A spindle stiffness testing method based on dynamic load

Abstract

The invention discloses a spindle stiffness testing method based on dynamic load. The method adopts a spindle stiffness testing system as a whole. The spindle stiffness testing system comprises a spindle structural parameter module, a comprehensive stiffness mapping module and a dynamic and static stiffness optimization module. The spindle structural parameter module is used to obtain structural parameters affecting the dynamic and static stiffness of the spindle. The comprehensive stiffness mapping module is used to establish a spindle performance parameter mapping model by using a neural network and a genetic algorithm. The dynamic and static stiffness optimization module is used to optimize the spindle. The spindle structural parameter module is electrically connected to the comprehensive stiffness mapping module, and the comprehensive stiffness mapping module is electrically connected to the dynamic and static stiffness optimization module. A simplified three-dimensional model of the spindle is established, corresponding loads and constraints are applied to obtain various structural parameters, neural model training is performed and output, and finally the spindle is optimized according to the model. The invention has the characteristics of improving machining accuracy and machining efficiency.

Description

Technical Field

[0001] The invention relates to the technical field of spindle stiffness testing, and in particular to a spindle stiffness testing method based on dynamic load. Background Art

[0002] The machine tool spindle refers to the axis on the machine tool that drives the workpiece or tool to rotate. As the core component of the machine tool, the performance of the spindle affects the machining accuracy and efficiency of the machine tool. The spindle stiffness, as an important performance indicator of the machine tool spindle, has an important influence on the size and shape error of the machined workpiece, the working performance and life of the spindle, and the vibration resistance of the machine tool. The dynamic and static stiffness of the machine tool directly affects the machining accuracy and efficiency of the machine tool. It is an important indicator to measure the performance of the machine tool. On the spindle of a high-speed finishing machine tool, it is a reflection of the machine tool's ability to resist vibration and external interference.

[0003] The existing detection of machine tool spindles can only apply vertical load to a certain position of the spindle. The load detection method using the same direction at a single position is prone to errors. At the same time, the detection of spindles of different sizes sometimes requires the replacement of different clamping mechanisms. Therefore, the accuracy and practicality of the existing testing devices are not ideal. At the same time, for the structuring of machine tools and spindles, when people optimize the stiffness of the spindle, the selection of design variables is often determined based on previous optimization experience. There is no more efficient optimization of sensitive deformation structural parameters, and the control of the dynamic and static stiffness of the spindle and the subsequent processing cost are not well unified. Therefore, it is necessary to design a spindle stiffness test method based on dynamic load to improve processing accuracy and processing efficiency. Summary of the invention

[0004] The purpose of the present invention is to provide a spindle stiffness testing method based on dynamic load to solve the problems raised in the above background technology.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions: a spindle stiffness testing method based on dynamic load, the method as a whole adopts a spindle stiffness testing system, the spindle stiffness testing system includes a spindle structural parameter module, a comprehensive stiffness mapping module and a dynamic and static stiffness optimization module, the spindle structural parameter module is used to obtain the structural parameters affecting the dynamic and static stiffness of the spindle according to the spindle deformation displacement and natural frequency, the comprehensive stiffness mapping module is used to use a neural network and a genetic algorithm to establish a spindle performance parameter mapping model with the obtained structural parameters, the dynamic and static stiffness optimization module is used to optimize the spindle so that the static and dynamic stiffness of the machine tool spindle is at an optimal level, the spindle structural parameter module is electrically connected to the comprehensive stiffness mapping module, and the comprehensive stiffness mapping module is electrically connected to the dynamic and static stiffness optimization module.

[0006] According to the above technical scheme, the spindle structural parameter module includes a three-dimensional modeling simplification module, a spindle static deformation module and a spindle dynamic stiffness module. The three-dimensional modeling simplification module is used to simplify the spindle structure and three-dimensionally model it. The spindle static deformation module is used to obtain the deformation displacement caused by the relevant structural parameters in the radial direction of the spindle under the action of load. The spindle dynamic stiffness module is used to obtain the relevant structural parameters that affect the spindle by utilizing the positive phase relationship between the natural frequency and dynamic stiffness of the spindle under the action of load. The three-dimensional modeling simplification module is electrically connected to the spindle static deformation module, and the spindle static deformation module is electrically connected to the spindle dynamic stiffness module.

