Multi-axis linkage based precision milling closed-loop optimization and real-time control process for ball screw

Abstract

The application provides a kind of based on multi-axis linkage's precision milling closed loop optimization and real-time control process of screw rod, including collecting workpiece geometric parameters, material properties and machine tool dynamics parameters, designing tool and fixture scheme, and setting initial machining parameters based on cutting database;Utilize NURBS curve to construct high-precision curved surface model of screw rod helical groove, optimize path through curvature adaptive discretization and quintic spline interpolation;Adopt real number coding genetic algorithm to iteratively optimize idle stroke path, shorten total path time and reversing frequency;Realize multi-axis synchronous control through inverse kinematics solution, and dynamically adjust tool inclination angle based on contact angle in real time;Use virtual simulation and finite element analysis to predict cutting force, thermal deformation and interference risk.The application generates continuous smooth multi-axis synchronous command, effectively reduces the positioning error of each axis, and dynamically corrects the tool inclination angle through the force feedback of the contact angle, effectively suppresses the cutting force fluctuation, and ensures that the roughness of helical groove tooth surface meets the requirements.

Description

Technical Field

[0001] This invention relates to the field of industrial production operation and control technology, specifically to a closed-loop optimization and real-time control process for precision milling of lead screws based on multi-axis linkage. Background Technology

[0002] In the field of precision milling of lead screws, traditional processes generally adopt a machining strategy of equally spaced discrete linear interpolation path planning and fixed tool tilt angle. However, traditional processes still have some limitations:

[0003] During machining, sudden acceleration changes are likely to occur in areas where the lead screw curvature changes drastically, leading to machine tool vibration and surface rippling. Existing machining paths frequently change direction and have high redundancy, resulting in extended machining time. Traditional inverse kinematics calculations rely on approximate linearized models and ignore the nonlinear coupling effect of the rotation axis (B / C axis), leading to the accumulation of tool pose errors, especially causing tooth surface interference when machining helical grooves with variable pitch. Furthermore, the current real-time monitoring and feedback mechanism for the cutting contact angle is imperfect, and the fixed tool tilt angle leads to fluctuations in cutting force, accelerating tool wear and inducing thermal deformation.

[0004] Therefore, there is an urgent need for a ball screw milling process that integrates high-precision path planning, multi-axis dynamic collaborative control, and data closed-loop optimization to overcome the comprehensive bottlenecks in machining accuracy, efficiency, and stability. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a closed-loop optimization and real-time control process for precision milling of lead screws based on multi-axis linkage, in order to solve the problems mentioned in the background technology.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a closed-loop optimization and real-time control process for precision milling of lead screws based on multi-axis linkage, comprising the following steps:

[0007] Collect workpiece geometric parameters, material properties, and machine tool dynamic parameters; design tool and fixture schemes; and set initial machining parameters based on the cutting database.

[0008] A high-precision surface model of the screw helical groove is constructed using NURBS curves, and the path is optimized through curvature adaptive discretization and quintic spline interpolation.

[0009] A real-number encoded genetic algorithm is used to iteratively optimize empty travel paths, thereby reducing the total path time and the number of reversals.

[0010] Multi-axis synchronous control is achieved through inverse kinematics calculation, and the tool tilt angle is dynamically adjusted in real time based on the contact angle;

[0011] Virtual simulation and finite element analysis are used to predict cutting forces, thermal deformation, and interference risks, and process parameters are corrected accordingly.

[0012] Perform multi-axis linkage machining and monitor the load in real time; detect lead, roughness and tooth profile accuracy through online measurement and coordinate measuring machine.

[0013] The path smoothness coefficient and genetic algorithm weights are optimized by combining processing data feedback, and the process knowledge base is updated.

[0014] As a preferred embodiment, the specific process of constructing a high-precision surface model of the lead screw helical groove using NURBS curves and optimizing the path through curvature adaptive discretization and quintic spline interpolation includes:

[0015] Based on the lead screw radius, lead and helix angle, a geometric model of the helical groove is generated by the NURBS curve parametric equation, and control points, weights and node vectors are defined to accurately represent the surface.

