A process parameter intelligent optimization method, device, equipment and medium
By constructing a cutting force and vibration mapping model and optimizing process parameters using multi-source data and genetic algorithms, the deformation problem in the machining of weakly rigid structural parts for aerospace equipment was solved, achieving efficient and low-loss machining results.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SICHUAN AEROSPACE FENGHUO SERVO CONTROL TECH CO LTD
- Filing Date
- 2026-02-06
- Publication Date
- 2026-06-05
AI Technical Summary
Weakly rigid structural components in aerospace equipment are prone to deformation during processing. Existing technologies make it difficult to effectively optimize process parameters, resulting in high product scrap rates and low processing efficiency.
By constructing a cutting force and cutting vibration mapping model, multi-source machining data and genetic algorithms are used for multi-objective nonlinear optimization. Combined with the TOPSIS algorithm, the globally optimal process parameters are selected, reducing the dependence on simulation computing power and reducing the number of trial cuts.
It achieves low-stress, low-vibration, and high-efficiency processing, reduces product scrap rate, and improves processing prediction accuracy and efficiency.
Smart Images
Figure CN122154409A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of industrial processing technology, and provides a method, apparatus, equipment and medium for intelligent optimization of process parameters. Background Technology
[0002] Currently, aerospace equipment widely employs weakly rigid structural components, which are typically composed of numerous thin-walled webs and side plates. Because their wall thickness is much smaller than the overall outline dimensions, their structural rigidity is poor, resulting in high material removal rates during machining. They are significantly affected by cutting forces and vibrations, making them highly susceptible to machining deformation, which severely impacts the dimensional accuracy and mechanical properties of the product. Therefore, current research on process parameter optimization often faces the following problems: I. Machining is a complex nonlinear process involving multiple physical fields such as thermo-mechanical coupling, elastoplastic deformation, creep, fracture, and friction. Studying it using CAE simulation or experimental methods is not only technically challenging and costly, but also inefficient.
[0003] Second, in actual production, the process parameters are often optimized by relying on experience or by measuring while processing, which often requires multiple trial cuts, which can easily lead to an increase in product scrap rate and a decrease in production efficiency. Summary of the Invention
[0004] This application provides a method, apparatus, equipment, and medium for intelligent optimization of process parameters, which can solve technical problems such as high product scrap rate and low processing efficiency in the prior art.
[0005] On the one hand, a method for intelligent optimization of process parameters is provided, the method comprising: Based on multi-source machining data, a cutting force mapping model and a cutting vibration mapping model are constructed; wherein, the multi-source machining data is obtained through vibration sensors, triaxial force gauges, current sensors, and secondary development libraries; Using a pre-defined genetic algorithm and pre-defined constraints, a multi-objective nonlinear optimization is performed on the cutting force mapping model and the cutting vibration mapping model to obtain the Pareto solution set of the process parameters; The Pareto solution set is optimized by using the preset TOPSIS algorithm to obtain the globally optimal process parameters.
[0006] Optionally, the step of constructing a cutting force mapping model and a cutting vibration mapping model based on multi-source machining data includes: Multiple orthogonal experiments were conducted using side milling with large depth of cut and small width of cut and end milling with large width of cut and small depth of cut to obtain multiple process parameters and multiple cutting forces for side milling and end milling. Multiple process parameters and multiple cutting forces for side milling and end milling are classified, summarized, and fitted to construct a cutting force mapping model.
[0007] Optionally, the step of constructing a cutting force mapping model and a cutting vibration mapping model based on multi-source machining data includes: The initial BP neural network model is trained using the multi-source machining data and the Levenberg-Marquardt method to obtain the cutting vibration mapping model; wherein, the initial BP neural network model has 4 hidden layers and the activation function is the Sigmoid function.
[0008] Optionally, the step of using a preset genetic algorithm and preset constraints to perform multi-objective nonlinear optimization on the cutting force mapping model and the cutting vibration mapping model to obtain the Pareto solution set of the process parameters includes: Under constraints of machine tool spindle speed, feed rate, power, tool, and cutting width, a pre-defined genetic algorithm is used to perform multi-objective nonlinear optimization on the cutting force mapping model and the cutting vibration mapping model to obtain the Pareto solution set of the process parameters.
