Rapid tool changing method and system for automobile auxiliary frame machining center

Abstract

The invention relates to the technical field of numerical control machine tool control, and discloses a rapid tool changing method and system for an automobile auxiliary frame machining center, and the method comprises the steps: analyzing a machining program, building a tool multi-dimensional feature space and time sequence association weight matrix, and generating a tool magazine topology layout considering stroke efficiency and static load balance; the tool magazine is controlled to execute the excitation sequence to conduct in-situ kinetic parameter identification, and tool quality parameters are corrected; generating a full-path position-speed coupling control map defining a local speed upper limit in combination with the geometric curvature of the path and the dynamic constraint; the process remaining processing time is compared with the shortest motion time in real time, and an adaptive speed scaling factor is calculated to generate a dynamic speed instruction. The problems of load unbalance and turning impact during high-speed operation of the heavy tool magazine are solved, mechanical vibration and system energy consumption are remarkably reduced on the premise that the machining takt is guaranteed, and tool changing safety is improved.

Description

Technical Field

[0001] This invention relates to the field of CNC machine tool control technology, specifically to a rapid tool change method and system for automotive subframe machining centers. Background Technology

[0002] As a key load-bearing component of the chassis system, the automotive subframe involves multiple complex processes such as face milling, boring, drilling, and tapping, typically requiring dozens of tools of different specifications. To meet the cycle time requirements of large-scale automated production, modern automotive parts manufacturing lines widely adopt heavy-duty gantry machining centers equipped with large-capacity chain tool magazines. These tool magazines not only need to support various tools with a large total weight, but also need to achieve rapid and precise tool selection and tool changing between processes. Their operating efficiency and stability directly determine the overall machining auxiliary time (Chip-to-Chip Time) and equipment utilization rate.

[0003] In existing technologies, the control logic of chain tool magazines typically employs a shortest path priority combined with fixed speed programming. During the machining preparation stage, the operator loads tools into available tool positions sequentially or randomly according to the process card. When the CNC system issues a tool change command, the PLC controller calculates the shortest distance between the target tool position and the current tool change point, and then drives the servo motor to move according to a preset trapezoidal or S-shaped speed curve. During this process, the servo drive usually uses a semi-closed-loop PID control algorithm based on fixed gain parameters. This means that regardless of changes in the tool magazine load, it attempts to complete the positioning action at the set maximum speed and acceleration to ensure timely tool change response.

[0004] While existing technologies can meet basic tool changing requirements, significant dynamic and energy efficiency deficiencies remain under heavy-load conditions. First, current speed planning strategies fail to adequately consider the geometric nonlinear characteristics of chain-type structures. The chain-type tool magazine's trajectory includes a straight segment and a sprocket engagement zone, with a significant difference in curvature radii between the two. While a uniform maximum speed index provides high efficiency on the straight segment, heavy tools experience a sharp decrease in curvature when passing through sprocket corners, generating substantial centrifugal force and normal acceleration. This local dynamic shift often exceeds the mechanical structure's tolerance limits, causing abnormal loads on the tool chucks, resulting in slight loosening, or even tool drop or severe chain vibration. Second, traditional fixed-parameter control cannot adapt to the system's time-varying characteristics. The total moment of inertia and frictional torque of the tool magazine vary significantly with the number of loaded tools, weight distribution, and degree of mechanical wear. Fixed-gain controllers are prone to overshoot oscillations or tracking lag under heavy or unbalanced load conditions, affecting positioning accuracy. Finally, the existing scheduling logic lacks the ability to perceive the process timing. Even if the current machining operation takes a long time, the tool magazine still blindly executes the pre-selected action with the highest dynamic performance. This unnecessary abrupt acceleration and deceleration not only causes ineffective energy loss in the servo system, but also accelerates the fatigue failure of the reducer and chain link pins, failing to achieve a reasonable balance between equipment lifespan and operating efficiency. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a rapid tool change method and system for automotive subframe machining centers. It solves the problems of large mechanical vibration caused by load imbalance and sudden changes in path curvature during high-speed operation of existing heavy-duty chain tool magazines, low safety of heavy tool mounting, and energy efficiency mismatch caused by the lack of adaptive scheduling capability for process sequence.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] The first aspect of this invention provides a method and system for rapid tool changing in an automotive subframe machining center, comprising the following steps:

[0008] S1, analyze the machining program and tool database, establish a multi-dimensional feature space containing the physical properties of the task tools, and quantify the tool call order to construct a temporal correlation weight matrix;

[0009] S2, Based on the time-series correlation weight matrix and geometric interference conditions, generate a tool magazine pre-selected topology layout with the goal of minimizing tool change stroke and static load balancing;

[0010] S3 controls the tool magazine to execute multi-frequency excitation motion sequences, collects servo feedback signals to identify in-situ dynamic parameters, and calculates the system's equivalent moment of inertia and gravitational torque distribution characteristics.