[0007] According to the above technical solution, the comprehensive stiffness mapping module includes a structural parameter module, a parameter variable module, a BP neural network module and a genetic algorithm combination module, the structural parameter module is used to record and sort the obtained structural parameters affecting the dynamic and static stiffness of the main shaft, the parameter variable module is used to select appropriate structural parameters and performance parameters as input and output variables of the model, the BP neural network module is used to select the BP neural network to train the structural parameters for multiple times and output a parameter model consistent with the expected results, the genetic algorithm combination module is used to apply the genetic algorithm to make up for the defects of the neural network that the convergence speed is slow and it is easy to fall into the local optimal solution and cause the training failure, the structural parameter module is electrically connected to the parameter variable module, and the BP neural network module is electrically connected to the genetic algorithm combination module;

[0008] The parameter variable module includes an input variable submodule and an output variable submodule. The input variable submodule is used to select the parameter with the greatest influence among many influential structural parameters as the input variable of the neural network, and the output variable submodule is used to select the performance parameters of the main shaft itself as the output variable of the neural network for model training. The input variable submodule is electrically connected to the output variable submodule.

[0009] According to the above technical scheme, the dynamic and static stiffness optimization module includes a structural optimization module, a parameter level combination module and a spindle optimization module. The structural optimization module is used to select the structural parameters of the machine tool spindle as design variables and establish a spindle structure optimization mathematical model. The parameter level combination module is used to apply the improved spindle performance mapping model to the constraint conditions to obtain the optimal structural parameter level combination. The spindle optimization module is used to optimize the spindle according to the structural parameters affecting the dynamic and static stiffness of the spindle and the spindle performance neural network model. The structural optimization module is electrically connected to the parameter level combination module, and the parameter level combination module is electrically connected to the spindle optimization module.

[0010] According to the above technical solution, the spindle stiffness testing method based on dynamic load comprises the following steps:

[0011] Step S1: Establish a simplified three-dimensional model of the spindle on the stiffness test bench, and perform a force analysis on the machine tool spindle according to the load and constraint conditions under actual working conditions;

[0012] Step S2: the spindle is controlled by an industrial computer, and the displacement change data in the axial deformation direction of the spindle is collected to screen out the structural parameters that have a significant impact on the change of the radial stiffness of the spindle, and the positive correlation between the natural frequency of the spindle and the dynamic stiffness is used to screen out the structural parameters that have a significant impact on the change of the dynamic stiffness of the spindle;

[0013] Step S3: using the structure parameters obtained by screening as input variables and the spindle performance parameters as output variables, a BP neural network model of the machine tool spindle is established, and the model is optimized by combining a genetic algorithm;

[0014] Step S4: Establish a structural optimization mathematical model, optimize the structural parameters of the machine tool spindle, and obtain the optimal structural parameter level combination.

[0015] According to the above technical solution, step S1 further includes the following steps:

[0016] Step S11: when establishing the three-dimensional model, the tiny features on the spindle components are deleted to simplify the structure of the machine tool spindle;

[0017] Step S12: Selecting a smaller overhang under the premise of meeting the spindle structure requirements, the spindle stiffness is affected by the bearing overhang;

[0018] Step S13: applying constraints and loads to the spindle by using an external cylindrical cutting processing method, simulating the stress state of the spindle under corresponding working conditions, and performing stress analysis on the spindle;

[0019] Step S14: During the process of the spindle being deformed by force, the right horizontal X-axis, the vertical Z-axis and the front horizontal Y-axis are established at the spindle core. The spindle itself and the front and rear support bearings are deformed by force, causing the spindle shaft end to be displaced.