[0009] Optionally, before constructing the cutting force mapping model and the cutting vibration mapping model based on multi-source machining data, the method further includes: A cutting data acquisition platform is established; wherein, the acquisition platform includes a vibration sensor, a triaxial force gauge, a current sensor, and a secondary development library; A three-dimensional force gauge was used to collect cutting force. Vibration signals are acquired using vibration sensors. Current is collected using a current sensor; Read the status and real-time information of the CNC machining center from the secondary development library.
[0010] Optionally, the cutting force mapping model is represented by the following formula:
[0011] in, , , These are three-dimensional cutting forces. The resultant force is the cutting force; , f and n are the axial depth of cut, radial depth of cut, feed per tooth, and spindle speed, respectively; x1, x2, x3, x4, y1, y2, y3, y4, z1, z2, z3, z4, h1, h2, h3, and h4 are the coefficients to be determined during fitting.
[0012] Optionally, the globally optimal process parameters are expressed using the following formula:
[0013] in, For the i-th globally optimal process parameter, The distance to the i-th positive ideal value. It is the distance to the i-th negative ideal value.
[0014] On the one hand, a process parameter intelligent optimization device is provided, the device comprising: The model building unit is used to construct a cutting force mapping model and a cutting vibration mapping model based on multi-source machining data; wherein, the multi-source machining data is obtained through vibration sensors, triaxial force gauges, current sensors, and secondary development libraries; The process parameter optimization unit is used to perform multi-objective nonlinear optimization of the cutting force mapping model and the cutting vibration mapping model using a preset genetic algorithm and preset constraints, so as to obtain the Pareto solution set of the process parameters. The optimal process parameter acquisition unit is used to select the optimal parameters from the Pareto solution set using a preset TOPSIS algorithm to obtain the globally optimal process parameters.
[0015] On one hand, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement any of the methods described above.
[0016] On the one hand, a storage medium is provided that stores computer program instructions thereon, which, when executed by a processor, implement any of the methods described above.
[0017] Compared with the prior art, the beneficial effects of this application are as follows: In this application, when performing intelligent optimization of process parameters, firstly, a cutting force mapping model and a cutting vibration mapping model can be constructed based on multi-source machining data; wherein, the multi-source machining data is obtained through vibration sensors, triaxial force gauges, current sensors, and secondary development libraries; then, a preset genetic algorithm and preset constraints can be used to perform multi-objective nonlinear optimization on the cutting force mapping model and the cutting vibration mapping model to obtain the Pareto solution set of process parameters; finally, a preset TOPSIS algorithm can be used to select the optimal parameters from the Pareto solution set to obtain the globally optimal process parameters.
[0018] Based on this, in this application, since parameter optimization relies on multi-source machining data, a cutting force mapping model and a cutting vibration mapping model are fitted and constructed through a data-driven approach, and a multi-objective genetic algorithm is used to optimize the machining process in multiple objectives. Therefore, it can effectively reduce the dependence on simulation computing power, reduce the number of trial cuts and product scrap rate, improve the prediction accuracy and machining efficiency of machining, and solve the optimal machining parameters under the working condition, ultimately achieving low-stress, low-vibration, and high-efficiency machining of components. Attached Figure Description
[0019] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the drawings described below are only embodiments of this application. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.
[0020] Figure 1 This is a schematic diagram of an application scenario provided by an embodiment of this application; Figure 2 A schematic flowchart of the intelligent optimization method for process parameters provided in this application embodiment; Figure 3 A schematic diagram of BP neural network training provided in an embodiment of this application; Figure 4 A schematic diagram of the mean square error of a BP neural network provided in an embodiment of this application; Figure 5 A schematic diagram illustrating the fit degree and regression coefficients of a BP neural network provided in an embodiment of this application; Figure 6 A schematic diagram comparing model predictions and actual processed values provided in an embodiment of this application; Figure 7 A schematic diagram of a multi-objective genetic algorithm provided in an embodiment of this application; Figure 8 A schematic diagram of the Pareto solution set provided in an embodiment of this application; Figure 9 This is a schematic diagram of a process parameter intelligent optimization device provided in an embodiment of this application.