[0016] Calculate the curvature of each point on the NURBS curve, dynamically adjust the step size, and generate a high-density discrete point set to match geometric features;

[0017] The discrete points are segmented and fitted into a quintic spline curve, and the position, velocity, acceleration and jerk are forced to be continuous between adjacent segments. The optimization objective is to minimize the change in the total acceleration of the path.

[0018] Solve for spline coefficients and generate a smooth G-code path, which includes coordinate, velocity, and acceleration commands.

[0019] As a preferred embodiment, the curvature k(u) of the curve at parameter u is calculated using the following formula:

[0020] ;

[0021] in, express, and Let represent the first and second derivatives of the curve at u, respectively;

[0022] The step size Δu is dynamically adjusted based on curvature, and the formula for calculating the step size is:

[0023] ;

[0024] in, For the maximum allowable step size, Represents the curve sensitivity coefficient. This represents the parameter value at the current discrete point.

[0025] As a preferred embodiment, the detailed process of using a real-number encoded genetic algorithm to iteratively optimize the empty travel path and shorten the total path time and the number of reversals includes:

[0026] The three-dimensional coordinate sequence of the empty travel path is encoded into a real number vector, and an initial population that satisfies the machine tool motion range constraint is randomly generated.

[0027] Path performance is quantified by a weighted sum of total path time and number of reversals.

[0028] The tournament selection process retains high-quality individuals, arithmetic crossover and Gaussian mutation introduce diversity, and out-of-bounds coordinates are corrected to ensure path feasibility.

[0029] The optimization process stops when the number of iterations reaches a preset threshold or the optimal fitness remains stable.

[0030] The real number sequence of the individual with the best fitness is decoded, a three-dimensional path is generated, and the path is transmitted to the actuator.

[0031] As a preferred embodiment, the process of quantifying path performance based on the weighted sum of total path time and number of reversals includes:

[0032] A fitness function is constructed based on total path time and number of reversals, and the performance of idle journeys is quantified by weighted summation.

[0033] The formula for calculating the fitness function is:

[0034] ;

[0035] Where F represents the weighted sum, T represents the total path time, and Ndir represents the number of reversals.

[0036] As a preferred embodiment, the specific process of achieving multi-axis synchronous control through inverse kinematics calculation and dynamically adjusting the tool tilt angle in real time based on the contact angle includes:

[0037] Construct a geometric kinematic model of the machine tool, and define the spatial mapping relationship and constraints between the tool coordinate system and each motion axis.

[0038] Based on the target path point coordinates and tool attitude angle, the real-time displacement parameters of each motion axis are derived through inverse kinematics calculation.

[0039] Multi-axis synchronous control commands are generated based on displacement parameters, and an interpolation algorithm is used to achieve dynamic matching and coordination of the motion trajectories of each axis.

[0040] The tool-workpiece contact angle is acquired in real time using a force sensor or vision measurement device, and the tilt angle deviation is determined by comparing it with the allowable threshold of the process.

[0041] If the contact angle exceeds the threshold, the tool tilt angle parameter is dynamically corrected and fed back to the inverse kinematics model.

[0042] As a preferred embodiment, the process of deriving the real-time displacement parameters of each motion axis through inverse kinematics calculation includes:

[0043] The machining path planning provides the target pose of the tool end in the base coordinate system;

[0044] By using the machine tool geometric model, the Jacobian matrix is ​​established as the differential relationship between the tool pose change and the displacement change of each axis.

[0045] The axial displacement increment is calculated by solving the pseudo-inverse of the Jacobian matrix;

[0046] The formula for calculating the displacement increment is:

[0047] ;

[0048] in, , ;

[0049] In the formula, represents the displacement increment vector of each axis. Describe the pseudo-inverse of the Jacobian matrix. The deviation vector between the target pose and the current actual pose. Indicates the current actual position of the tool;

[0050] Through the formula: Update axis displacement parameters;

[0051] in, This represents the axis displacement parameter after the k-th iteration, and the update termination condition is: .

[0052] As a preferred embodiment, the process of using virtual simulation and finite element analysis to predict cutting forces, thermal deformation, and interference risks, and to correct process parameters, includes:

[0053] Based on the workpiece CAD model, tool geometry parameters and material constitutive relations, a finite element multiphysics coupled simulation model including cutting force, heat conduction and contact behavior is constructed.