[0020] According to the above technical solution, step S2 further includes the following steps:

[0021] Step S21: Using the parameter drive of the three-dimensional model, collect and record any structural parameter set of the main axis {D1, D2, ...D n The corresponding principal axis radial deformation force set {L1, L2, ...L n};

[0022] Step S22: Calculate the influence of the structural parameters on the maximum deformation in the X direction of the spindle, and use the spindle radial stiffness to represent the influence. The spindle radial stiffness calculation formula is:

[0023]

[0024] In the formula, F i is the radial stiffness of the spindle, Δr is the load on the spindle set under different working conditions, and L max The maximum deformation amount of the main axis radial deformation force concentration;

[0025] Step S23: By performing constrained modal analysis on the original model of the main shaft, the axial deformation force set {l1, l2, ...l n};

[0026] Step S24: Calculate the stiffness of the main shaft under the action of the sinusoidal alternating load, that is, calculate the dynamic stiffness F as a function of the excitation frequency. j , the calculation formula is:

[0027]

[0028] Where A is the amplitude of the main shaft vibration under load, l i is the axial deformation force of the main shaft corresponding to the excitation frequency function, and ω is the corresponding natural frequency in the first order.

[0029] According to the above technical solution, step S3 further includes the following steps:

[0030] Step S31: Use a neural network to construct a mapping model to characterize the nonlinear relationship between the structural parameters and the dynamic and static stiffness of the main shaft. The BP neural network modifies the weights and thresholds after performing error analysis based on the results obtained from the structural parameter training and the expected results. The training is repeated multiple times to output a parameter model consistent with the expected results.

[0031] Step S32: Sorting the structural parameters that affect the radial stiffness and the dynamic stiffness respectively in descending order according to the magnitude of the influence relationship, selecting the top three structural parameters that are most affected as the input variables of the mapping model, and under the constraints of the spindle performance index, selecting the maximum displacement and first-order natural frequency corresponding to the spindle, and the five parameters of the set constraint condition index spindle mass, displacement amplitude and maximum stress as the output variables of the final mapping model;

[0032] Step S33: After preprocessing the input data in combination with the genetic algorithm, the BP neural network determines the corresponding network topology, initializes the BP neural network weights and threshold length, and performs multiple network training to obtain the error value, which is used as the fitness value for subsequent prediction output.

[0033] According to the above technical solution, in step S33, since the standard BP neural network has the defect of slow convergence speed and easy to fall into the local optimal solution leading to training failure, the global optimization and local optimization characteristics of the genetic algorithm are used to first determine a better search space in the solution space, so that the network training is in the optimal solution position on the error surface from the beginning.

[0034] According to the above technical solution, step S4 further includes the following steps:

[0035] Step S41: Selecting the machine tool spindle structural parameters as design variables, establishing a spindle structural optimization mathematical model, optimizing the machine tool spindle structural parameters, applying the improved spindle performance mapping model to the constraint conditions, and obtaining the optimal structural parameter level combination;

[0036] Step S42: Utilizing the analysis results of the sensitivity of the spindle structural parameters to the dynamic and static stiffness, combined with the spindle performance neural network model, the spindle is optimized while ensuring that the structural quality of the machine tool spindle remains unchanged, so that the static and dynamic stiffness of the machine tool spindle are both at an optimal level.

[0037] Compared with the prior art, the beneficial effects achieved by the present invention are as follows: the present invention, by providing a spindle structural parameter module, a comprehensive stiffness mapping module and a static and dynamic stiffness optimization module, simplifies the three-dimensional modeling of the spindle, uses a single variable method to obtain the spindle structural parameters that significantly affect the static and dynamic stiffness changes of the spindle, then comprehensively considers the static and dynamic stiffness of the spindle, uses a neural network to establish a spindle performance mapping model, improves the neural network model through a BP neural network model combined with a genetic algorithm, obtains the relationship between the spindle structural parameters and the performance indicators, and finally establishes a spindle structural optimization mathematical model to optimize the structure of the machine tool spindle. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0039] Figure 1 It is a schematic diagram of the system module composition of the present invention. DETAILED DESCRIPTION