[0021] The diagram is labeled as follows: 10-Intelligent process parameter optimization device, 101-Processor, 102-Memory, 103-I / O interface, 104-Database, 90-Intelligent process parameter optimization device, 901-Model building unit, 902-Process parameter optimization unit, 903-Optimal process parameter acquisition unit, and 904-Data acquisition unit. Detailed Implementation
[0022] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application. Unless otherwise specified, the embodiments and features in the embodiments of this application can be arbitrarily combined with each other. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than that shown here.
[0023] Currently, aerospace equipment widely employs weakly rigid structural components, which are typically composed of numerous thin-walled webs and side plates. Because their wall thickness is much smaller than the overall outline dimensions, their structural rigidity is poor, resulting in high material removal rates during machining. They are significantly affected by cutting forces and vibrations, making them highly susceptible to machining deformation, which severely impacts the dimensional accuracy and mechanical properties of the product. Therefore, current research on process parameter optimization often faces the following problems: I. Machining is a complex nonlinear process involving multiple physical fields such as thermo-mechanical coupling, elastoplastic deformation, creep, fracture, and friction. Studying it using CAE simulation or experimental methods is not only technically challenging and costly, but also inefficient.
[0024] Second, in actual production, the process parameters are often optimized by relying on experience or by measuring while processing, which often requires multiple trial cuts, which can easily lead to an increase in product scrap rate and a decrease in production efficiency.
[0025] Based on this, embodiments of this application provide an intelligent optimization method for process parameters. In this method, firstly, a cutting force mapping model and a cutting vibration mapping model can be constructed based on multi-source machining data; wherein, the multi-source machining data is obtained through vibration sensors, triaxial force gauges, current sensors, and secondary development libraries; then, a multi-objective nonlinear optimization of the cutting force mapping model and the cutting vibration mapping model can be performed using a preset genetic algorithm and preset constraints to obtain the Pareto solution set of process parameters; finally, the optimal parameters can be selected from the Pareto solution set using a preset TOPSIS algorithm to obtain the globally optimal process parameters.
[0026] Based on this, in this application, since parameter optimization relies on multi-source machining data, a cutting force mapping model and a cutting vibration mapping model are fitted and constructed through a data-driven approach, and a multi-objective genetic algorithm is used to optimize the machining process in multiple objectives. Therefore, it can effectively reduce the dependence on simulation computing power, reduce the number of trial cuts and product scrap rate, improve the prediction accuracy and machining efficiency of machining, and solve the optimal machining parameters under the working condition, ultimately achieving low-stress, low-vibration, and high-efficiency machining of components.
[0027] After introducing the design concept of the embodiments of this application, the following is a brief introduction to the application scenarios to which the technical solutions of the embodiments of this application can be applied. It should be noted that the application scenarios described below are only for illustrating the embodiments of this application and are not intended to limit the scope. In specific implementation, the technical solutions provided by the embodiments of this application can be flexibly applied according to actual needs.
[0028] like Figure 1 The diagram shown illustrates an application scenario provided by an embodiment of this application. This application scenario may include a process parameter intelligent optimization device 10.
[0029] The intelligent process parameter optimization device 10 can be used to intelligently optimize process parameters. For example, it can be an in-vehicle computer, a personal computer (PC), a server, or a laptop. The intelligent process parameter optimization device 10 may include one or more processors 101, memory 102, I / O interfaces 103, and a database 104. Specifically, the processor 101 can be a central processing unit (CPU) or a digital processing unit, etc. The memory 102 can be volatile memory, such as random-access memory (RAM); the memory 102 can also be non-volatile memory, such as read-only memory, flash memory, hard disk drive (HDD), or solid-state drive (SSD); or the memory 102 can be any other medium capable of carrying or storing desired program code in the form of instructions or data structures that can be accessed by a computer, but is not limited thereto. The memory 102 can be a combination of the above-mentioned memories. The memory 102 can store some program instructions of the intelligent optimization method for process parameters provided in the embodiments of this application. When these program instructions are executed by the processor 101, they can be used to implement the steps of the intelligent optimization method for process parameters provided in the embodiments of this application, so as to solve the technical problems of high product scrap rate and low processing efficiency in the prior art. The database 104 can be used to store data such as cutting force mapping model, cutting vibration mapping model, Pareto solution set and global optimal process parameters involved in the solution provided in the embodiments of this application.