[0011] S4, use the identification results to correct the multidimensional feature space, and perform global load verification based on the identified gravitational torque amplitude to calculate the global speed derating factor.

[0012] S5. Combining the corrected multidimensional feature space parameters, the global speed derating coefficient, and the local geometric curvature of the tool magazine path, a full-path position-speed coupling control map defining the local speed upper limit is generated.

[0013] S6. In real time, compare the remaining processing time of the current process with the shortest motion time planned based on the map, calculate the adaptive speed scaling factor, and generate dynamic speed commands to drive the tool magazine to run.

[0014] Preferably, the step S1 of establishing a multidimensional feature space containing the physical properties of the task tool specifically includes:

[0015] For each task tool, its geometric envelope dimensions are extracted, and a preset equivalent density coefficient and volume correction coefficient are matched according to the tool type. The initial estimate of the tool mass is calculated based on the cylindrical model, and a feature vector containing geometric attributes, the initial mass estimate, and parameter confidence labels is constructed.

[0016] The construction of the temporal correlation weight matrix in step S1 specifically includes: extracting the tool change instruction sequence in the machining program, counting the frequency of switching between any two tools in the process flow, and filling the corresponding row and column positions of the matrix with the frequency as the weight value to form a symmetric matrix that quantifies the physical adjacency requirements between tools.

[0017] Preferably, the step S2 of generating the tool magazine pre-selected topology layout specifically includes:

[0018] Establish geometric interference constraints for adjacent tool positions, requiring that the sum of the rotational envelope radii of two adjacent tools plus a preset safety margin be less than or equal to the physical tool position center distance;

[0019] Establish a full-cycle static gravity moment evaluation model, traverse all angular positions of the tool magazine rotation, and calculate the maximum absolute value of the gravity moment;

[0020] With minimizing the total cost of the weighted tool change stroke and the peak value of the static gravity torque calculated by the evaluation model as the comprehensive optimization objectives, a heuristic search algorithm is used to solve for the optimal arrangement sequence of tools on the physical tool positions.

[0021] Preferably, the in-situ dynamic parameter identification in step S3 specifically includes:

[0022] A generalized servo dynamic equilibrium equation describing the rotational motion of the tool magazine is established. This equation includes an equivalent rotational inertia term, a viscous friction term, and a periodic gravitational load torque term expanded using Fourier series.

[0023] Control the servo motor to perform a combined excitation motion that includes low-frequency large-amplitude scanning and high-frequency small-amplitude dithering;

[0024] An angular acceleration estimate is extracted from the position feedback signal using a state observer, and a discrete time series dataset is constructed by combining it with the current and torque signals.

[0025] The dataset was solved using the recursive least squares method to separate the actual total moment of inertia, friction coefficient, and gravitational torque distribution coefficient.

[0026] Preferably, the generation of the full-path position-velocity coupled control map in step S5 specifically includes:

[0027] The motion trajectory of the tool magazine is discretized into position units, and the path curvature radius corresponding to each position unit is extracted.

[0028] For each position unit, based on the nominal locking force of the tool magazine chuck and the corrected actual tool mass, the upper limit of the dynamic velocity constrained by centrifugal force is calculated, so that the centrifugal force generated by the tool is less than the ratio of the locking force to the safety factor.

[0029] For each position unit, based on the minimum distance between the tool tip scanning envelope and the stationary parts of the machine tool, the upper limit of geometric velocity constrained by spatial obstacle avoidance is calculated, so that the sliding distance of the tool after triggering an emergency stop is less than the minimum distance.

[0030] The smaller value between the upper limit of dynamic velocity and the upper limit of geometric velocity is selected as the original velocity limit value for this position unit.

[0031] Preferably, the step S5 of generating the full-path position-velocity coupled control map further includes:

[0032] In the tool magazine trajectory, a look-ahead transition zone is set in the area where the straight segment transitions to the arc segment. Within the look-ahead transition zone, the speed limit value is forced to decay according to a preset curve to smooth the speed boundary jump caused by the curvature change.

[0033] A kinematic inverse smoothing filter is applied to the original speed limit value sequence of the entire path. A reverse recursive scan is performed from the end point of the path to the starting point. The speed value of the previous position unit is limited according to the maximum braking angle acceleration of the servo motor, thereby generating a continuous and smooth position-velocity coupled control spectrum.

