Intelligent control method and system for steel plate cutting

Abstract

The present invention provides an intelligent control method and system for steel plate cutting, which relates to the field of intelligent control technology. A cutting factor indicator set is established, a historical cutting database is obtained to screen the factor indicators to obtain a target factor indicator set, and then a cutting depth prediction model is trained; a preset cutting scheme is obtained and a model analysis is performed to output a first cutting depth prediction result, and a first candidate scheme is screened in combination with a target cutting efficiency requirement threshold; a target cutting cost requirement threshold and a target cutting accuracy requirement threshold are obtained, and a scheme analysis is performed to determine a target cutting scheme, and steel plate cutting control is performed. This solves the technical problem that the screening method for steel plate cutting control schemes in the prior art is not intelligent enough, resulting in insufficient fit between the scheme and the current working conditions, resulting in low cutting efficiency and insufficient cutting accuracy. Scheme evaluation and screening are performed based on multi-dimensional indicators, and the optimal scheme is determined for steel plate cutting, which can effectively improve cutting efficiency and ensure cutting accuracy.

Description

Technical Field

[0001] The present invention relates to the field of intelligent control technology, and in particular to an intelligent control method and system for steel plate cutting. Background Art

[0002] Abrasive water jet steel plate cutting technology is widely used due to its advantages such as no open flame, high cutting efficiency, high environmental adaptability and small size, but it has certain disadvantages in cutting performance. Nowadays, conventional steel plate cutting methods are based on fixed processes, and cutting requirements are achieved by adjusting cutting parameters, ignoring the variables in the cutting process. Therefore, the key to solve the problem is to efficiently and accurately calculate the cutting parameters that are compatible with the current cutting conditions. The current technology still has certain flaws and needs further technological innovation.

[0003] In the prior art, the screening method for steel plate cutting control schemes is not intelligent enough, resulting in insufficient compatibility between the schemes and current working conditions, leading to low cutting efficiency and insufficient cutting accuracy. Summary of the invention

[0004] The present application provides an intelligent control method and system for steel plate cutting, which is used to solve the technical problem that the screening method for steel plate cutting control scheme in the prior art is not intelligent enough, resulting in insufficient compatibility between the scheme and the current working conditions, leading to low cutting efficiency and insufficient cutting accuracy.

[0005] In view of the above problems, the present application provides an intelligent control method and system for steel plate cutting.

[0006] In a first aspect, the present application provides an intelligent control method for steel plate cutting, the method comprising:

[0007] Analyze abrasive water jet steel plate cutting and establish a cutting factor index set, wherein the cutting factor index set includes multiple factor indicators;

[0008] Acquire a historical cutting database, and screen the multiple factor indicators based on the historical cutting database to obtain a target factor indicator set, wherein the target factor indicator set includes jet pressure, traverse speed, and target spacing;

[0009] Using the data in the historical cutting database to train a cutting depth prediction model;

[0010] Acquire a preset cutting scheme, wherein the preset cutting scheme is embedded with a first cutting constraint, and the first cutting constraint includes a first jet pressure, a first traverse speed, and a first target spacing;

[0011] Inputting the first jet pressure, the first traverse speed, and the first target spacing into the cutting depth prediction model to obtain a first cutting depth prediction result;

[0012] Obtaining a target cutting efficiency requirement threshold, and screening the preset cutting scheme in combination with the first cutting depth prediction result to obtain a first candidate scheme;

[0013] The target cutting cost requirement threshold and the target cutting accuracy requirement threshold are obtained in sequence, and the first candidate solution is analyzed to determine a target cutting solution, wherein the target cutting solution is used for steel plate cutting control.

[0014] In a second aspect, the present application provides an intelligent control system for steel plate cutting, the system comprising:

[0015] An index set building module, the index set building module is used to analyze abrasive water jet steel plate cutting and build a cutting factor index set, wherein the cutting factor index set includes a plurality of factor indicators;

[0016] An indicator screening module, the indicator screening module is used to obtain a historical cutting database, and screen the multiple factor indicators based on the historical cutting database to obtain a target factor indicator set, wherein the target factor indicator set includes jet pressure, traverse speed, and target spacing;

[0017] A model training module, wherein the model training module is used to train a cutting depth prediction model using the data in the historical cutting database;

[0018] A scheme acquisition module, wherein the scheme acquisition module is used to acquire a preset cutting scheme, wherein the preset cutting scheme is embedded with a first cutting constraint, and the first cutting constraint includes a first jet pressure, a first traverse speed, and a first target spacing;

[0019] A result prediction module, the result prediction module is used to input the first jet pressure, the first traverse speed and the first target spacing into the cutting depth prediction model to obtain a first cutting depth prediction result;

[0020] A scheme screening module, wherein the scheme screening module is used to obtain a target cutting efficiency requirement threshold, and screen the preset cutting scheme in combination with the first cutting depth prediction result to obtain a first candidate scheme;

[0021] A scheme determination module, the scheme determination module is used to sequentially obtain a target cutting cost requirement threshold and a target cutting accuracy requirement threshold, and analyze the first candidate scheme to determine a target cutting scheme, wherein the target cutting scheme is used for steel plate cutting control.