[0030] In this embodiment, the intelligent process parameter optimization device 10 can obtain parameter optimization instructions through the I / O interface 103. Then, the processor 101 of the intelligent process parameter optimization device 10 will solve the technical problems such as high product scrap rate and low processing efficiency in the prior art according to the program instructions of the intelligent process parameter optimization method provided in this embodiment in the memory 102. In addition, data such as cutting force mapping model, cutting vibration mapping model, Pareto solution set and global optimal process parameters can be stored in the database 104.
[0031] Of course, the methods provided in the embodiments of this application are not limited to... Figure 1 The application scenarios shown can also be used in other possible scenarios, and this application embodiment does not impose any limitations. Figure 1The functions that the various devices in the application scenarios shown can achieve will be described in subsequent method embodiments, and will not be elaborated on here. Below, the methods of the embodiments of this application will be described in conjunction with the accompanying drawings.
[0032] like Figure 2 The diagram shown is a flowchart illustrating a method for intelligent optimization of process parameters provided in an embodiment of this application. This method can... Figure 1 The process parameter intelligent optimization equipment 10 is used to perform the operation. Specifically, the process flow of this method is described below.
[0033] Step 201: Use the constructed cutting data acquisition platform to acquire data and obtain multi-source machining data.
[0034] Specifically, firstly, a cutting data acquisition platform can be built, which can integrate vibration sensors, triaxial force gauges, current sensors, and secondary development libraries into the host computer software for real-time data acquisition.
[0035] Then, a triaxial force gauge (e.g., a piezoelectric triaxial force gauge) can be used to acquire the cutting force. When the cutting force load is transmitted to the force gauge platform through the workpiece, the charge of the piezoelectric material changes, and the electrical signal is transmitted to the acquisition software through an amplification circuit.
[0036] Next, a vibration sensor (e.g., a laser displacement sensor) can be used to collect vibration signals. The laser point is positioned on the smooth part of the tool holder. During the machining process, the tool changes position, and the displacement sensor records the displacement change in a very short time.
[0037] Then, a current sensor can be used to collect the current. The spindle current can reflect the load of the CNC machining center to a certain extent. The current sensor can be a Hall current sensor, which uses the Hall effect to convert the current signal into a voltage signal. The voltage signal is collected by a data acquisition card and finally transmitted to the PC to output the current using a communication bus protocol.
[0038] Next, the status and real-time information of the CNC machining center can be read from the secondary development library. For example, using communication protocols such as TCP / IP, the machining real-time coordinates, error compensation, and actual spindle speed can be read in real time through the secondary development package interface provided by the machine tool.
[0039] Finally, host computer software can be written to integrate the secondary development library and signal acquisition quantities into the signal acquisition host computer software.
[0040] Step 202: Construct a cutting force mapping model based on multi-source machining data.
[0041] The multi-source processing data is obtained through vibration sensors, triaxial force gauges, current sensors, and secondary development libraries.
[0042] Specifically, in practical applications, changes in tool parameters, process parameters, and initial residual stress lead to changes in cutting force and residual stress during machining. The cutting force and its introduced residual stress are significant factors causing machining deformation. In actual machining, the workpiece and tool remain largely unchanged. Therefore, in this application, when investigating machining deformation, orthogonal experiments are conducted to fit the cutting force by changing the process parameters.
[0043] Based on this, firstly, multiple orthogonal experiments can be conducted using both large depth-of-cut and small width-of-cut side milling and large width-of-cut and small depth-of-cut end milling to obtain multiple process parameters and multiple cutting forces for side milling and end milling. That is, based on the two machining methods of "large depth-of-cut and small width-of-cut" side milling and "large width-of-cut and small depth-of-cut" end milling, different combinations of process parameters can be selected using orthogonal experimental design to cover the expected ideal range with as few experiments as possible. Cutting experiments are conducted according to the designed orthogonal parameters, and parameter information of the machining process is collected.