[0054] The virtual simulation platform is driven to perform dynamic simulation of the cutting process, and the cutting force time-domain waveform, thermal deformation gradient field and spatial distribution of tool-workpiece interference region are analyzed simultaneously.

[0055] The peak cutting force, extreme temperature, and minimum interference distance output by the simulation are compared with the preset process safety threshold to generate a risk level matrix and a list of key sensitive parameters.

[0056] Based on the risk matrix, the spindle speed curve, feed acceleration curve, and coolant injection strategy are optimized to generate a process parameter correction scheme under multi-objective constraints.

[0057] As a preferred embodiment, the process of combining processing data feedback to optimize the path smoothness coefficient and the genetic algorithm weights, and updating the process knowledge base, includes:

[0058] Based on the lead deviation distribution, tooth profile error spectrum, and path optimization parameters generated by the genetic algorithm in step three, a feature matrix including path smoothing coefficient, time-commutation weight ratio, and process parameter coupling degree is constructed.

[0059] By analyzing the nonlinear relationship between lead error and λ and the interaction effect between tooth profile accuracy and time-commutation weight ratio through partial least squares regression, a correlation response surface model of error source-process parameters is established.

[0060] Based on the response surface model, a multi-objective particle swarm optimization algorithm is used to simultaneously optimize the smoothness coefficient of the quintic spline path and the time-commutation weight ratio of the genetic algorithm.

[0061] The optimized parameter combinations and corresponding machining quality data are encoded as rule entries and stored in a relational database according to the material type-tool combination-machine tool model triplet.

[0062] When the coefficient of variation (CV) of a new workpiece batch is greater than or equal to 5%, the system automatically matches the historical best parameter combination based on the updated knowledge base and triggers a re-optimization process.

[0063] This invention provides a closed-loop optimization and real-time control process for precision ball screw milling based on multi-axis linkage, which has the following advantages: Based on inverse kinematics calculation and a pseudo-inverse iterative algorithm using the Jacobian matrix, a real-time mapping model of tool pose and axis displacement is constructed. Combined with a quintic spline interpolation algorithm, continuous and smooth multi-axis synchronous commands are generated, effectively reducing positioning errors of each axis. Simultaneously, the tool tilt angle is dynamically corrected through contact angle force feedback, effectively suppressing cutting force fluctuations and ensuring that the surface roughness of the helical groove meets requirements. In path planning, NURBS curve curvature adaptive discretization technology and quintic spline smoothing optimization are adopted, combined with a real-number encoded genetic algorithm to optimize the idle travel path, significantly improving machining efficiency and machine tool motion stability. In closed-loop optimization, a response surface model is constructed based on cutting simulation and online measurement data. The path smoothing coefficient is synchronously optimized using a multi-objective particle swarm optimization algorithm, achieving high lead accuracy and low tooth profile error precision machining. Simultaneously, relying on a process knowledge base, the optimal historical parameter combination is automatically matched for new batches of workpieces, reducing process debugging time and comprehensively improving machining quality stability and production efficiency. Attached Figure Description

[0064] Figure 1 This is a flowchart of the closed-loop optimization and real-time control process for precision milling of ball screws based on multi-axis linkage, as described in this invention. Detailed Implementation

[0065] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0066] The following disclosure provides many different embodiments or examples for implementing various structures of the invention. To simplify the disclosure, specific examples of components and arrangements are described below. These are merely examples and are not intended to limit the invention. Furthermore, reference numerals and / or letters may be repeated in different examples; such repetition is for simplification and clarity and does not in itself indicate a relationship between the various embodiments and / or arrangements discussed. In addition, examples of various specific processes and materials are provided in this invention, but those skilled in the art will recognize the application of other processes and / or the use of other materials.

[0067] like Figure 1 As shown, this embodiment of the invention provides a closed-loop optimization and real-time control process for precision milling of lead screws based on multi-axis linkage, including the following steps:

[0068] Step 1: Collect workpiece geometric parameters, material properties, and machine tool dynamic parameters; design tool and fixture schemes; and set initial machining parameters based on the cutting database.

[0069] Specifically, this involves obtaining the geometric parameters (lead, thread angle, length, etc.) and material properties (hardness, coefficient of thermal expansion, etc.) of the lead screw workpiece, as well as the dynamic parameters (maximum speed, acceleration, and accuracy range of each axis) of the multi-axis CNC machine tool.