[0040] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0041] See also Figure 1The present invention provides a technical solution: a spindle stiffness test method based on dynamic load, the method adopts a spindle stiffness test system as a whole, the spindle stiffness test system includes a spindle structural parameter module, a comprehensive stiffness mapping module and a dynamic and static stiffness optimization module, the spindle structural parameter module is used to obtain the structural parameters affecting the dynamic and static stiffness of the spindle according to the spindle deformation displacement and natural frequency, the comprehensive stiffness mapping module is used to use neural network and genetic algorithm to establish a spindle performance parameter mapping model with the obtained structural parameters, the dynamic and static stiffness optimization module is used to optimize the spindle so that the static and dynamic stiffness of the machine tool spindle is at an optimal level, and the spindle structure The structural parameter module is electrically connected to the comprehensive stiffness mapping module, and the comprehensive stiffness mapping module is electrically connected to the dynamic and static stiffness optimization module. The dynamic and static stiffness of the spindle directly affects the processing accuracy and efficiency of the subsequent machine tools. By establishing a simplified three-dimensional model of the spindle and applying corresponding loads to obtain the radial deformation displacement of the spindle, and through the characteristics that the dynamic stiffness of the spindle increases with the frequency, the overall structural parameters that have a greater impact on the spindle are obtained. This is used as the input variable of the neural network and combined with the performance parameters of the spindle itself as the output variable for model training and output. Finally, the spindle is optimized according to the model to improve the subsequent processing accuracy and efficiency.

[0042] The spindle structural parameter module includes a three-dimensional modeling simplification module, a spindle static deformation module and a spindle dynamic stiffness module. The three-dimensional modeling simplification module is used to simplify the spindle structure and perform three-dimensional modeling. The spindle static deformation module is used to obtain the deformation displacement caused by the relevant structural parameters in the radial direction of the spindle under the action of load. The spindle dynamic stiffness module is used to obtain the relevant structural parameters that affect the spindle under the action of load by using the positive relationship between the natural frequency and dynamic stiffness of the spindle. The three-dimensional modeling simplification module is electrically connected to the spindle static deformation module, and the spindle static deformation module is electrically connected to the spindle dynamic stiffness module. Some spindle structures that may affect the subsequent accuracy are deleted, simple three-dimensional modeling is performed, and loads and constraints under corresponding working conditions are applied to obtain the structural parameters that affect the dynamic and static stiffness in the radial and axial directions of the spindle.

[0043] The comprehensive stiffness mapping module includes a structural parameter module, a parameter variable module, a BP neural network module and a genetic algorithm combination module. The structural parameter module is used to record and sort the obtained structural parameters that affect the dynamic and static stiffness of the main shaft. The parameter variable module is used to select appropriate structural parameters and performance parameters as input and output variables of the model. The BP neural network module is used to select a parameter model whose output is consistent with the expected result after the BP neural network trains the structural parameters for multiple times. The genetic algorithm combination module is used to apply the genetic algorithm to make up for the defects of the neural network that the convergence speed is slow and it is easy to fall into the local optimal solution and cause the training failure. The structural parameter module is electrically connected to the parameter variable module, and the BP neural network module is electrically connected to the genetic algorithm combination module.

[0044] The parameter variable module includes an input variable submodule and an output variable submodule. The input variable submodule is used to select the most influential parameter among many influential structural parameters as the input variable of the neural network, and the output variable submodule is used to select the performance parameters of the spindle itself as the output variable of the neural network for model training. The input variable submodule is electrically connected to the output variable submodule.

[0045] The dynamic and static stiffness optimization module includes a structural optimization module, a parameter level combination module and a spindle optimization module. The structural optimization module is used to select the structural parameters of the machine tool spindle as design variables and establish a spindle structural optimization mathematical model. The parameter level combination module is used to apply the improved spindle performance mapping model to the constraint conditions to obtain the optimal structural parameter level combination. The spindle optimization module is used to optimize the spindle according to the structural parameters that affect the dynamic and static stiffness of the spindle and the spindle performance neural network model. The structural optimization module is electrically connected to the parameter level combination module, and the parameter level combination module is electrically connected to the spindle optimization module.