[0044] Then, the cutting data can be imported into statistical software, and the calculation formulas for the process parameters and cutting forces of side milling and end milling can be obtained by iterative calculation using the Levenberg-Marquardt method. The R-squared (coefficient of determination COD) of both can be evaluated. If the R-squared is close to the value of 1, it can be said that the collected data is reasonable and effective.
[0045] Next, all experimental data from side milling and end milling can be categorized, summarized, and fitted to construct a cutting force mapping model between cutting parameters and cutting forces under this working condition. This cutting force mapping model can be expressed by the following formula:
[0046] in, , , These are three-dimensional cutting forces. The resultant force is the cutting force; , f, n, and y1, y2, y3, y4, z1, z2, z3, z4, h1, h2, h3, h4 are the coefficients to be determined during the fitting process. These coefficients (all constants) can be obtained by substituting experimental data into the above cutting force mapping model. Based on this, the cutting force mapping model also needs to be adjusted for the degree of fit (…). The closer the result is to 1, the better the fit and the higher the prediction accuracy of the model.
[0047] Step 203: Construct a cutting vibration mapping model based on multi-source machining data.
[0048] Specifically, the multi-source machining data and the Levenberg-Marquardt method can be used to train the initial BP neural network model to obtain the cutting vibration mapping model.
[0049] In constructing the initial BP neural network model, the input and output layer parameters must first be selected. Combining the process parameters selected in the orthogonal experiment in step 202, cutting depth, cutting width, feed per tooth, and spindle speed can be chosen as input layer parameters, and cutting vibration as the output layer response. A certain amount of samples is selected for training. The number of hidden layers is set to 4. 70% of the sample size is used as the training set for training and fitting, 15% of the sample size is used as the validation set to solve for the optimal network depth, backpropagation termination point, and number of hidden layer neurons, and 15% of the sample size is used as the test set to evaluate the model's accuracy and performance.
[0050] Then, the Sigmoid function can be chosen as the activation function for the neural network, and the Levenberg-Marquardt method can be used as the training function for fast and accurate training. After the parameters are set, the training of the neural network can begin. Figure 3 The diagram shown is a schematic representation of a BP neural network training method provided in an embodiment of this application.
[0051] Next, model accuracy is evaluated using mean squared error (MSE), input-output correlation (R), and the error between the analog quantity and the actual vibration. The mean squared error evaluation metric refers to the result calculated after multiple iterations using training, validation, and test sets of data. Figure 4 The diagram shown illustrates the mean square error of a BP neural network provided in this embodiment. If the mean square error tends to be stable without significant fluctuations, and the target error is small, monotonically decreasing, and converges quickly, it demonstrates the excellent performance of the BP network. Figure 5 The diagram shown is a schematic of the fitting degree and regression coefficient of the BP neural network provided in the embodiment of this application. The R value evaluation index refers to the correlation between the actual data distribution and the fitting curve. The closer it is to 1, the better the correlation and the more accurate the prediction model.
[0052] Finally, the new process parameters can be substituted into the cutting vibration mapping model, and error analysis can be performed between the actual vibration quantity and the predicted simulated quantity, such as... Figure 6 The diagram shown is a comparison between the model prediction value and the actual processed value provided in the embodiment of this application. If the difference is less than 10%, it can be proven that the prediction model is accurate and effective.
[0053] Step 204: Using a preset genetic algorithm and preset constraints, perform multi-objective nonlinear optimization on the cutting force mapping model and the cutting vibration mapping model to obtain the Pareto solution set of the process parameters.
[0054] That is, under constraints of machine tool spindle speed, feed rate, power, tool, and cutting width, a preset genetic algorithm can be used to perform multi-objective nonlinear optimization on the cutting force mapping model and the cutting vibration mapping model to obtain the Pareto solution set of the process parameters.
[0055] Specifically, firstly, the objective function is determined. To achieve high-quality machining and reduce workpiece deformation, this application requires reducing cutting force and cutting vibration. Simultaneously, improving production efficiency must also be considered.