[0070] Choose either a ball end mill or a form milling cutter, and determine the cutter diameter, cutting edge length, and coating type; design a dedicated fixture to ensure workpiece clamping stability and coaxiality ≤0.005 mm.

[0071] Based on the material cutting database, preset spindle speed, feed rate, depth of cut and coolant parameters are used.

[0072] Step 2: Construct a high-precision surface model of the lead screw helical groove using NURBS curves, and optimize the path through curvature adaptive discretization and quintic spline interpolation;

[0073] In this embodiment, the specific process of step two includes: generating a helical groove geometric model based on the lead screw radius, lead and helix angle using NURBS curve parametric equations, and defining control points, weights and node vectors to accurately characterize the surface;

[0074] Specifically, the spiral is discretized into m control points {Pi | i = 0, 1, 2, ..., m-1} to construct a non-uniform rational B-spline (NURBS) curve;

[0075] Through the curve equation:

[0076] ;

[0077] in, This represents the coordinates of the curve points corresponding to parameter u∈[0,1]. () represents the i-th p-th order B-spline basis function. Indicates the coordinates of the control points. represents the control point weights, and p represents the curve order.

[0078] Calculate the curvature of each point on the NURBS curve, dynamically adjust the step size (smaller step size for larger curvature, larger step size for smaller curvature), and generate a high-density discrete point set to match geometric features;

[0079] In this embodiment, the curvature k(u) of the curve at parameter u is calculated using the following formula:

[0080] ;

[0081] in, express, and Let represent the first and second derivatives of the curve at u, respectively;

[0082] The step size Δu is dynamically adjusted based on curvature, and the formula for calculating the step size is:

[0083] ;

[0084] in, For the maximum allowable step size, Represents the curve sensitivity coefficient. This represents the parameter value at the current discrete point;

[0085] The discrete points are segmented and fitted into a quintic spline curve, and the position, velocity, acceleration and jerk are forced to be continuous between adjacent segments. The optimization objective is to minimize the change in the total acceleration of the path.

[0086] Solve for spline coefficients and generate a smooth G-code path, which includes coordinate, velocity, and acceleration commands.

[0087] It should be noted that after the path is output, it also includes verification of the path's geometric accuracy, motion stability, and processing effect. The specific verification methods are existing technical means, which will not be described in detail here.

[0088] Step 3: Use a real-number encoded genetic algorithm to iteratively optimize the empty travel path, thereby reducing the total path time and the number of reversals;

[0089] Specifically, the detailed process of using a real-number encoded genetic algorithm to iteratively optimize empty travel paths and shorten the total path time and number of reversals includes:

[0090] The three-dimensional coordinate sequence of the empty travel path is encoded into a real number vector, and an initial population that satisfies the machine tool motion range constraint is randomly generated.

[0091] The path performance is quantified by a weighted sum of the total path time (the sum of segmented distances and speeds) and the number of direction changes (the number of times the angle between the direction vectors exceeds the limit).

[0092] The tournament selection process retains high-quality individuals, arithmetic crossover and Gaussian mutation introduce diversity, and out-of-bounds coordinates are corrected to ensure path feasibility.

[0093] The optimization process stops when the number of iterations reaches a preset threshold or the optimal fitness remains stable.

[0094] Decode the real number sequence of the individual with the best fitness, generate a three-dimensional path, and transmit it to the actuator;

[0095] Specifically, the process of quantifying path performance based on a weighted sum of total path time and number of reversals includes:

[0096] A fitness function is constructed based on total path time and number of reversals, and the performance of idle journeys is quantified by weighted summation.

[0097] The formula for calculating the fitness function is:

[0098] ;

[0099] Where F represents the weighted sum, T represents the total path time, and Ndir represents the number of reversals.

[0100] Step 4: Achieve multi-axis synchronous control through inverse kinematics calculation, and dynamically adjust the tool tilt angle in real time based on the contact angle;

[0101] Specifically, the process of achieving multi-axis synchronous control through inverse kinematics calculation and dynamically adjusting the tool tilt angle in real time based on the contact angle includes:

[0102] Construct a geometric kinematic model of the machine tool, and define the spatial mapping relationship and constraints between the tool coordinate system and each motion axis.

