Surface roughness prediction method and system based on multi-order modal superposition for flank milling

Abstract

This disclosure relates to a method and system for predicting the surface roughness of side milling based on multi-mode superposition, including: acquiring multi-mode parameters of the cutting tool and machining parameters; determining an ideal tool tip trajectory based on the machining parameters; establishing a tool vibration response model under periodic cutting force excitation based on the principle of modal superposition and the multi-mode parameters, and calculating the vibration amplitude distribution at different heights along the tool axis; extracting the time-domain vibration response at the tool tip height as a vibration disturbance term, superimposing it onto the ideal tool tip trajectory to construct the tool tip motion trajectory; constructing a surface morphology model based on the tool tip motion trajectory to obtain the theoretical distribution law of surface roughness with the tool axial height; constructing a neural network prediction model based on the multi-mode parameters, machining parameters, and theoretical distribution law, training the model using a loss function based on the tool vibration response model, and obtaining the surface roughness prediction result of the tool. This disclosure can improve the accuracy and reliability of roughness prediction.

Description

Technical Field

[0001] This disclosure relates to the field of CNC machining and surface quality prediction technology, and in particular to a method and system for predicting the surface roughness of side milling based on multi-mode superposition. Background Technology

[0002] During side milling, the cutting tool generates forced vibration under the excitation of periodic cutting force. Especially with slender tools or tools with large overhangs, there are significant differences in the vibration response of the tool at different heights along the axial direction. This difference directly affects the actual movement trajectory of the tool tip, thereby changing the morphology and roughness distribution of the machined surface.

[0003] Existing surface roughness prediction methods are mostly based on geometric models or empirical formulas, failing to fully consider the influence of multi-mode vibrations of the cutting tool, resulting in insufficient prediction accuracy. Therefore, it is necessary to establish a surface roughness prediction method for side milling that considers the influence of multi-mode vibrations. Summary of the Invention

[0004] To address the shortcomings of existing methods, which neglect vibration effects or lack mechanistic support, resulting in insufficient accuracy and reliability in surface roughness prediction, this disclosure proposes a multi-mode superposition-based method for predicting the surface roughness of side-milled surfaces.

[0005] According to one aspect of this disclosure, a method for predicting the surface roughness of side milling based on multi-mode superposition is provided, comprising:

[0006] S10. Obtain the multi-mode parameters of the tool and the machining process parameters. The multi-mode parameters are obtained through experimental modal analysis or finite element simulation, including natural frequency, damping ratio and mode shape. The machining process parameters include at least one of spindle speed, feed rate and depth of cut.

[0007] S20. Determine the ideal tool tip trajectory according to the machining process parameters, and establish a tool vibration response model under periodic cutting force excitation based on the modal superposition principle according to the multi-mode parameters, and calculate the vibration amplitude distribution at different heights along the tool axis.

[0008] S30. Based on the vibration amplitude distribution, extract the time-domain vibration response at the height of the tool tip as a vibration disturbance term, and superimpose the vibration disturbance term onto the ideal tool tip trajectory to construct the tool tip motion trajectory;

[0009] S40. Construct a surface morphology model based on the tool tip motion trajectory to obtain the theoretical distribution law of surface roughness with the tool axial height;

[0010] S50. Based on the multi-mode parameters, the machining process parameters, and the theoretical distribution law, a neural network prediction model is constructed, and a loss function is constructed based on the tool vibration response model. The neural network prediction model is then trained to obtain the surface roughness prediction results at different axial heights of the tool.

[0011] Preferably, a tool vibration response model under periodic cutting force excitation is established based on the principle of modal superposition, including:

[0012] Discretize the tool axial direction into multiple nodes;

[0013] Based on the principle of modal superposition, the frequency domain response of each node under periodic cutting force excitation is calculated.

[0014] Preferably, the tool vibration response model is expressed as:

[0015] ,

[0016] In the formula, Let m be the steady-state response amplitude of the i-th mode, and m be the mode order. For mode shape functions, The excitation angular frequency.

[0017] Preferably, the trajectory of the blade tip is represented as follows:

[0018] ,

[0019] In the formula, For the ideal blade tip trajectory, This is the vibration disturbance term.