[0056] Assume the cutting force mapping model is simplified as follows:
[0057] The simplified representation of the cutting vibration mapping model is as follows:
[0058] The simplified formula for calculating processing efficiency is as follows:
[0059] Therefore, the three optimization objective functions are as follows: , ,
[0060] Then, the optimization constraints are determined. These constraints include machine tool spindle speed constraints, feed rate constraints, power constraints, tool constraints, and cutting width constraints.
[0061] Based on this, the constraints on the machine tool spindle speed and feed rate are as follows:
[0062]
[0063] in, , These are the upper and lower limits of the machine tool spindle speed, respectively; Number of teeth; , These are the upper and lower limits of the feed rate, respectively.
[0064] The power constraint is as follows:
[0065] in, This refers to the actual cutting force. This refers to the actual processing speed; This refers to the actual processing speed; For the transmission efficiency of machine tools; This refers to the motor power of the machine tool.
[0066] The tool constraint conditions are shown in the following formula:
[0067] in, , These are the upper and lower limits of the feed per tooth.
[0068] The cutting width constraint is as follows:
[0069] in, d is the cutting width, and d is the tool diameter.
[0070] Next, the parameters of the multi-objective genetic algorithm can be set according to the amount of data and model characteristics. For example, the maximum number of iterations can be set to 100, the population size to 150, the crowding degree to 0.8, the crossover number to 90, the mutation percentage to 0.8, the mutation probability to 0.05, and the step size of the constraint variable to 0.0001. Figure 7 The diagram shown is a schematic of a multi-objective genetic algorithm provided in an embodiment of this application.
[0071] Finally, after multiple rounds of iterative calculations, the three-dimensional Pareto solution set of the optimized process parameters can be obtained, such as... Figure 8 The diagram shown illustrates a Pareto solution set provided in an embodiment of this application. A Pareto solution describes a processing state of optimal resource allocation, referring to a situation in multi-objective optimization where no solution is inferior to the others and is superior to at least one other solution on one objective function. To determine the processing parameters, further parameter optimization of the solution set is required.
[0072] Step 205: Use the preset TOPSIS algorithm to select the optimal parameters for the Pareto solution set and obtain the globally optimal process parameters.
[0073] In practical applications, when evaluating process parameters, higher machining efficiency is generally considered better; therefore, machining efficiency can be referred to as a maximum-scale indicator. Conversely, lower cutting vibration and cutting force are considered better; therefore, cutting vibration and cutting force can be referred to as minimum-scale indicators. Based on this, to facilitate data calculation, improve computational stability, and accelerate computational convergence, this application employs a "TOPSIS decision algorithm based on entropy weighting" as the preset TOPSIS algorithm to obtain globally optimal process parameters.
[0074] Specifically, in this application, to facilitate data calculation and output, the process parameters need to be matrixed. If the solution set has a results and b evaluation indicators, the matrix is as follows:
[0075] Based on this, the initial matrix of the Pareto solution set (which is a 3D-Pareto solution set) is:
[0076] Then, for ease of calculation and processing, this application also needs to perform positive transformation on the evaluation indicators, that is, the extremely large indicators are not processed, and the extremely small indicators are processed by taking the reciprocal, as shown in the following formula:
[0077] Next, to improve computational stability, accelerate convergence, and eliminate the influence of dimensions on the calculation, this application also performs data normalization. The method used here is sum-of-squares normalization to process each index, as shown in the following formula:
[0078]
[0079] Then, we can also solve for the positive and negative ideal vectors. and The value is used as an indicator to evaluate the Pareto solution set, as shown in the following formula: ,
[0080] ,
[0081] ,
[0082] Next, we can determine the distance of each solution to the positive and negative ideal values of each evaluation index. The larger the value, the closer it is to the ideal value. The smaller the value, the farther away it is from the negative ideal value. The formula for calculating the ideal value is as follows: ,
[0083] Finally, to more intuitively evaluate each parameter, this application also introduces the relative similarity index of process parameter evaluation. The calculation method is shown in the formula. After calculating the relative similarity of each Pareto process parameter solution, they are sorted. The larger the value, the better the process parameter. The largest value is the globally optimal process parameter. The globally optimal process parameter is expressed by the following formula:
[0084] in, For the i-th globally optimal process parameter, The distance to the i-th positive ideal value. The distance to the i-th negative ideal value. Furthermore, based on the aforementioned "TOPSIS decision algorithm based on entropy weight method," low-stress, low-vibration, and high-efficiency processing parameters can be obtained, which are globally optimal process parameters.