[0103] Based on the target path point coordinates and tool attitude angle, the real-time displacement parameters of each motion axis are derived through inverse kinematics calculation.

[0104] Multi-axis synchronous control commands are generated based on displacement parameters, and an interpolation algorithm is used to achieve dynamic matching and coordination of the motion trajectories of each axis.

[0105] The tool-workpiece contact angle is acquired in real time using a force sensor or vision measurement device, and the tilt angle deviation is determined by comparing it with the allowable threshold of the process.

[0106] If the contact angle exceeds the threshold, the tool tilt angle parameter is dynamically corrected and fed back to the inverse kinematics model.

[0107] Specifically, the process of deriving the real-time displacement parameters of each motion axis through inverse kinematics calculation includes:

[0108] The machining path planning provides the target pose of the tool end in the base coordinate system;

[0109] By using the machine tool geometric model, the Jacobian matrix is ​​established as the differential relationship between the tool pose change and the displacement change of each axis.

[0110] The axial displacement increment is calculated by solving the pseudo-inverse of the Jacobian matrix;

[0111] The formula for calculating the displacement increment is:

[0112] ;

[0113] in, , ;

[0114] In the formula, represents the displacement increment vector of each axis. Describe the pseudo-inverse of the Jacobian matrix. The deviation vector between the target pose and the current actual pose. Indicates the current actual position of the tool;

[0115] Through the formula: Update axis displacement parameters;

[0116] in, This represents the axis displacement parameter after the k-th iteration, and the update termination condition is:

[0117] It should be noted that, This is the allowable error threshold.

[0118] Step 5: Use virtual simulation and finite element analysis to predict cutting forces, thermal deformation, and interference risks, and correct process parameters accordingly;

[0119] In this embodiment, the process of using virtual simulation and finite element analysis to predict cutting forces, thermal deformation, and interference risks, and to correct process parameters, includes:

[0120] Based on the workpiece CAD model, tool geometry parameters and material constitutive relations, a finite element multiphysics coupled simulation model including cutting force, heat conduction and contact behavior is constructed.

[0121] The virtual simulation platform is driven to perform dynamic simulation of the cutting process, and the cutting force time-domain waveform, thermal deformation gradient field and spatial distribution of tool-workpiece interference region are analyzed simultaneously.

[0122] The peak cutting force, extreme temperature, and minimum interference distance output by the simulation are compared with the preset process safety threshold to generate a risk level matrix and a list of key sensitive parameters.

[0123] Based on the risk matrix, the spindle speed curve, feed acceleration curve, and coolant injection strategy are optimized to generate a process parameter correction scheme under multi-objective constraints.

[0124] Step Six: Perform multi-axis linkage machining and monitor the load in real time. Detect lead, roughness, and tooth profile accuracy through online measurement and coordinate measuring machine.

[0125] Step 7: Optimize the path smoothness coefficient and genetic algorithm weights by combining processing data feedback, and update the process knowledge base.

[0126] Extract the lead deviation distribution, tooth profile error spectrum and path optimization parameters generated by the genetic algorithm in step six from the online measurement in step six, and construct a feature matrix including path smoothing coefficient, time-commutation weight ratio and process parameter coupling degree.

[0127] By analyzing the nonlinear relationship between lead error and λ and the interaction effect between tooth profile accuracy and time-commutation weight ratio through partial least squares regression, a correlation response surface model of error source-process parameters is established.

[0128] Based on the response surface model, a multi-objective particle swarm optimization algorithm is used to simultaneously optimize the smoothness coefficient of the quintic spline path and the time-commutation weight ratio of the genetic algorithm.

[0129] The optimized parameter combinations and corresponding machining quality data are encoded as rule entries and stored in a relational database according to the material type-tool combination-machine tool model triplet.

[0130] When the coefficient of variation (CV) of a new workpiece batch is greater than or equal to 5%, the system automatically matches the historical optimal parameter combination based on the updated knowledge base and triggers the re-optimization process in steps three through six.