[0020] Preferably, the surface morphology model of the processed surface is represented as follows:

[0021] ,

[0022] In the formula, To account for the actual trajectory of the tool tip after vibration disturbance, T is the cycle of the tool teeth.

[0023] Preferably, the neural network prediction model is trained by constructing a loss function based on the tool vibration response model, including:

[0024] A loss function is constructed that includes a data error term and a physical constraint term, wherein the physical constraint term is constructed based on the tool vibration response model, and the data error term is constructed based on the deviation between the predicted value and the true value of the neural network model.

[0025] The neural network prediction model is trained with the loss function as the target.

[0026] Preferably, the loss function is expressed as:

[0027] ,

[0028] ,

[0029] ,

[0030] In the formula, For data error terms, For physical constraints, These are the weighting coefficients. The predicted value of the neural network model. The true value of the neural network model. The result is the calculation result of the tool vibration response model, where N is the number of samples.

[0031] According to one aspect of this disclosure, a surface roughness prediction system for side milling based on multi-mode superposition is provided, comprising:

[0032] The parameter acquisition module acquires the multi-mode parameters of the tool and the machining process parameters. The multi-mode parameters are obtained through experimental modal analysis or finite element simulation and include natural frequency, damping ratio and mode shape. The machining process parameters include at least one of spindle speed, feed rate and depth of cut.

[0033] The vibration response analysis module determines the ideal tool tip trajectory based on the machining process parameters, establishes a tool vibration response model under periodic cutting force excitation based on the modal superposition principle according to the multi-mode parameters, and calculates the vibration amplitude distribution at different heights along the tool axis.

[0034] The trajectory construction module extracts the time-domain vibration response at the height of the tool tip as a vibration disturbance term based on the vibration amplitude distribution, and superimposes the vibration disturbance term onto the ideal tool tip trajectory to construct the tool tip motion trajectory.

[0035] The morphology analysis module constructs a surface morphology model based on the tool tip movement trajectory to obtain the theoretical distribution law of surface roughness with the axial height of the tool.

[0036] The roughness prediction module constructs a neural network prediction model based on the multi-mode parameters, the machining process parameters, and the theoretical distribution law, and constructs a loss function based on the tool vibration response model to train the neural network prediction model, thereby obtaining the surface roughness prediction results of the tool at different axial heights.

[0037] According to one aspect of this disclosure, an electronic device is provided, comprising: a processor; a memory for storing processor-executable instructions; wherein the processor is configured to: execute the above-described method for predicting the surface roughness of side milling based on multi-mode superposition.

[0038] According to one aspect of this disclosure, a computer-readable storage medium is provided that stores computer program instructions thereon, which, when executed by a processor, implement the above-described method for predicting the surface roughness of side milling based on multi-mode superposition to generate a bit stream.

[0039] Compared to the prior art, the beneficial effects of this disclosure are as follows:

[0040] 1) This disclosure obtains multi-mode parameters of the tool through experimental modal analysis or finite element simulation, and establishes a tool vibration response model under periodic cutting force excitation based on the modal superposition principle. It can accurately describe the vibration amplitude distribution characteristics of the tool at different heights along the axial direction, and provide accurate vibration input conditions for surface roughness prediction.

[0041] 2) This disclosure superimposes the tool vibration response as a disturbance term onto the ideal tool tip trajectory to construct a machining surface morphology model that considers the influence of vibration. This realizes the coupled analysis of multi-mode vibration of the tool and the surface forming mechanism, and can quantitatively describe the influence law of tool vibration on the machining surface morphology.

[0042] 3) This disclosure constructs a neural network prediction model that integrates physical mechanism constraints, and embeds the tool vibration response model as a physical constraint term into the loss function, so that the neural network follows both experimental data and physical laws during the training process, thereby improving the generalization ability and prediction accuracy of the model and realizing accurate prediction of surface roughness at different heights along the tool axis.

[0043] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure.

[0044] Other features and aspects of this disclosure will become clear from the following detailed description of exemplary embodiments with reference to the accompanying drawings. Attached Figure Description

[0045] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this disclosure and, together with the specification, serve to illustrate the technical solutions of this disclosure.