[0022] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0023] An intelligent control method for steel plate cutting provided in an embodiment of the present application analyzes abrasive water jet steel plate cutting and constructs a cutting factor index set, including multiple factor indexes, obtains a historical cutting database, and screens the multiple factor indexes to obtain a target factor index set, including jet pressure, lateral movement speed, and target spacing; uses the data in the historical cutting database to train a cutting depth prediction model; obtains a preset cutting scheme, the preset cutting scheme is embedded with a first cutting constraint, including a first jet pressure, a first lateral movement speed, and a first target spacing, and inputs the first jet pressure, the first lateral movement speed, and the first target spacing into the cutting depth prediction model , output the first cutting depth prediction result, obtain the target cutting efficiency requirement threshold, and screen the scheme in combination with the first cutting depth prediction result to obtain the first candidate scheme; obtain the target cutting cost requirement threshold and the target cutting accuracy requirement threshold in turn, perform scheme analysis to determine the target cutting scheme, and perform steel plate cutting control. This solves the problem that the screening method for steel plate cutting control schemes in the prior art is not intelligent enough, resulting in insufficient fit between the scheme and the current working conditions, leading to low cutting efficiency and insufficient cutting accuracy. This method evaluates and screens the scheme based on multi-dimensional indicators, and determines the optimal scheme for steel plate cutting, which can effectively improve cutting efficiency and ensure cutting accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 A schematic flow chart of an intelligent control method for steel plate cutting is provided for this application;

[0025] Figure 2 A schematic diagram of a process for obtaining a target factor index set in an intelligent control method for steel plate cutting is provided for this application;

[0026] Figure 3 A schematic diagram of a process for obtaining a cutting depth prediction model in an intelligent control method for steel plate cutting is provided for this application;

[0027] Figure 4 A schematic diagram of the structure of an intelligent control system for steel plate cutting is provided for this application.

[0028] Explanation of the accompanying drawings: indicator set building module 11, indicator screening module 12, model training module 13, solution acquisition module 14, result prediction module 15, solution screening module 16, solution determination module 17. DETAILED DESCRIPTION

[0029] The present application provides an intelligent control method and system for steel plate cutting, establishes a cutting factor indicator set, obtains a historical cutting database to screen the factor indicators to obtain a target factor indicator set, and then trains a cutting depth prediction model; obtains a preset cutting plan and performs model analysis to output a first cutting depth prediction result, and screens the plans in combination with a target cutting efficiency requirement threshold to obtain a first candidate plan; obtains a target cutting cost requirement threshold and a target cutting accuracy requirement threshold in turn, performs plan analysis to determine a target cutting plan, and performs steel plate cutting control, so as to solve the technical problems in the prior art that the screening method for steel plate cutting control plans is not intelligent enough, resulting in insufficient fit between the plan and the current working conditions, leading to low cutting efficiency and insufficient cutting accuracy.

[0030] Embodiment 1

[0031] like Figure 1 As shown, the present application provides an intelligent control method for steel plate cutting, the method comprising:

[0032] Step S100: analyzing abrasive water jet steel plate cutting and establishing a cutting factor index set, wherein the cutting factor index set includes a plurality of factor indexes;

[0033] Specifically, abrasive water jet steel plate cutting technology has a wide range of applications due to its advantages such as no open flame, high cutting efficiency, high environmental adaptability and small size. However, it has certain disadvantages in cutting performance. The key problem is to efficiently and accurately calculate the cutting parameters that are compatible with the current cutting conditions, which needs further technological innovation. The intelligent control method for steel plate cutting provided in the present application performs cutting impact analysis based on multi-dimensional indicators, and then models the scheme for analysis and screening to determine the optimal cutting scheme for target adaptability. Specifically, the abrasive water jet steel plate cutting process is evaluated, and hydraulic factors, abrasive factors and working condition factors are used as analysis dimensions. Factor indicators are identified respectively to determine the influencing factors in the equipment cutting process, and the determined multiple factor indicators are dimensionally attributed and integrated to form the cutting factor indicator set, that is, the set of indicators that affect the cutting depth. The cutting factor indicator set is a preliminary extracted factor indicator, and further screening and analysis are performed on this basis.

[0034] Furthermore, the analysis of abrasive water jet steel plate cutting and the establishment of a cutting factor index set, step S100 of the present application also includes:

[0035] Step S110: sequentially obtaining a hydraulic factor index set, an abrasive factor index set, and a working condition factor index set for the abrasive water jet steel plate cutting;

[0036] Step S120: wherein the hydraulic factor index set includes jet pressure and nozzle diameter;

[0037] Step S130: wherein the abrasive factor index set includes abrasive type, abrasive size, abrasive shape, and abrasive ratio;

[0038] Step S140: wherein the working condition factor index set includes traverse speed, target spacing, and impact strength;

[0039] Step S150: performing a union operation on the hydraulic factor index set, the abrasive factor index set and the working condition factor index set to obtain the cutting factor index set.