[0046] A spindle stiffness testing method based on dynamic load, the spindle stiffness testing method comprising the following steps:

[0047] Step S1: Establish a simplified three-dimensional model of the spindle on the stiffness test bench, and perform a force analysis on the machine tool spindle according to the load and constraint conditions under actual working conditions;

[0048] Step S2: the spindle is controlled by an industrial computer, and the displacement change data in the axial deformation direction of the spindle is collected to screen out the structural parameters that have a significant impact on the change of the radial stiffness of the spindle, and the positive correlation between the natural frequency of the spindle and the dynamic stiffness is used to screen out the structural parameters that have a significant impact on the change of the dynamic stiffness of the spindle;

[0049] Step S3: using the structure parameters obtained by screening as input variables and the spindle performance parameters as output variables, a BP neural network model of the machine tool spindle is established, and the model is optimized by combining a genetic algorithm;

[0050] Step S4: Establish a structural optimization mathematical model, optimize the structural parameters of the machine tool spindle, and obtain the optimal structural parameter level combination.

[0051] Step S1 further comprises the following steps:

[0052] Step S11: When establishing the three-dimensional model, the tiny features on the spindle components are deleted to simplify the structure of the machine tool spindle. The tiny features on the spindle require extremely high precision for mesh division, which makes the mesh after division dense, resulting in an exponential increase in the amount of calculation. It is difficult for ordinary industrial computers to process and the accuracy of the calculation cannot be guaranteed, so it is deleted;

[0053] Step S12: Select a smaller overhang under the premise of meeting the spindle structure requirements. The spindle stiffness is affected by the bearing overhang. The overhang refers to the distance from the front end of the spindle to the midpoint of the front support of the spindle. As the front end bearing overhang increases, the spindle stiffness decreases.

[0054] Step S13: applying constraints and loads to the spindle by using an external cylindrical cutting processing method, simulating the stress state of the spindle under corresponding working conditions, and performing stress analysis on the spindle;

[0055] Step S14: During the process of the spindle being subjected to force and deformation, the right horizontal X-axis, the vertical Z-axis and the front horizontal Y-axis are established at the spindle core. The spindle itself and the front support bearing and the rear support bearing are all subjected to force and deformed, causing the spindle end to be displaced. The displacement of the end of the spindle includes the deformation of the spindle end caused by the bending deformation of the spindle itself and the deformation of the bearing.

[0056] Step S2 further comprises the following steps:

[0057] Step S21: Using the parameter drive of the three-dimensional model, collect and record any structural parameter set of the main axis {D1, D2, ...D n The corresponding principal axis radial deformation force set {L1, L2, ...L n};

[0058] Step S22: Calculate the influence of the structural parameters on the maximum deformation in the X direction of the spindle, and use the spindle radial stiffness to represent the influence. The spindle radial stiffness calculation formula is:

[0059]

[0060] In the formula, F i is the radial stiffness of the spindle, Δr is the load on the spindle set under different working conditions, and L max The maximum deformation amount of the spindle radial deformation force concentration. When the spindle load is constant, the spindle radial stiffness is affected by the spindle maximum deformation. The X-axis direction is the spindle sensitive deformation direction. The change of the spindle maximum deformation in the X direction can clearly reflect the change of the spindle's final maximum deformation.

[0061] Step S23: By performing constrained modal analysis on the original model of the main shaft, the axial deformation force set {l1, l2, ...l n}, Modal analysis is the process of analyzing modal parameters. Mode refers to the natural vibration characteristics of a mechanical structure. Each mode has a specific natural frequency, damping ratio, and modal vibration shape;

[0062] Step S24: Calculate the stiffness of the main shaft under the action of the sinusoidal alternating load, that is, calculate the dynamic stiffness F as a function of the excitation frequency.j , the calculation formula is:

[0063]

[0064] Where A is the amplitude of the main shaft vibration under load, l i is the axial deformation force corresponding to the excitation frequency function of the spindle, ω is the corresponding natural frequency in the first order. The natural frequency value of the spindle increases with the increase of the order, and the dynamic characteristics of mechanical products are mainly determined by the low-order modes. Therefore, the amplitude value A of the vibration of the spindle under the load and the axial deformation force of each parameter at the first-order natural frequency are selected to characterize the dynamic stiffness of the spindle. That is, at the first-order natural frequency, the greater the axial deformation force, the better the dynamic characteristics of the spindle.