[0046] Figure 1 A flowchart of a side milling surface roughness prediction method based on multi-mode superposition is shown in an embodiment of this disclosure;

[0047] Figure 2A flowchart of side milling surface roughness prediction and process optimization considering vibration disturbance is shown in an embodiment of this disclosure;

[0048] Figure 3 A schematic diagram of multiple vibration modes of the tool in an embodiment of this disclosure is shown;

[0049] Figure 4 A schematic diagram of the blade tip trajectory model in an embodiment of this disclosure is shown;

[0050] Figure 5 A neural network structure diagram is shown in an embodiment of this disclosure;

[0051] Figure 6 A block diagram of a side milling surface roughness prediction system based on multi-mode superposition in an embodiment of this disclosure is shown. Detailed Implementation

[0052] Various exemplary embodiments, features, and aspects of this disclosure will now be described in detail with reference to the accompanying drawings. The same reference numerals in the drawings denote elements that have the same or similar functions. Although various aspects of the embodiments are shown in the drawings, they are not necessarily drawn to scale unless specifically indicated otherwise.

[0053] The term “exemplary” as used herein means “serving as an example, embodiment, or illustration.” Any embodiment illustrated herein as “exemplary” is not necessarily to be construed as superior to or better than other embodiments.

[0054] In this document, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent three cases: A alone, A and B simultaneously, and B alone. Furthermore, the term "at least one" in this document means any combination of at least two of any one or more elements. For example, including at least one of A, B, and C can mean including any one or more elements selected from the set consisting of A, B, and C.

[0055] Furthermore, to better illustrate this disclosure, numerous specific details are set forth in the following detailed description. Those skilled in the art will understand that this disclosure can be practiced without certain specific details. In some instances, methods, means, components, and circuits well known to those skilled in the art have not been described in detail in order to highlight the main points of this disclosure.

[0056] To make the objectives, technical solutions, and advantages of the embodiments of this disclosure clearer, the technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of this disclosure, but not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this disclosure without creative effort are within the scope of protection of this disclosure.

[0057] Based on the above ideas, this disclosure proposes a method for predicting the surface roughness of side milling based on the superposition of multiple modes. Figure 1 A flowchart is shown for a method for predicting the surface roughness of side milling based on multi-mode superposition. The method includes:

[0058] S10. Obtain the multi-mode parameters of the tool and the machining process parameters. The multi-mode parameters are obtained through experimental modal analysis or finite element simulation, including natural frequency, damping ratio and mode shape. The machining process parameters include at least one of spindle speed, feed rate and depth of cut.

[0059] S20. Determine the ideal tool tip trajectory according to the machining process parameters, and establish a tool vibration response model under periodic cutting force excitation based on the modal superposition principle according to the multi-mode parameters, and calculate the vibration amplitude distribution at different heights along the tool axis.

[0060] S30. Based on the vibration amplitude distribution, extract the time-domain vibration response at the height of the tool tip as a vibration disturbance term, and superimpose the vibration disturbance term onto the ideal tool tip trajectory to construct the tool tip motion trajectory;

[0061] S40. Construct a surface morphology model based on the tool tip motion trajectory to obtain the theoretical distribution law of surface roughness with the tool axial height;

[0062] S50. Based on the multi-mode parameters, the machining process parameters, and the theoretical distribution law, a neural network prediction model is constructed, and a loss function is constructed based on the tool vibration response model. The neural network prediction model is then trained to obtain the surface roughness prediction results at different axial heights of the tool.

[0063] In this embodiment, the process for predicting and optimizing the surface roughness of side milling considering vibration disturbance is as follows: Figure 2As shown, the specific steps are as follows: First, the frequency response function (FRF) is obtained through tool impact testing and vibration signal measurement. Then, multi-factor data acquisition for side milling is carried out, simultaneously acquiring modal parameters, cutting force data, and machining process parameters. Based on the acquired data, modal analysis and vibration modeling are performed to establish the vibration response relationship. On this basis, considering vibration disturbance, the actual tool tip trajectory is derived. Next, a surface morphology model is constructed to obtain the surface roughness calculation formula, and a neural network model is further constructed. Through training, a roughness prediction function is obtained. Then, the roughness prediction result is output and compared with the actual measurement result. If the prediction result is not similar to the actual result, the neural network model is optimized. If the result is similar, the machining process is optimized based on the prediction model, and the machining parameters are adjusted to achieve low roughness machining.

