quantity, process, and working time of each weld, and establish a welding information database; The model prediction module is configured to perform the following actions: train the historical data of the welding information database based on the extreme gradient boosting algorithm to predict the welding task operation duration, and correct the operation time predicted by the extreme gradient boosting algorithm based on the real-time welding operation time data of the predicted weld, iteratively optimize the model, and ultimately achieve the goal of continuously reducing the quota-to-actual ratio.

8. An electronic device, characterized in that: include: A memory and a processor, wherein the memory stores a computer program executable by the processor, and when the processor executes the computer program, the method for quickly predicting the ship welding quota working hours according to any one of claims 1 to 6 is implemented.

9. A storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is read and executed, the method for quickly predicting the ship welding quota working hours according to any one of claims 1 to 6 is implemented.

Images