[0131] This application presents a closed-loop optimization and real-time control process for multi-axis linkage lead screw precision milling. Based on inverse kinematics calculation and a pseudo-inverse iterative algorithm using the Jacobian matrix, a real-time mapping model of tool pose and axis displacement is constructed. Combined with a quintic spline interpolation algorithm, continuous and smooth multi-axis synchronous commands are generated, effectively reducing positioning errors on each axis. Simultaneously, the tool tilt angle is dynamically corrected through force feedback of the contact angle γ, effectively suppressing cutting force fluctuations and ensuring that the surface roughness of the helical groove meets requirements. In path planning, NURBS curve curvature adaptive discretization technology and quintic spline smoothing optimization are employed, along with a real-number encoded genetic algorithm to optimize the idle travel path, significantly improving machining efficiency and machine tool motion stability. In closed-loop optimization, a response surface model is constructed based on cutting simulation and online measurement data. A multi-objective particle swarm optimization algorithm is used to synchronously optimize the path smoothing coefficient, achieving high lead accuracy and low tooth profile error precision machining. Furthermore, relying on a process knowledge base, the system automatically matches historically optimal parameter combinations for new batches of workpieces, reducing process debugging time and comprehensively improving machining quality stability and production efficiency.

[0132] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A multi-axis linkage based precision ball screw milling closed loop optimization and real-time control process, characterized in that, The method comprises the following steps: Collecting workpiece geometry parameters, material properties and machine tool dynamics parameters, designing a tool and fixture scheme, and setting initial machining parameters based on a cutting database; Using NURBS curves to construct a high-precision curved surface model of the screw helical groove, and optimizing the path through curvature adaptive discretization and quintic spline interpolation; Using a real number coding genetic algorithm to iteratively optimize the idle stroke path and shorten the total path time and the number of reversals; Achieving multi-axis synchronous control through inverse kinematics calculation, and dynamically adjusting the tool tilt angle in real time based on the contact angle; Using virtual simulation and finite element analysis to predict the cutting force, thermal deformation and interference risk, and correcting the process parameters; Performing multi-axis linkage machining and real-time monitoring of the load, and through online measurement and three-coordinate machine detection of the lead, roughness and tooth profile precision; Combining the machining data feedback to optimize the path fairing coefficient and the genetic algorithm weight, and updating the process knowledge base; The specific process of using NURBS curves to construct a high-precision curved surface model of the screw helical groove, and optimizing the path through curvature adaptive discretization and quintic spline interpolation comprises: Based on the screw radius, lead and helix angle, generating a helical groove geometry model through the NURBS curve parameter equation, defining the control points, weights and node vectors to accurately represent the curved surface; Calculating the curvature of each point on the NURBS curve, dynamically adjusting the discretization step, and generating a high-density discrete point set to match the geometric features; Fitting the discrete points into quintic spline curves in segments, forcing the adjacent segments to meet the continuity of position, velocity, acceleration and jerk, and optimizing the objective to minimize the total acceleration variation of the path; Solving the spline coefficients and generating a smoothed G-code path, which includes coordinate, velocity and acceleration instructions; The curvature k(u) of the curve at the parameter u is calculated as: ； wherein denotes, and denotes the first and second derivative of the curve at u, respectively; The discretization step Δu is dynamically adjusted according to the curvature, and the step calculation formula is: ； wherein, is the maximum allowed step size, denotes the curve sensitivity coefficient, denotes the current discrete point parameter value; The detailed process of using a real number coding genetic algorithm to iteratively optimize the idle stroke path and shorten the total path time and the number of reversals comprises: Encoding the three-dimensional coordinate sequence of the idle stroke path into a real number vector, and randomly generating an initial population that meets the machine tool motion range constraints; Quantifying the path performance based on the weighted sum of the total path time and the number of reversals; Retaining high-quality individuals through tournament selection, introducing diversity through arithmetic crossover and Gaussian mutation, and correcting out-of-bound coordinates to ensure path feasibility; Stopping the optimization process when the number of iterations reaches the preset threshold or the optimal fitness is continuously stable; Decoding the real number sequence of the individual with the optimal fitness, generating a three-dimensional path, and transmitting it to the actuator.