[0064] This disclosure further extends the above-described method with detailed possible implementations, specifically including:

[0065] S10. Obtain the multi-mode parameters of the tool and the machining process parameters. The multi-mode parameters are obtained through experimental modal analysis or finite element simulation, including natural frequency, damping ratio and mode shape. The machining process parameters include at least one of spindle speed, feed rate and depth of cut.

[0066] In one embodiment, a four-tooth carbide end mill with a manufacturing diameter of 12 mm is used as the measurement target for surface roughness prediction, and the measurement is performed according to the method of this disclosure. The modal parameters of the tool, including its natural frequencies, are obtained using an impact hammer modal testing method. Damping ratio and mode shape function The mode shapes of the tool in the first few orders are obtained through modal testing or finite element analysis. The multi-order mode shape diagrams are shown below. Figure 3 As shown.

[0067] In the experiment, a force hammer was used to excite the free end of the tool, and an accelerometer was used to collect the tool vibration response signal. The excitation signal and response signal were recorded using a signal acquisition system, thereby obtaining the tool's frequency response function (FRF) as follows:

[0068] ,

[0069] In the formula, The vibration response spectrum, The excitation force spectrum is shown in the table below. The first four modal parameters of the tool can be obtained by peak identification and curve fitting of the frequency response function.

[0070] Modal order Natural frequency (Hz) Damping ratio 1 820 0.021 2 2140 0.018 3 3950 0.016 4 6120 0.015

[0071] The modal parameter set is represented as:

[0072] ,

[0073] In the formula, For mode shape functions, Let be the natural frequency of the i-th mode. Let be the damping ratio of the i-th mode.

[0074] S20. Determine the ideal tool tip trajectory according to the machining process parameters, establish a tool vibration response model under periodic cutting force excitation based on the multi-mode parameters and the principle of modal superposition, and calculate the vibration amplitude distribution at different heights along the tool axis.

[0075] In this embodiment, the tool vibrates under the action of periodic cutting force during milling, and its dynamic equation is as follows:

[0076] ,

[0077] In the formula, For the quality matrix, Here is the damping matrix. Here is the stiffness matrix. For tool displacement response, For tool speed response, For tool acceleration response, The cutting force is the excitation. The result is obtained through modal transformation:

[0078] ,

[0079] In the formula, Let be the generalized coordinates of the i-th modal.

[0080] A tool vibration response model under periodic cutting force excitation is established based on the modal superposition principle, including: discretizing the tool axial direction into multiple nodes; and calculating the frequency domain response of each node under periodic cutting force excitation based on the modal superposition principle.

[0081] The i-th modal response equation is obtained as follows:

[0082] ,

[0083] The modal forces are:

[0084] ,

[0085] If the cutting force is a periodic excitation, it is expressed as:

[0086] ,

[0087] The steady-state response amplitude of the i-th mode is:

[0088] ,

[0089] The vibration response model of the cutting tool at any axial height z is expressed as follows:

[0090] ,

[0091] In the formula, Let m be the steady-state response amplitude of the i-th mode, and m be the mode order. For mode shape functions, The excitation angular frequency.

[0092] Discretizing the tool axis into n nodes yields the vibration amplitude distribution matrix as follows:

[0093] ,

[0094] In this embodiment, The vibration response model is then expressed as:

[0095] ,

[0096] To describe the vibration characteristics of the tool at different positions along the axial direction, the tool axial direction is discretized into n=20 nodes, and the vibration amplitude distribution matrix can be obtained as follows:

[0097] .

[0098] S30. Based on the vibration amplitude distribution, extract the time-domain vibration response at the height of the tool tip as a vibration disturbance term, and superimpose the vibration disturbance term onto the ideal tool tip trajectory to construct the tool tip motion trajectory.

[0099] In this embodiment, a tool tip trajectory model considering vibration disturbance is constructed, and the ideal tool tip trajectory is:

[0100] ,

[0101] In the formula, This is the feed per tooth.