[0040] Specifically, the hydraulic factor is used as the indicator collection dimension. The jet pressure is proportional to the cutting depth, and the nozzle diameter will affect the pump pressure to a certain extent. The jet pressure and the nozzle diameter are used as the hydraulic factor indicators. The specific characteristic attributes of abrasives are different, and the corresponding cutting parameters are different. The abrasive type, abrasive size, abrasive shape and abrasive ratio are used as the original information to be cut, as the abrasive factor indicator set, and the abrasive factor indicator set will affect the specific working conditions. The traverse speed and cutting depth show a negatively correlated indicator attenuation relationship. The target spacing and the impact strength are indicators that affect the cutting depth. The indicators are integrated to generate the working condition factor indicator set. Based on the hydraulic factor indicator set, the abrasive factor indicator set and the working condition factor indicator set, the indicators are combined, and the indicator combination result is used as the cutting factor indicator set. Indicator analysis and integration are performed based on multiple dimensions to ensure the accuracy and pertinence of the extracted indicators and improve the completeness of indicator coverage.

[0041] Step S200: obtaining a historical cutting database, and screening the multiple factor indicators based on the historical cutting database to obtain a target factor indicator set, wherein the target factor indicator set includes jet pressure, traverse speed, and target spacing;

[0042] Furthermore, if Figure 2 As shown, the historical cutting database is obtained, and the multiple factor indicators are screened based on the historical cutting database to obtain a target factor indicator set. Step S200 of the present application also includes:

[0043] Step S210: the historical cutting database includes multiple groups of cutting data, and the first cutting data from the multiple groups of cutting data is extracted;

[0044] Step S220: obtaining a first factor indicator and a second factor indicator from the plurality of factor indicators;

[0045] Step S230: sequentially obtaining a first indicator parameter of the first factor indicator and a second indicator parameter of the second factor indicator in the first cutting data;

[0046] Step S240: obtaining a first cutting depth in the first cutting data;

[0047] Step S250: respectively calculating a first correlation between the first index parameter and the first cutting depth, and a second correlation between the second index parameter and the first cutting depth;

[0048] Step S260: sorting the first relevance degree and the second relevance degree in descending order to obtain a target descending list;

[0049] Step S270: extracting the target descending list based on a preset sorting threshold, and reversely matching the corresponding factor indicators to form the target factor indicator set.

[0050] Specifically, multiple groups of historical cutting data are collected to construct the historical cutting database, and further the index parameters are extracted and the correlation analysis of cutting depth is performed for the factor indexes respectively, and the factor indexes that meet the threshold standard are screened as the target factor index set, including the jet pressure, the lateral movement speed and the target spacing, and the corresponding influence degrees decrease in sequence.

[0051] Specifically, a preset time period is set, that is, a time interval for data collection, cutting data collection is performed based on the preset time period, multiple groups of cutting data are obtained, the multiple groups of cutting data are integrated and regularized, and the historical cutting database is generated. Based on the historical cutting database, the historical cutting database is classified into two categories, and a group of data is randomly extracted as the first cutting data. Extraction is performed based on the multiple factor indicators to obtain the first factor indicator and the second factor indicator, wherein the value of the factor indicator is not limited to two, and is consistent with the currently determined number of indicators, which is used to distinguish the factor indicators. The first cutting data is traversed, and the first factor indicator and the second factor indicator are used as index directions to extract the first indicator parameter and the second indicator parameter, wherein the value of the indicator parameter includes but is not limited to the current number, and is subject to the actual extracted number. Further based on the first cutting data, the first cutting depth is extracted.

[0052] Then, the first index parameter and the second index parameter are respectively calculated for the cutting depth correlation. For example, the influence degree on the cutting depth can be determined by verifying the single index parameter. The higher the influence degree, the higher the corresponding correlation degree. The first correlation degree and the second correlation degree are determined, wherein the number of correlation degrees is consistent with the number of index parameters and the number of factor indicators. For example, the correlation degree of the index parameters is calculated based on the grey correlation algorithm. The first correlation degree and the second correlation degree are arranged in descending order to generate a correlation degree sequence as the target descending list. By extracting historical cutting data for analysis, accidental conditions are avoided, the universality and accuracy of the results are guaranteed, and the actual fit of the results is improved. The preset sorting threshold is further configured, that is, the critical value for limiting the correlation degree. Based on the preset sorting threshold, the target descending sequence is limited and intercepted, and multiple correlation degrees with higher values ​​are extracted, and the corresponding factor indicators are reversely matched to form the target factor indicator set. The target factor indicator set is a plurality of factor indicators with higher influence on the cutting depth. Subsequent cutting analysis is performed based on the target factor indicator set to screen out low-impact factor indicators and improve the efficiency of subsequent analysis.