[0065] Step S3 further comprises the following steps:

[0066] Step S31: A mapping model is constructed using a neural network to characterize the nonlinear relationship between the structural parameters and the dynamic and static stiffness of the spindle. The BP neural network modifies the weights and thresholds after performing error analysis based on the results obtained from the structural parameter training and the expected results. The training is repeated multiple times and a parameter model consistent with the expected results is output. When studying the mapping relationship between the structural parameters and the spindle stiffness, the structural parameters that can quickly respond to the dynamic and static stiffness of the spindle should be selected as input quantities. When the structural parameters change, the corresponding dynamic and static stiffness spindle performance indicators can be quickly calculated through the model as output quantities.

[0067] Step S32: Sorting the structural parameters that affect the radial stiffness and the dynamic stiffness respectively in descending order according to the magnitude of the influence relationship, selecting the top three structural parameters that are most affected as the input variables of the mapping model, and under the constraints of the spindle performance index, selecting the maximum displacement and first-order natural frequency corresponding to the spindle, and the five parameters of the set constraint condition index spindle mass, displacement amplitude and maximum stress as the output variables of the final mapping model;

[0068] Step S33: After preprocessing the input data in combination with the genetic algorithm, the BP neural network determines the corresponding network topology, and initializes the BP neural network weights and threshold length, performs multiple network training to obtain the error value, and uses it as the fitness value for subsequent prediction output. The BP neural network, as a multi-layer feedforward neural network, has the greatest feature that it does not require precise function expressions but instead performs simulation training and weight training through a large amount of data learning. The mapping relationship between the connection threshold and weight between neurons is established with a simple structure and fast calculation speed, and can solve multi-objective optimization problems more conveniently.

[0069] In step S33, since the standard BP neural network has the defects of slow convergence speed and easy to fall into the local optimal solution leading to training failure, the global optimization and local optimization characteristics of the genetic algorithm are used to first determine a better search space in the solution space, so that the network training is at the optimal solution position on the error surface from the beginning. In the genetic algorithm, the fitness function is used to evaluate the fitness of the individual to the environment. The larger the fitness function value, the better the individual. When performing the selection operation, the fitness is used as the basis for the survival of the fittest in the group. Individuals with high fitness parameters are more likely to be inherited to the next generation group, and individuals with low fitness are less likely to be inherited to the next generation.

[0070] Step S4 further comprises the following steps:

[0071] Step S41: Selecting the machine tool spindle structural parameters as design variables, establishing a spindle structural optimization mathematical model, optimizing the machine tool spindle structural parameters, applying the improved spindle performance mapping model to the constraint conditions, and obtaining the optimal structural parameter level combination;

[0072] Step S42: Utilizing the analysis results of the sensitivity of the spindle structural parameters to the dynamic and static stiffness, combined with the spindle performance neural network model, the spindle is optimized while ensuring that the structural quality of the machine tool spindle remains unchanged, so that the static and dynamic stiffness of the machine tool spindle are both at an optimal level.