[0102] The processing parameters used in this embodiment are shown in the table below.

[0103] parameter numerical values Spindle speed 6000r / min Feed per tooth 0.05mm / z Axial depth 5mm Radial depth 2mm

[0104] The trajectory of the tool tip after considering vibration disturbance is expressed as follows:

[0105] ,

[0106] In the formula, For the ideal blade tip trajectory, For the vibration disturbance term, the tool tip trajectory model in this embodiment is as follows: Figure 4 As shown.

[0107] The vibration disturbance term is:

[0108] .

[0109] S40. Construct a surface morphology model based on the tool tip motion trajectory to obtain the theoretical distribution law of surface roughness with the axial height of the tool.

[0110] In this embodiment, a surface topography model is constructed based on the tool tip trajectory, represented as follows:

[0111] ,

[0112] In the formula, To account for the actual trajectory of the tool tip after vibration disturbance, T is the cycle of the tool teeth.

[0113] The formula for calculating surface roughness with a sampling length of L is:

[0114] ,

[0115] In this embodiment, the sampling length is The theoretical roughness value is calculated using this method:

[0116] .

[0117] S50. Based on the multi-mode parameters, the machining process parameters, and the theoretical distribution law, a neural network prediction model is constructed, and a loss function is constructed based on the tool vibration response model. The neural network prediction model is then trained to obtain the surface roughness prediction results at different axial heights of the tool.

[0118] In this embodiment, a neural network model incorporating physical mechanism constraints is constructed, and an input feature vector is constructed, represented as follows:

[0119] ,

[0120] In the formula, This is the axial cutting depth. Radial cutting depth, This refers to the theoretical roughness.

[0121] Milling experiments were conducted under different machining parameters, resulting in 120 sets of experimental data samples. These samples were then used to construct an input feature vector, represented as follows:

[0122] ,

[0123] The neural network adopts a three-layer structure: a 9-node input layer, a 12-node hidden layer, and a 1-node output layer. The neural network structure is as follows: Figure 5 As shown, the network forward propagation formula is:

[0124] ,

[0125] In the formula, Let l be the weight matrix of the l-th layer. This is the output of the neuron in the previous layer. This is the output of the l-th layer neuron.

[0126] The activation function is:

[0127] ,

[0128] The predicted roughness of the output layer is:

[0129] ,

[0130] In the formula, This is the output layer weight matrix. This is the output result of the previous layer. This is the output layer bias.

[0131] Further, based on the tool vibration response model, a loss function is constructed to train the neural network prediction model, including: constructing a loss function containing a data error term and a physical constraint term, wherein the physical constraint term is constructed based on the tool vibration response model, and the data error term is constructed based on the deviation between the predicted value and the true value of the neural network model; and training the neural network prediction model with the loss function as the target.

[0132] Neural network training uses a loss function, which is expressed as:

[0133] ,

[0134] ,

[0135] ,

[0136] In the formula, For data error terms, For physical constraints, The weighting coefficient is taken as follows: , The predicted value of the neural network model. The true value of the neural network model. The result is the calculation result of the tool vibration response model, where N is the number of samples.

[0137] The roughness prediction function obtained through training is:

[0138] ,

[0139] In the formula, For the input feature vector, These are the parameters for the neural network model.

[0140] By inputting feature data under new machining parameter conditions, the surface roughness at different heights along the tool axis can be predicted.

[0141] This disclosure obtains multi-mode parameters of the cutting tool, including natural frequency, damping ratio, and mode shape, through experimental modal analysis or finite element simulation. Based on the principle of modal superposition, a tool vibration response model under periodic cutting force excitation is established to obtain the vibration amplitude distribution at different heights along the tool axis. On this basis, the vibration disturbance term is superimposed onto the ideal tool tip trajectory to construct a side milling surface morphology model considering the influence of vibration. Modal features and machining parameters are extracted, and a neural network prediction model integrating physical mechanism constraints is constructed. The model is trained by introducing a loss function with physical constraints to achieve accurate prediction of the surface roughness of the side milling. Furthermore, by combining the tool axis discretization and modal superposition methods, the distribution characteristics of multi-mode vibration and the tool trajectory forming mechanism are deeply integrated. This significantly improves the roughness prediction accuracy and the ability to depict the spatial distribution along the tool axis while ensuring the interpretability of the prediction model. This effectively overcomes the shortcomings of existing methods that ignore the influence of multi-mode vibration or lack mechanistic support, and enhances the method's process optimization and surface quality control capabilities in scenarios such as high overhang side milling and machining of slender tools.