[0085] In summary, this application utilizes multi-source sensors and a CNC secondary development library to collect and analyze cutting information from CNC machining centers. Orthogonal cutting experiments are conducted using two common cutting methods in actual machining scenarios: end milling and side milling, to construct a multi-source data-driven cutting force prediction model. A data-driven cutting vibration prediction model is established based on the collected vibration dataset and a BP neural network. A multi-objective genetic algorithm is used to optimize the machining process by reducing cutting force and vibration while increasing machining efficiency. The TOPSIS decision algorithm is applied to obtain the globally optimal process parameters. Therefore, this provides an effective approach to achieving low-stress, low-vibration, and high-efficiency machining of components. It not only reduces the dependence on simulation computing power and the number of trial cuts and product scrap rates, but also improves the prediction accuracy of cutting force and cutting vibration, and can accurately solve for the optimal machining parameters under specific working conditions, reducing machining deformation.
[0086] Based on the same inventive concept, embodiments of this application provide a process parameter intelligent optimization device 90, such as... Figure 9 As shown, the intelligent process parameter optimization device 90 includes: Model building unit 901 is used to build a cutting force mapping model and a cutting vibration mapping model based on multi-source machining data; wherein, the multi-source machining data is obtained through vibration sensors, triaxial force gauges, current sensors and secondary development libraries; The process parameter optimization unit 902 is used to perform multi-objective nonlinear optimization of the cutting force mapping model and the cutting vibration mapping model using a preset genetic algorithm and preset constraints, so as to obtain the Pareto solution set of the process parameters. The optimal process parameter acquisition unit 903 is used to select the optimal parameters from the Pareto solution set using the preset TOPSIS algorithm to obtain the globally optimal process parameters.
[0087] Optionally, model building unit 901 is also used for: Multiple orthogonal experiments were conducted using side milling with large depth of cut and small width of cut and end milling with large width of cut and small depth of cut to obtain multiple process parameters and multiple cutting forces for side milling and end milling. Multiple process parameters and multiple cutting forces for side milling and end milling are classified, summarized, and fitted to construct a cutting force mapping model.
[0088] Optionally, model building unit 901 is also used for: A cutting vibration mapping model was obtained by training an initial BP neural network model using multi-source machining data and the Levenberg-Marquardt method. The initial BP neural network model has 4 hidden layers and the activation function is the Sigmoid function.
[0089] Optionally, the process parameter optimization unit 902 is also used for: Under constraints of machine tool spindle speed, feed rate, power, tool, and cutting width, a pre-defined genetic algorithm is used to perform multi-objective nonlinear optimization on the cutting force mapping model and the cutting vibration mapping model to obtain the Pareto solution set of the process parameters.
[0090] Optionally, the intelligent process parameter optimization device 90 also includes a data acquisition unit 904, used for: A cutting data acquisition platform was built; the platform includes a vibration sensor, a triaxial force gauge, a current sensor, and a secondary development library. A three-dimensional force gauge was used to collect cutting force. Vibration signals are acquired using vibration sensors. Current is collected using a current sensor; Read the status and real-time information of the CNC machining center from the secondary development library.
[0091] The intelligent process parameter optimization device 90 can be used for execution. Figure 2 The method executed by the intelligent process parameter optimization device in the illustrated embodiment can be referenced for understanding the functions that each functional module of the intelligent process parameter optimization device 90 can achieve. Figure 2 The embodiments shown are described in detail below.
[0092] In some possible implementations, various aspects of the methods provided in this application can also be implemented as a program product comprising program code that, when run on a computer device, causes the computer device to perform the steps of the methods according to the various exemplary embodiments of this application described above. For example, the computer device may perform actions such as... Figure 2The method performed by the intelligent process parameter optimization device in the illustrated embodiment.
[0093] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks. Alternatively, if the integrated units of this application are implemented as software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of this application, or the parts that contribute to the prior art, can be embodied in the form of software products. These computer software products are stored in a storage medium and include several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROM, RAM, magnetic disks, or optical disks.