[0053] Step S300: using the data in the historical cutting database to train and obtain a cutting depth prediction model;

[0054] Furthermore, if Figure 3 As shown, the cutting depth prediction model is obtained by training with the data in the historical cutting database, and step S300 of the present application also includes:

[0055] Step S310: extracting second cutting data from the multiple sets of cutting data;

[0056] Step S320: wherein the second cutting data includes a second jet pressure, a second traverse speed, a second target spacing, and a second cutting depth;

[0057] Step S330: using the second jet pressure, the second traverse speed, the second target spacing, and the second cutting depth as training data;

[0058] Step S340: Divide the training data into a first data group and a second data group;

[0059] Step S350: training the first data set to obtain a first model, and training the second data set to obtain a second model;

[0060] Step S360: fusing the first model and the second model to obtain the cutting depth prediction model.

[0061] Specifically, based on the historical cutting database, sample cutting data is extracted, data mapping association is performed to determine the constructed data, and neural network training is further performed to generate the cutting depth prediction model.

[0062] Specifically, based on the historical cutting database, the second cutting data in the multiple groups of cutting data is extracted, wherein there is no union interval between the first cutting data and the second cutting data. Factor indicator parameters are extracted based on the second cutting data to determine the two jet pressures, the second lateral movement speed, the second target spacing and the second cutting depth, and mapping and associating the factor indicator parameters are performed to generate multiple parameter sequences as the training data, wherein different parameters have obvious differences in the effects on the cutting depth. The training data is classified into two categories to obtain the first data group and the second data group, wherein in any parameter sequence, the corresponding second jet pressure, the second lateral movement speed and the second target spacing are node identification data, and the second cutting depth is node decision data. Neural network training is performed based on the first data group to generate the first model; neural network training is performed based on the second data group to generate the second model.

[0063] Among them, the first model and the second model are trained in the same way, but with different training data, and the final model operation mechanism is different. The first model and the second model are fused based on the integrated fusion method, and a strong learner is constructed based on multiple weak learners to optimize the model performance, and then the dimensional analysis method is assisted to improve the optimization model operation mechanism to form the cutting depth prediction model. The cutting depth prediction model has higher analysis efficiency and output accuracy, and the cutting plan is predicted based on the cutting depth prediction model.

[0064] Step S400: obtaining a preset cutting scheme, wherein the preset cutting scheme is embedded with a first cutting constraint, and the first cutting constraint includes a first jet pressure, a first traverse speed, and a first target spacing;

[0065] Step S500: inputting the first jet pressure, the first traverse speed and the first target spacing into the cutting depth prediction model to obtain a first cutting depth prediction result;

[0066] Specifically, by conducting a big data survey, a plurality of currently applied cutting schemes are statistically integrated as the preset cutting scheme. The preset cutting scheme is embedded with the first cutting constraint, that is, a plurality of screened factor indicators, including the first jet pressure, the first lateral movement speed and the first target spacing, wherein the cutting depth is positively correlated with the jet pressure and negatively correlated with the lateral movement speed, and increases first and then decreases as the target distance increases. Furthermore, the first jet pressure, the first lateral movement speed and the first target spacing are extracted and mapped for each scheme in the preset cutting scheme, and are input into the cutting depth prediction model. By performing hierarchical data recognition matching and decision mapping, the corresponding cutting depth prediction result is directly output, and the cutting depth prediction result is correspondingly marked based on the preset cutting scheme to generate the first cutting depth prediction result, that is, the cutting depth per second. The acquisition of the first cutting depth prediction result provides a reference basis for subsequent scheme screening.

[0067] Step S600: obtaining a target cutting efficiency requirement threshold, and screening the preset cutting scheme in combination with the first cutting depth prediction result to obtain a first candidate scheme;

[0068] Step S700: sequentially obtaining a target cutting cost requirement threshold and a target cutting accuracy requirement threshold, and analyzing the first candidate solution to determine a target cutting solution, wherein the target cutting solution is used for steel plate cutting control.

[0069] Specifically, the first cutting depth prediction result is traversed to extract the cutting efficiency of each scheme in the preset cutting scheme. The target cutting efficiency requirement threshold is further determined, that is, the critical value for limiting the cutting efficiency, and it is determined whether the cutting efficiency of each scheme meets the target cutting efficiency requirement threshold, and multiple cutting efficiencies greater than or equal to the target cutting efficiency requirement threshold are extracted. By performing reverse matching of the cutting schemes, multiple cutting schemes that meet the cutting efficiency requirements in the preset cutting schemes are determined as the first candidate schemes, and the first candidate schemes are the primary screening scheme set.

[0070] Furthermore, cutting cost and cutting accuracy are used as evaluation criteria, and each of the first candidate solutions is evaluated to determine the evaluation result of the first candidate solution. The target cutting cost requirement threshold and the target cutting accuracy requirement threshold are further configured, that is, the critical value of the parameters for measuring cutting cost and cutting accuracy is defined, and the evaluation result of the first candidate solution is threshold determined, and the solution that meets the target cutting cost requirement threshold and the target cutting accuracy requirement threshold is screened to determine the preferred cutting solution, and the solutions are further ranked to extract the first cutting solution, that is, the best cutting solution, as the target cutting solution, and the cutting control of the steel plate is performed based on the target cutting solution to achieve efficient and accurate cutting of the steel plate.