[0142] As another aspect of the embodiments of this disclosure, a side milling surface roughness prediction system 100 based on multi-mode superposition is also provided, such as... Figure 6 As shown, it includes:

[0143] The parameter acquisition module 1 acquires the multi-mode parameters of the tool and the machining process parameters. The multi-mode parameters are obtained through experimental modal analysis or finite element simulation and include natural frequency, damping ratio and mode shape. The machining process parameters include at least one of spindle speed, feed rate and depth of cut.

[0144] Vibration response analysis module 2 determines the ideal tool tip trajectory based on the machining process parameters, establishes a tool vibration response model under periodic cutting force excitation based on the multi-mode parameters and the principle of modal superposition, and calculates the vibration amplitude distribution at different heights along the tool axis.

[0145] The trajectory construction module 3 extracts the time-domain vibration response at the height of the tool tip as a vibration disturbance term based on the vibration amplitude distribution, and superimposes the vibration disturbance term into the ideal tool tip trajectory to construct the tool tip motion trajectory.

[0146] The morphology analysis module 4 constructs a surface morphology model based on the tool tip movement trajectory to obtain the theoretical distribution law of surface roughness with the axial height of the tool.

[0147] The roughness prediction module 5 constructs a neural network prediction model based on the multi-mode parameters, the machining process parameters, and the theoretical distribution law, and constructs a loss function based on the tool vibration response model to train the neural network prediction model, thereby obtaining the surface roughness prediction results of the tool at different axial heights.

[0148] Without causing contradictions, the above-described modules in the system of the present disclosure embodiments can implement any of the above-described methods.

[0149] Based on the description of the above embodiments, it can be seen that the embodiments of this disclosure can achieve the following technical effects:

[0150] 1) The embodiments of this disclosure obtain multi-mode parameters of the tool through experimental modal analysis or finite element simulation, and establish a tool vibration response model under periodic cutting force excitation based on the modal superposition principle. This model can accurately describe the vibration amplitude distribution characteristics of the tool at different heights along the axial direction, and provide accurate vibration input conditions for surface roughness prediction.

[0151] 2) In this embodiment, the tool vibration response is superimposed as a disturbance term onto the ideal tool tip trajectory to construct a machining surface morphology model that considers the influence of vibration. This realizes the coupled analysis of multi-mode vibration of the tool and the surface forming mechanism, and can quantitatively describe the influence law of tool vibration on the machining surface morphology.

[0152] 3) The embodiments of this disclosure construct a neural network prediction model that integrates physical mechanism constraints. The tool vibration response model is embedded as a physical constraint term in the loss function, so that the neural network follows both experimental data and physical laws during the training process, thereby improving the generalization ability and prediction accuracy of the model and realizing accurate prediction of surface roughness at different heights along the tool axis.

[0153] This disclosure also proposes an electronic device, including: a processor; and a memory for storing processor-executable instructions; wherein the processor is configured to perform the aforementioned method for predicting the surface roughness of side milling based on multi-mode superposition. The electronic device may be provided as a terminal, a server, or other type of device.

[0154] This disclosure also proposes a computer-readable storage medium storing computer program instructions, which, when executed by a processor, implement the aforementioned method for predicting the surface roughness of side milling based on multi-mode superposition to generate a bitstream. The computer-readable storage medium may be a non-volatile computer-readable storage medium.

[0155] Those skilled in the art will understand that, in the above-described method and system for predicting the surface roughness of side milling based on multi-mode superposition in specific embodiments, the order in which each step is written does not imply a strict execution order and does not constitute any limitation on the implementation process. The specific execution order of each step should be determined by its function and possible internal logic.