[0073] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0074] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A spindle stiffness testing method based on dynamic load, characterized in that: The method adopts a spindle stiffness test system as a whole, and the spindle stiffness test system includes a spindle structural parameter module, a comprehensive stiffness mapping module and a dynamic and static stiffness optimization module. The spindle structural parameter module is used to obtain structural parameters that affect the dynamic and static stiffness of the spindle according to the spindle deformation displacement and natural frequency. The comprehensive stiffness mapping module is used to establish a spindle performance parameter mapping model with the obtained structural parameters using a neural network and a genetic algorithm. The dynamic and static stiffness optimization module is used to optimize the spindle so that the static and dynamic stiffness of the machine tool spindle is at an optimal level. The spindle structural parameter module is electrically connected to the comprehensive stiffness mapping module, and the comprehensive stiffness mapping module is electrically connected to the dynamic and static stiffness optimization module. The spindle stiffness testing method comprises the following steps: Step S1: Establish a simplified three-dimensional model of the spindle on the stiffness test bench, and perform a force analysis on the machine tool spindle according to the load and constraint conditions under actual working conditions; Step S2: the spindle is controlled by an industrial computer, and the displacement change data in the axial deformation direction of the spindle is collected to screen out the structural parameters that have a significant impact on the change of the radial stiffness of the spindle, and the positive correlation between the natural frequency of the spindle and the dynamic stiffness is used to screen out the structural parameters that have a significant impact on the change of the dynamic stiffness of the spindle; Step S3: using the structure parameters obtained by screening as input variables and the spindle performance parameters as output variables, a BP neural network model of the machine tool spindle is established, and the model is optimized by combining a genetic algorithm; Step S4: Establishing a structural optimization mathematical model, optimizing the structural parameters of the machine tool spindle, and obtaining the optimal structural parameter level combination; The step S2 further comprises the following steps: Step S21: Apply the parameter drive of the three-dimensional model to collect and record any structural parameter set D1, D2, ... D n The corresponding principal axis radial deformation force set in the model is L1, L2, ...L n ; Step S22: Calculate the influence of the structural parameters on the maximum deformation in the X direction of the spindle, and use the spindle radial stiffness to represent the influence. The spindle radial stiffness calculation formula is: In the formula, F i is the radial stiffness of the spindle, Δr is the load on the spindle set under different working conditions, and L max The maximum deformation amount of the main axis radial deformation force concentration; Step S23: By performing constrained modal analysis on the original model of the main shaft, the axial deformation force set {l1, l2, ...l n }; Step S24: Calculate the stiffness of the main shaft under the action of the sinusoidal alternating load, that is, calculate the dynamic stiffness F as a function of the excitation frequency. j , the calculation formula is: Where A is the amplitude of the main shaft vibration under load, l i is the axial deformation force of the main shaft corresponding to the excitation frequency function, and ω is the corresponding natural frequency in the first order.

2. A spindle stiffness testing method based on dynamic load according to claim 1, characterized in that: The spindle structural parameter module includes a three-dimensional modeling simplification module, a spindle static deformation module and a spindle dynamic stiffness module. The three-dimensional modeling simplification module is used to simplify the spindle structure and three-dimensionally model it. The spindle static deformation module is used to obtain the deformation displacement caused by the relevant structural parameters in the radial direction of the spindle under the action of load. The spindle dynamic stiffness module is used to obtain the relevant structural parameters that affect the spindle by utilizing the positive phase relationship between the natural frequency and dynamic stiffness of the spindle under the action of load. The three-dimensional modeling simplification module is electrically connected to the spindle static deformation module, and the spindle static deformation module is electrically connected to the spindle dynamic stiffness module.

3. A spindle stiffness testing method based on dynamic load according to claim 2, characterized in that: The comprehensive stiffness mapping module includes a structural parameter module, a parameter variable module, a BP neural network module and a genetic algorithm combination module. The structural parameter module is used to record and sort the obtained structural parameters that affect the dynamic and static stiffness of the main shaft. The parameter variable module is used to select appropriate structural parameters and performance parameters as input and output variables of the model. The BP neural network module is used to select a parameter model that outputs a consistent result with the expected result after the BP neural network performs multiple training on the structural parameters. The genetic algorithm combination module is used to apply the genetic algorithm to compensate for the defects of the neural network that the convergence speed is slow and it is easy to fall into the local optimal solution and cause the training failure. The structural parameter module is electrically connected to the parameter variable module, and the BP neural network module is electrically connected to the genetic algorithm combination module. The parameter variable module includes an input variable submodule and an output variable submodule. The input variable submodule is used to select the parameter with the greatest influence among many influential structural parameters as the input variable of the neural network, and the output variable submodule is used to select the performance parameters of the main shaft itself as the output variable of the neural network for model training. The input variable submodule is electrically connected to the output variable submodule.