[0071] Furthermore, the target cutting efficiency requirement threshold is obtained, and the preset cutting scheme is screened in combination with the first cutting depth prediction result to obtain a first candidate scheme. Step S600 of the present application also includes:

[0072] Step S610: The preset cutting scheme includes a unidirectional single cutting scheme, a bidirectional single cutting scheme, a unidirectional double cutting scheme, and a bidirectional double cutting scheme;

[0073] Step S620: According to the first cutting depth prediction result, respectively obtaining a first cutting efficiency of the unidirectional single cutting scheme, a second cutting efficiency of the bidirectional single cutting scheme, a third cutting efficiency of the unidirectional double cutting scheme, and a fourth cutting efficiency of the bidirectional double cutting scheme;

[0074] Step S630: Compare the first cutting efficiency, the second cutting efficiency, the third cutting efficiency and the fourth cutting efficiency according to the target cutting efficiency requirement threshold to obtain the first candidate solution.

[0075] Specifically, the preset cutting scheme includes a variety of cutting modes, including single cutting, multiple cutting in the same direction, and multiple cutting in alternating directions. Based on the above-mentioned multiple cutting modes, the preset cutting scheme is divided to determine the unidirectional single cutting scheme, the bidirectional single cutting scheme, the unidirectional double cutting scheme, and the bidirectional double cutting scheme, wherein the bidirectional double cutting is a double cutting operation performed in alternating directions, and there are differences in the adapted cutting schemes under different cutting conditions. The first cutting depth prediction result is traversed, and the results of the unidirectional single cutting scheme, the bidirectional single cutting scheme, the unidirectional double cutting scheme, and the bidirectional double cutting scheme are matched respectively to determine the first cutting efficiency, the second cutting efficiency, the third cutting efficiency, and the fourth cutting efficiency, wherein the cutting effect is additional annotation information. The target cutting efficiency requirement threshold is obtained, that is, the critical efficiency for limiting the cutting efficiency. Based on the target cutting efficiency requirement threshold, the first cutting efficiency, the second cutting efficiency, the third cutting efficiency and the fourth cutting efficiency are proofread and determined, and the cutting efficiency that meets the target cutting efficiency requirement threshold is extracted. The schemes are reverse matched to determine multiple schemes that meet the threshold standard in the preset cutting schemes as the first candidate schemes, wherein the first candidate schemes include a single scheme or multiple schemes. The schemes are initially screened based on the cutting efficiency to accurately identify the cutting schemes that meet the cutting efficiency standards.

[0076] Further, before sequentially obtaining the target cutting cost requirement threshold and the target cutting accuracy requirement threshold, analyzing the first candidate solution, and determining the target cutting solution, wherein the target cutting solution is used for steel plate cutting control, step S700 of the present application further includes:

[0077] Step S710: forming a preset solution evaluation feature set, wherein the preset solution evaluation feature set includes solution cost and solution accuracy;

[0078] Step S720: evaluating and analyzing the first candidate solution according to the preset solution evaluation feature set to obtain a first candidate solution evaluation result;

[0079] Step S730: wherein the evaluation result of the first candidate solution includes a solution cost evaluation result and a solution accuracy evaluation result.

[0080] Furthermore, the first candidate solution is evaluated and analyzed according to the preset solution evaluation feature set to obtain a first candidate solution evaluation result. Step S720 of the present application also includes:

[0081] Step S721: obtaining a first solution test of the first candidate solution;

[0082] Step S722: Obtaining a first test record of the first solution test;

[0083] Step S723: analyzing the first test record according to the preset scheme evaluation feature set to obtain a first test cost and a first test accuracy in sequence;

[0084] Step S724: taking the first test cost and the first test accuracy as the first solution evaluation result;

[0085] Step S725: Based on the evaluation result of the first solution, form the evaluation result of the first candidate solution.

[0086] Specifically, the first candidate solution is comprehensively evaluated to determine the cutting solution with the best cutting effect and the highest degree of fit with the current working conditions. The solution cost and the solution accuracy are used as evaluation feature directions to form the preset solution evaluation feature set. Each solution in the first candidate solution is evaluated and analyzed based on the solution cost and the solution accuracy to generate the evaluation result of the first candidate solution.

[0087] Specifically, the scheme evaluation is performed by conducting experimental analysis to improve the actual fit of the evaluation result. Based on the first candidate scheme, a scheme is randomly selected as the first scheme, and a trial cutting is performed based on the first scheme to determine the first scheme experiment. Preferably, based on the control variable method, the unidirectional single cutting scheme, the unidirectional double cutting scheme, the bidirectional single cutting scheme and the bidirectional double cutting scheme are respectively tested. Exemplarily, the test execution can be performed by constructing a simulated cutting model to avoid test loss and improve the test efficiency. As the first test experiment progresses, the first test record is obtained. Exemplarily, the steel plate is cut from a determined direction, and its cutting depth per second is counted to obtain the cutting efficiency; the cutting energy consumption analysis is performed, such as the power consumption of unidirectional cutting, and the cutting cost is calculated; after the cutting is completed, the test detection is performed to determine the cutting accuracy of the scheme. The first test record is traversed, and the mapping record data of the test cost and the test accuracy are extracted to determine the first test cost and the first test accuracy. The first test cost and the first test accuracy are integrated and summarized as the evaluation result of the first scheme. Each scheme in the first candidate scheme is evaluated and analyzed separately, and the corresponding scheme evaluation result is determined and associated with the mapping scheme to generate the evaluation result of the first candidate scheme.