2. A multi-axis linkage based precision ball screw milling closed loop optimization and real-time control process according to claim 1, wherein, The process of quantifying the path performance based on the weighted sum of the total path time and the number of reversals comprises: Constructing a fitness function based on the total path time and the number of reversals, and quantifying the idle stroke performance through weighted sum; The calculation formula of the fitness function is: ； Where F represents the weighted sum, T represents the total path time, and Ndir represents the number of reversals.

3. The multi-axis linkage based precision ball screw milling closed loop optimization and real-time control process according to claim 1, wherein, The specific process of achieving multi-axis synchronous control through inverse kinematics calculation, and dynamically adjusting the tool tilt angle in real time based on the contact angle comprises: Constructing a machine tool geometry kinematics model, defining the spatial mapping relationship and constraint conditions of the tool coordinate system and each motion axis; According to the target path point coordinates and the tool posture angle, the real-time displacement parameters of each motion axis are derived through inverse kinematics calculation; Based on the displacement parameters, multi-axis synchronous control instructions are generated, and the dynamic matching and coordination of the motion trajectories of each axis are realized by using interpolation algorithm The tool-workpiece contact angle is obtained in real time through force sensor or visual measurement device, and compared with the process allowable threshold to determine the inclination deviation If the contact angle exceeds the threshold, the tool inclination angle parameter is dynamically corrected and fed back to the inverse kinematics model.

4. The multi-axis linkage based precision ball screw closed-loop optimized and real-time controlled milling process according to claim 3, wherein, The process of deriving the real-time displacement parameters of each motion axis through inverse kinematics calculation includes: The target pose of the tool end in the base coordinate system given by the machining path planning; Through the machine tool geometric model, the differential relationship Jacobian matrix between tool pose change and axis displacement change is established; The axis displacement increment is calculated through the pseudo-inverse of Jacobian matrix; The calculation formula of displacement increment is: ； wherein , ; wherein denotes the incremental vector of the displacement of each axis, denotes the pseudo-inverse of the Jacobian matrix, a deviation vector of the target pose from the current actual pose, denotes the current actual pose of the tool By the formula: updating the axis displacement parameters; wherein, represents the axis displacement parameter after the kth iteration, and the update termination condition is: .

5. The multi-axis linkage based precision ball screw milling closed loop optimization and real-time control process according to claim 1, wherein, The process of predicting cutting force, thermal deformation and interference risk by using virtual simulation and finite element analysis, and correcting process parameters includes: Based on the workpiece CAD model, tool geometric parameters and material constitutive relationship, a finite element multi-physical field coupling simulation model is constructed, which includes cutting force, heat conduction and contact behavior; Drive the virtual simulation platform to perform dynamic simulation of the cutting process, and simultaneously analyze the time domain waveform of cutting force, thermal deformation gradient field and spatial distribution of tool-workpiece interference area; Compare the cutting force peak value, temperature extreme value and minimum interference distance output by the simulation with the preset process safety threshold to generate a risk level matrix and a list of key sensitive parameters; Based on the risk matrix, the spindle speed curve, feed acceleration curve and cooling liquid injection strategy are optimized, and a process parameter correction scheme under multi-objective constraints is generated.

6. A multi-axis linkage based ball screw precision milling closed loop optimization and real-time control process according to claim 1, wherein, The process of updating the process knowledge base by combining the feedback of machining data to optimize the path smoothing coefficient and the weight of genetic algorithm includes: Based on the lead deviation distribution, tooth profile error spectrum and path optimization parameters generated by genetic algorithm in step three, a feature matrix is constructed, which includes path smoothing coefficient, time-reversing weight ratio and process parameter coupling degree; Through partial least squares regression analysis, the nonlinear relationship between lead error and λ, and the interactive effect between tooth profile accuracy and time-reversing weight ratio are analyzed, and a correlation response surface model of error source-process parameters is established; Based on the response surface model, the smoothing coefficient of quintic spline path and the time-reversing weight ratio of genetic algorithm are simultaneously optimized by using multi-objective particle swarm optimization algorithm; The optimized parameter combination is encoded with the corresponding machining quality data as a rule entry, and is stored in a relational database according to the material type-tool combination-machine tool type ternary classification; When the new workpiece batch variation coefficient CV is greater than or equal to 5%, the historical optimal parameter combination is automatically matched based on the updated knowledge base, and the re-optimization process is triggered.

Images