4. A spindle stiffness testing method based on dynamic load according to claim 3, characterized in that: The dynamic and static stiffness optimization module includes a structural optimization module, a parameter level combination module and a spindle optimization module. The structural optimization module is used to select the machine tool spindle structural parameters as design variables and establish a spindle structural optimization mathematical model. The parameter level combination module is used to apply the improved spindle performance mapping model to the constraint conditions to obtain the optimal structural parameter level combination. The spindle optimization module is used to optimize the spindle according to the structural parameters that affect the dynamic and static stiffness of the spindle and the spindle performance neural network model. The structural optimization module is electrically connected to the parameter level combination module, and the parameter level combination module is electrically connected to the spindle optimization module.

5. A spindle stiffness testing method based on dynamic load according to claim 4, characterized in that: The step S1 further comprises the following steps: Step S11: when establishing the three-dimensional model, the tiny features on the spindle components are deleted to simplify the structure of the machine tool spindle; Step S12: Selecting a smaller overhang under the premise of meeting the spindle structure requirements, the spindle stiffness is affected by the bearing overhang; Step S13: applying constraints and loads to the spindle by using an external cylindrical cutting processing method, simulating the stress state of the spindle under corresponding working conditions, and performing stress analysis on the spindle; Step S14: During the process of the spindle being deformed by force, the right horizontal X-axis, the vertical Z-axis and the front horizontal Y-axis are established at the spindle core. The spindle itself and the front and rear support bearings are deformed by force, causing the spindle shaft end to be displaced.

6. A spindle stiffness testing method based on dynamic load according to claim 5, characterized in that: The step S3 further comprises the following steps: Step S31: Use a neural network to construct a mapping model to characterize the nonlinear relationship between the structural parameters and the dynamic and static stiffness of the main shaft. The BP neural network modifies the weights and thresholds after performing error analysis based on the results obtained from the structural parameter training and the expected results. The training is repeated multiple times to output a parameter model consistent with the expected results. Step S32: Sorting the structural parameters that affect the radial stiffness and the dynamic stiffness respectively in descending order according to the magnitude of the influence relationship, selecting the top three structural parameters that are most affected as the input variables of the mapping model, and under the constraints of the spindle performance index, selecting the maximum displacement and first-order natural frequency corresponding to the spindle, and the five parameters of the set constraint condition index spindle mass, displacement amplitude and maximum stress as the output variables of the final mapping model; Step S33: After preprocessing the input data in combination with the genetic algorithm, the BP neural network determines the corresponding network topology, initializes the BP neural network weights and threshold length, and performs multiple network training to obtain the error value, which is used as the fitness value for subsequent prediction output.

7. The spindle stiffness testing method based on dynamic load according to claim 6, characterized in that: In step S33, since the standard BP neural network has the defect of slow convergence speed and easy to fall into the local optimal solution leading to training failure, the global optimization and local optimization characteristics of the genetic algorithm are used to first determine a better search space in the solution space, so that the network training is at the optimal solution position on the error surface from the beginning.

8. The spindle stiffness testing method based on dynamic load according to claim 7, characterized in that: The step S4 further comprises the following steps: Step S41: Selecting the machine tool spindle structural parameters as design variables, establishing a spindle structural optimization mathematical model, optimizing the machine tool spindle structural parameters, applying the improved spindle performance mapping model to the constraint conditions, and obtaining the optimal structural parameter level combination; Step S42: Utilizing the analysis results of the sensitivity of the spindle structural parameters to the dynamic and static stiffness, combined with the spindle performance neural network model, the spindle is optimized while ensuring that the structural quality of the machine tool spindle remains unchanged, so that the static and dynamic stiffness of the machine tool spindle are both at an optimal level.

Images