[0088] The evaluation result of the first candidate solution includes multiple evaluation result sequences, namely solution-solution cost evaluation result-solution accuracy evaluation result. By conducting a test evaluation on the first candidate solution, the accuracy of the evaluation result is guaranteed and the existence of evaluation errors is reduced. Then, the evaluation result of the first candidate solution is used as a screening basis to select and determine the best cutting solution.

[0089] Embodiment 2

[0090] Based on the same inventive concept as the intelligent control method for steel plate cutting in the aforementioned embodiment, Figure 4 As shown, the present application provides an intelligent control system for steel plate cutting, the system comprising:

[0091] An index set building module 11, the index set building module 11 is used to analyze abrasive water jet steel plate cutting and build a cutting factor index set, wherein the cutting factor index set includes a plurality of factor indicators;

[0092] An indicator screening module 12, the indicator screening module 12 is used to obtain a historical cutting database, and screen the multiple factor indicators based on the historical cutting database to obtain a target factor indicator set, wherein the target factor indicator set includes jet pressure, traverse speed, and target spacing;

[0093] A model training module 13, wherein the model training module 13 is used to train a cutting depth prediction model using the data in the historical cutting database;

[0094] A scheme acquisition module 14, wherein the scheme acquisition module 14 is used to acquire a preset cutting scheme, wherein the preset cutting scheme is embedded with a first cutting constraint, and the first cutting constraint includes a first jet pressure, a first traverse speed, and a first target spacing;

[0095] A result prediction module 15, wherein the result prediction module 15 is used to input the first jet pressure, the first traverse speed and the first target spacing into the cutting depth prediction model to obtain a first cutting depth prediction result;

[0096] A scheme screening module 16, wherein the scheme screening module 16 is used to obtain a target cutting efficiency requirement threshold, and screen the preset cutting scheme in combination with the first cutting depth prediction result to obtain a first candidate scheme;

[0097] The scheme determination module 17 is used to sequentially obtain a target cutting cost requirement threshold and a target cutting accuracy requirement threshold, and analyze the first candidate scheme to determine a target cutting scheme, wherein the target cutting scheme is used for steel plate cutting control.

[0098] Furthermore, the system further comprises:

[0099] An index acquisition module, the index acquisition module is used to sequentially acquire a hydraulic factor index set, an abrasive factor index set, and a working condition factor index set of the abrasive water jet steel plate cutting;

[0100] A hydraulic factor index set analysis module, wherein the hydraulic factor index set analysis module is used, and the hydraulic factor index set includes jet pressure and nozzle diameter;

[0101] An abrasive factor index set analysis module, wherein the abrasive factor index set analysis module is used, wherein the abrasive factor index set includes abrasive type, abrasive size, abrasive shape, and abrasive ratio;

[0102] A working condition factor index set analysis module, wherein the working condition factor index set analysis module is used, and the working condition factor index set includes traverse speed, target spacing, and impact strength;

[0103] A cutting factor index set acquisition module is used to perform a union operation on the hydraulic factor index set, the abrasive factor index set and the working condition factor index set to obtain the cutting factor index set.

[0104] Furthermore, the system further comprises:

[0105] A first cutting data extraction module, wherein the historical cutting database includes a plurality of groups of cutting data, and the first cutting data extraction module is used to extract first cutting data from the plurality of groups of cutting data;

[0106] A factor indicator acquisition module, the factor indicator acquisition module is used to acquire a first factor indicator and a second factor indicator from the multiple factor indicators;

[0107] An indicator parameter acquisition module, the indicator parameter acquisition module is used to sequentially acquire a first indicator parameter of the first factor indicator and a second indicator parameter of the second factor indicator in the first cutting data;

[0108] A cutting depth acquisition module, the cutting depth acquisition module is used to acquire a first cutting depth in the first cutting data;

[0109] a correlation calculation module, the correlation calculation module being used to respectively calculate a first correlation between the first index parameter and the first cutting depth, and a second correlation between the second index parameter and the first cutting depth;

[0110] A list generation module, the list generation module is used to descend the first relevance degree and the second relevance degree to obtain a target descending list;

[0111] An indicator matching module is used to extract the target descending list based on a preset sorting threshold, and reversely match the corresponding factor indicators to form the target factor indicator set.

[0112] Furthermore, the system further comprises:

[0113] A second cutting data extraction module, the second cutting data extraction module is used to extract second cutting data from the multiple groups of cutting data;

[0114] A second cutting data analysis module, wherein the second cutting data analysis module is used, wherein the second cutting data includes a second jet pressure, a second traverse speed, a second target spacing, and a second cutting depth;

[0115] a training data determination module, the training data determination module being used to use the second jet pressure, the second traverse speed, the second target spacing, and the second cutting depth as training data;

[0116] A data partitioning module, the data partitioning module is used to partition the training data into a first data group and a second data group;

[0117] A model training module, the model training module is used to train the first data group to obtain a first model, and train the second data group to obtain a second model;

[0118] A model fusion module is used to fuse the first model and the second model to obtain the cutting depth prediction model.