[0094] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.
[0095] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.
Claims
1. A method for intelligent optimization of process parameters, characterized in that, The method includes: Based on multi-source machining data, a cutting force mapping model and a cutting vibration mapping model are constructed; wherein, the multi-source machining data is obtained through vibration sensors, triaxial force gauges, current sensors, and secondary development libraries; Using a pre-defined genetic algorithm and pre-defined constraints, a multi-objective nonlinear optimization is performed on the cutting force mapping model and the cutting vibration mapping model to obtain the Pareto solution set of the process parameters; The Pareto solution set is optimized by using the preset TOPSIS algorithm to obtain the globally optimal process parameters.
2. The method as described in claim 1, characterized in that, The steps of constructing a cutting force mapping model and a cutting vibration mapping model based on multi-source machining data include: Multiple orthogonal experiments were conducted using side milling with large depth of cut and small width of cut and end milling with large width of cut and small depth of cut to obtain multiple process parameters and multiple cutting forces for side milling and end milling. Multiple process parameters and multiple cutting forces for side milling and end milling are classified, summarized, and fitted to construct a cutting force mapping model.
3. The method as described in claim 1, characterized in that, The steps of constructing a cutting force mapping model and a cutting vibration mapping model based on multi-source machining data include: The initial BP neural network model is trained using the multi-source machining data and the Levenberg-Marquardt method to obtain the cutting vibration mapping model; wherein, the initial BP neural network model has 4 hidden layers and the activation function is the Sigmoid function.
4. The method as described in claim 1, characterized in that, The step of using a preset genetic algorithm and preset constraints to perform multi-objective nonlinear optimization on the cutting force mapping model and the cutting vibration mapping model to obtain the Pareto solution set of the process parameters includes: Under constraints of machine tool spindle speed, feed rate, power, tool, and cutting width, a pre-defined genetic algorithm is used to perform multi-objective nonlinear optimization on the cutting force mapping model and the cutting vibration mapping model to obtain the Pareto solution set of the process parameters.
5. The method as described in claim 1, characterized in that, Before constructing the cutting force mapping model and the cutting vibration mapping model based on multi-source machining data, the method further includes: A cutting data acquisition platform is established; wherein, the acquisition platform includes a vibration sensor, a triaxial force gauge, a current sensor, and a secondary development library; A three-dimensional force gauge was used to collect cutting force. Vibration signals are acquired using vibration sensors. Current is collected using a current sensor; Read the status and real-time information of the CNC machining center from the secondary development library.
6. The method as described in claim 1, characterized in that, The cutting force mapping model is represented by the following formula: in, , , These are three-dimensional cutting forces. The resultant force is the cutting force; , f and n are the axial depth of cut, radial depth of cut, feed per tooth, and spindle speed, respectively; x1, x2, x3, x4, y1, y2, y3, y4, z1, z2, z3, z4, h1, h2, h3, and h4 are the coefficients to be determined during fitting.
7. The method as described in claim 1, characterized in that, The globally optimal process parameters are expressed by the following formula: in, For the i-th globally optimal process parameter, The distance to the i-th positive ideal value. It is the distance to the i-th negative ideal value.
8. A process parameter intelligent optimization device, characterized in that, The device includes: The model building unit is used to construct a cutting force mapping model and a cutting vibration mapping model based on multi-source machining data; wherein, the multi-source machining data is obtained through vibration sensors, triaxial force gauges, current sensors, and secondary development libraries; The process parameter optimization unit is used to perform multi-objective nonlinear optimization of the cutting force mapping model and the cutting vibration mapping model using a preset genetic algorithm and preset constraints, so as to obtain the Pareto solution set of the process parameters. The optimal process parameter acquisition unit is used to select the optimal parameters from the Pareto solution set using a preset TOPSIS algorithm to obtain the globally optimal process parameters.
9. An electronic device, characterized in that, The device includes: Memory, used to store program instructions; A processor is configured to invoke program instructions stored in the memory and execute the method described in any one of claims 1-7 according to the obtained program instructions.
10. A storage medium, characterized in that, The storage medium stores computer-executable instructions for causing a computer to perform the method described in any one of claims 1-7.