[0156] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of an instruction containing one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than those shown in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0157] The various embodiments of this disclosure have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, or technical improvements to the technology in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A flank milling surface roughness prediction method based on multi-order modal superposition, characterized in that, include: S10. Obtain the multi-mode parameters of the tool and the machining process parameters. The multi-mode parameters are obtained through experimental modal analysis or finite element simulation, including natural frequency, damping ratio and mode shape. The machining process parameters include at least one of spindle speed, feed rate and depth of cut. S20. Determine the ideal tool tip trajectory according to the machining process parameters, and establish a tool vibration response model under periodic cutting force excitation based on the modal superposition principle according to the multi-mode parameters, and calculate the vibration amplitude distribution at different heights along the tool axis. S30. Based on the vibration amplitude distribution, extract the time-domain vibration response at the height of the tool tip as a vibration disturbance term, and superimpose the vibration disturbance term onto the ideal tool tip trajectory to construct the tool tip motion trajectory; S40. Construct a surface morphology model based on the tool tip motion trajectory to obtain the theoretical distribution law of surface roughness with the tool axial height; S50. Based on the multi-mode parameters, the machining process parameters, and the theoretical distribution law, a neural network prediction model is constructed, and a loss function is constructed based on the tool vibration response model. The neural network prediction model is then trained to obtain the surface roughness prediction results at different axial heights of the tool.

2. The method of claim 1, wherein, A tool vibration response model under periodic cutting force excitation is established based on the principle of modal superposition, including: Discretize the tool axial direction into multiple nodes; Based on the principle of modal superposition, the frequency domain response of each node under periodic cutting force excitation is calculated.

3. The method of claim 2, wherein, The tool vibration response model is expressed as follows: ， wherein is the amplitude of the steady state response of the i-th mode, m is the mode order, is the mode shape function, is the excitation angular frequency.

4. The method of claim 1, wherein, The trajectory of the blade tip is represented as follows: ， wherein is the ideal tip trajectory, is the vibrational disturbance term.

5. The method of claim 1, wherein, The surface morphology model of the processed surface is represented as follows: ， In the formula, T is the tool tooth passing period for considering the actual trajectory of the tool tip after the vibration disturbance.

6. The method of claim 1, wherein, Based on the tool vibration response model, a loss function is constructed, and the neural network prediction model is trained, including: A loss function is constructed that includes a data error term and a physical constraint term, wherein the physical constraint term is constructed based on the tool vibration response model, and the data error term is constructed based on the deviation between the predicted value and the true value of the neural network model. The neural network prediction model is trained with the loss function as the target.

7. The method of claim 6, wherein, The loss function is expressed as: ， ， ， In the formula, is a data error term, is a physical constraint term, is a weight coefficient, is a predicted value of the neural network model, is a true value of the neural network model, is a calculation result of the tool vibration response model, and N is a sample number.

8. A side milling surface roughness system based on multi-mode superposition, characterized in that, include: The parameter acquisition module acquires the multi-mode parameters of the tool and the machining process parameters. The multi-mode parameters are obtained through experimental modal analysis or finite element simulation and include natural frequency, damping ratio and mode shape. The machining process parameters include at least one of spindle speed, feed rate and depth of cut. The vibration response analysis module determines the ideal tool tip trajectory based on the machining process parameters, establishes a tool vibration response model under periodic cutting force excitation based on the modal superposition principle according to the multi-mode parameters, and calculates the vibration amplitude distribution at different heights along the tool axis. The trajectory construction module extracts the time-domain vibration response at the height of the tool tip as a vibration disturbance term based on the vibration amplitude distribution, and superimposes the vibration disturbance term onto the ideal tool tip trajectory to construct the tool tip motion trajectory. The morphology analysis module constructs a surface morphology model based on the tool tip movement trajectory to obtain the theoretical distribution law of surface roughness with the axial height of the tool. The roughness prediction module constructs a neural network prediction model based on the multi-mode parameters, the machining process parameters, and the theoretical distribution law, and constructs a loss function based on the tool vibration response model to train the neural network prediction model, thereby obtaining the surface roughness prediction results of the tool at different axial heights.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the side milling surface roughness method based on multi-mode superposition as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it generates a bitstream by implementing the side milling surface roughness method based on multi-mode superposition as described in any one of claims 1 to 7.

Images