[0119] Furthermore, the system further comprises:

[0120] A scheme analysis module, wherein the preset cutting schemes include a unidirectional single-cutting scheme, a bidirectional single-cutting scheme, a unidirectional double-cutting scheme, and a bidirectional double-cutting scheme;

[0121] a cutting efficiency acquisition module, the cutting efficiency acquisition module being used to respectively acquire a first cutting efficiency of the unidirectional single cutting scheme, a second cutting efficiency of the bidirectional single cutting scheme, a third cutting efficiency of the unidirectional double cutting scheme, and a fourth cutting efficiency of the bidirectional double cutting scheme according to the first cutting depth prediction result;

[0122] An efficiency comparison module is used to compare the first cutting efficiency, the second cutting efficiency, the third cutting efficiency and the fourth cutting efficiency according to the target cutting efficiency requirement threshold to obtain the first candidate solution.

[0123] Furthermore, the system further comprises:

[0124] A feature set building module, the feature set building module is used to build a preset solution evaluation feature set, wherein the preset solution evaluation feature set includes solution cost and solution accuracy;

[0125] A solution evaluation module, the solution evaluation module is used to evaluate and analyze the first candidate solution according to the preset solution evaluation feature set to obtain an evaluation result of the first candidate solution;

[0126] An evaluation result analysis module is used in which the evaluation result of the first candidate solution includes a solution cost evaluation result and a solution accuracy evaluation result.

[0127] Furthermore, the system further comprises:

[0128] A scheme experiment acquisition module, the scheme experiment acquisition module is used to acquire a first scheme experiment of the first candidate scheme;

[0129] An experiment record acquisition module, the experiment record acquisition module is used to acquire a first experiment record of the first scheme experiment;

[0130] An experimental parameter analysis module, the experimental parameter analysis module is used to analyze the first test record according to the preset scheme evaluation feature set, and obtain a first test cost and a first test accuracy in sequence;

[0131] An evaluation result acquisition module, wherein the evaluation result acquisition module is used to use the first test cost and the first test accuracy as the first scheme evaluation result;

[0132] A candidate solution evaluation result composition module is used to compose the first candidate solution evaluation result according to the first solution evaluation result.

[0133] Through the above-mentioned detailed description of an intelligent control method for steel plate cutting in this specification, those skilled in the art can clearly know an intelligent control method and system for steel plate cutting in this embodiment. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the description of the method part.

[0134] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An intelligent control method for steel plate cutting, characterized in that: include: Analyze abrasive water jet steel plate cutting and establish a cutting factor index set, wherein the cutting factor index set includes multiple factor indicators; Acquire a historical cutting database, and screen the multiple factor indicators based on the historical cutting database to obtain a target factor indicator set, wherein the target factor indicator set includes jet pressure, traverse speed, and target spacing; Using the data in the historical cutting database to train a cutting depth prediction model; Acquire a preset cutting scheme, wherein the preset cutting scheme is embedded with a first cutting constraint, and the first cutting constraint includes a first jet pressure, a first traverse speed, and a first target spacing; Inputting the first jet pressure, the first traverse speed, and the first target spacing into the cutting depth prediction model to obtain a first cutting depth prediction result; Obtaining a target cutting efficiency requirement threshold, and screening the preset cutting scheme in combination with the first cutting depth prediction result to obtain a first candidate scheme; Obtaining a target cutting cost requirement threshold and a target cutting accuracy requirement threshold in sequence, and analyzing the first candidate solution to determine a target cutting solution, wherein the target cutting solution is used for steel plate cutting control; The method of using the data in the historical cutting database to train a cutting depth prediction model comprises: extracting second cutting data from the plurality of sets of cutting data; Wherein, the second cutting data includes a second jet pressure, a second traverse speed, a second target spacing, and a second cutting depth; using the second jet pressure, the second traverse speed, the second target spacing, and the second cutting depth as training data; Dividing the training data into a first data group and a second data group; Training the first data set to obtain a first model, and training the second data set to obtain a second model; The cutting depth prediction model is obtained by fusing the first model and the second model.

2. The intelligent control method according to claim 1, characterized in that: The analysis of abrasive water jet steel plate cutting and the establishment of a cutting factor index set include: Sequentially obtain a hydraulic factor index set, an abrasive factor index set, and a working condition factor index set for abrasive water jet steel plate cutting; Wherein, the hydraulic factor index set includes jet pressure and nozzle diameter; The abrasive factor index set includes abrasive type, abrasive size, abrasive shape, and abrasive ratio; The working condition factor index set includes traverse speed, target spacing, and impact strength; A union operation is performed on the hydraulic factor index set, the abrasive factor index set and the working condition factor index set to obtain the cutting factor index set.

3. The intelligent control method according to claim 2, characterized in that: The acquiring of the historical cutting database and screening the multiple factor indicators based on the historical cutting database to obtain a target factor indicator set includes: The historical cutting database includes multiple groups of cutting data, and the first cutting data from the multiple groups of cutting data is extracted; Obtaining a first factor indicator and a second factor indicator from among the multiple factor indicators; Sequentially acquiring a first indicator parameter of the first factor indicator and a second indicator parameter of the second factor indicator in the first cutting data; Acquire a first cutting depth in the first cutting data; respectively calculating a first correlation between the first index parameter and the first cutting depth, and a second correlation between the second index parameter and the first cutting depth; Descending the first relevance degree and the second relevance degree to obtain a target descending list; The target descending list is extracted based on a preset sorting threshold, and the corresponding factor indicators are reversely matched to form the target factor indicator set.

4. The intelligent control method according to claim 1, characterized in that: The obtaining of the target cutting efficiency requirement threshold and screening the preset cutting scheme in combination with the first cutting depth prediction result to obtain a first candidate scheme include: The preset cutting schemes include a unidirectional single cutting scheme, a bidirectional single cutting scheme, a unidirectional double cutting scheme, and a bidirectional double cutting scheme; According to the first cutting depth prediction result, respectively obtaining a first cutting efficiency of the unidirectional single cutting scheme, a second cutting efficiency of the bidirectional single cutting scheme, a third cutting efficiency of the unidirectional double cutting scheme, and a fourth cutting efficiency of the bidirectional double cutting scheme; The first cutting efficiency, the second cutting efficiency, the third cutting efficiency and the fourth cutting efficiency are compared according to the target cutting efficiency requirement threshold to obtain the first candidate solution.

5. The intelligent control method according to claim 1, characterized in that: The target cutting cost requirement threshold and the target cutting accuracy requirement threshold are obtained in sequence, and the first candidate solution is analyzed to determine the target cutting solution, wherein the target cutting solution is used for steel plate cutting control before the target cutting solution is used, including: Establishing a preset scheme evaluation feature set, wherein the preset scheme evaluation feature set includes scheme cost and scheme accuracy; Evaluate and analyze the first candidate solution according to the preset solution evaluation feature set to obtain an evaluation result of the first candidate solution; Among them, the evaluation result of the first candidate solution includes a solution cost evaluation result and a solution accuracy evaluation result.

6. The intelligent control method according to claim 5, characterized in that: The step of evaluating and analyzing the first candidate solution according to the preset solution evaluation feature set to obtain the first candidate solution evaluation result includes: Obtain a first solution test of the first candidate solution; Obtain a first test record of the first scheme test; Analyze the first test record according to the preset scheme evaluation feature set to obtain a first test cost and a first test accuracy in sequence; Taking the first test cost and the first test accuracy as the first scheme evaluation result; Based on the evaluation result of the first solution, the evaluation result of the first candidate solution is formed.

7. An intelligent control system for steel plate cutting, characterized in that: include: An index set building module, the index set building module is used to analyze abrasive water jet steel plate cutting and build a cutting factor index set, wherein the cutting factor index set includes a plurality of factor indicators; An indicator screening module, the indicator screening module is used to obtain a historical cutting database, and screen the multiple factor indicators based on the historical cutting database to obtain a target factor indicator set, wherein the target factor indicator set includes jet pressure, traverse speed, and target spacing; A model training module, wherein the model training module is used to train a cutting depth prediction model using the data in the historical cutting database; A scheme acquisition module, wherein the scheme acquisition module is used to acquire a preset cutting scheme, wherein the preset cutting scheme is embedded with a first cutting constraint, and the first cutting constraint includes a first jet pressure, a first traverse speed, and a first target spacing; A result prediction module, the result prediction module is used to input the first jet pressure, the first traverse speed and the first target spacing into the cutting depth prediction model to obtain a first cutting depth prediction result; A scheme screening module, wherein the scheme screening module is used to obtain a target cutting efficiency requirement threshold, and screen the preset cutting scheme in combination with the first cutting depth prediction result to obtain a first candidate scheme; A scheme determination module, the scheme determination module is used to sequentially obtain a target cutting cost requirement threshold and a target cutting accuracy requirement threshold, and analyze the first candidate scheme to determine a target cutting scheme, wherein the target cutting scheme is used for steel plate cutting control; The system further comprises: A second cutting data extraction module, the second cutting data extraction module is used to extract second cutting data from the multiple groups of cutting data; A second cutting data analysis module, wherein the second cutting data analysis module is used, wherein the second cutting data includes a second jet pressure, a second traverse speed, a second target spacing, and a second cutting depth; a training data determination module, the training data determination module being used to use the second jet pressure, the second traverse speed, the second target spacing, and the second cutting depth as training data; A data partitioning module, the data partitioning module is used to partition the training data into a first data group and a second data group; A model training module, the model training module is used to train the first data group to obtain a first model, and train the second data group to obtain a second model; A model fusion module is used to fuse the first model and the second model to obtain the cutting depth prediction model.

Images