[0034] Preferably, step S4, which uses the identification result to correct the multidimensional feature space, specifically includes:

[0035] The total load inertia is obtained by calculating the difference between the actual total rotational inertia obtained from the identification and the inertia of the unloaded tool magazine.

[0036] The total load inertia is compared with the sum of the estimated inertia of all tools in the multidimensional feature space to obtain the inertia estimation error.

[0037] The inertia estimation error is allocated to the mass parameters of each tool according to the estimated mass weight of each tool, the data of the multidimensional feature space is updated, and the parameter confidence level is marked as verified.

[0038] Preferably, the global load verification in step S4 specifically includes:

[0039] The full-cycle gravitational moment curve is reconstructed based on the identified gravitational moment distribution coefficient, and the maximum amplitude of the gravitational moment is obtained.

[0040] If the maximum amplitude of the gravitational torque is within the preset ratio range of the rated torque of the servo motor, it is determined to be a heavy load condition. The global speed derating factor is calculated based on the difference between the peak torque of the motor and the maximum amplitude of the gravitational torque.

[0041] When generating the position-velocity coupled control map, the global velocity derating coefficient is used as the top-level constraint to uniformly scale the velocity upper limit of the entire path.

[0042] Preferably, the calculation of the adaptive speed scaling factor in step S6 follows the following logic:

[0043] Use virtual interpolation technology to estimate the remaining processing time before the end of the current process;

[0044] If the remaining processing time is greater than the sum of the shortest motion time and the reserved buffer time, then the adaptive speed scaling factor is set to the ratio of the shortest motion time to the effective remaining time, where the effective remaining time is the remaining processing time minus the reserved buffer time.

[0045] If the calculated scaling factor is less than the preset minimum sustaining threshold, the scaling factor will be clamped at the minimum sustaining threshold to maintain the low-speed steady-state response of the servo system.

[0046] A second aspect of the present invention provides a quick tool change system for an automotive subframe machining center, comprising:

[0047] The model building module is used to execute step S1, establishing a multi-dimensional feature space containing the physical properties of the task tool and a process temporal correlation weight matrix;

[0048] The topology optimization module is used to execute step S2 and generate a tool magazine pre-selected topology layout that balances stroke efficiency and static load balancing.

[0049] The parameter identification module is used to execute step S3, perform in-situ dynamic parameter identification, and calculate the system's equivalent moment of inertia and gravitational torque distribution characteristics.

[0050] The parameter correction and verification module is used to execute step S4, correct the multidimensional feature space using the identification results and calculate the global speed derating coefficient.

[0051] The map generation module is used to execute step S5, which generates a full path position-velocity coupling control map that defines the local velocity upper limit based on the corrected parameters, global velocity derating factor and path curvature.

[0052] The adaptive control module is used to execute step S6, calculate the speed scaling factor based on the comparison result of the remaining processing time and the shortest motion time, and generate dynamic speed commands in combination with the map.

[0053] This invention provides a method and system for rapid tool change in an automotive subframe machining center. It offers the following advantages:

[0054] 1. This invention constructs a full-path position-velocity coupled control graph, establishing a mapping relationship between dynamic constraints and path geometric curvature, enabling the system to automatically limit the upper limit of operating speed at high curvature positions such as the sprocket meshing area; combined with a tool magazine topology layout based on static load balancing, it effectively suppresses excessive centrifugal force and instantaneous mechanical impact generated by heavy tools when cornering at high speed, avoiding tool loosening or falling off accidents caused by insufficient chuck locking force.

[0055] 2. This invention introduces an in-situ dynamic parameter identification mechanism, which separates the actual rotational inertia, viscous friction coefficient, and gravity distribution characteristics of the system in real time through a multi-frequency excitation sequence, and uses the identification results to perform closed-loop correction of the tool mass parameters in the control model. This method eliminates the model mismatch problem caused by mechanical component wear, changes in lubrication status, or errors in tool parameter input, ensuring that the servo system can maintain optimal dynamic response characteristics under different load conditions.

[0056] 3. This invention utilizes an adaptive scheduling strategy based on process timing to compare the remaining processing time of the current process with the shortest movement time of the tool magazine in real time and calculate an adaptive speed scaling factor. In non-urgent working conditions with time redundancy, the system can actively reduce the running speed and acceleration of the tool magazine. Without extending the overall processing cycle time, it significantly reduces the high-frequency peak current consumption of the servo motor and the heat loss of the reduction mechanism, thus extending the service life of key mechanical components. Attached Figure Description

[0057] Figure 1 This is the overall flowchart of the tool magazine adaptive scheduling and control method based on multidimensional dynamic constraints of the present invention;

[0058] Figure 2 This is a flowchart illustrating the initial feature space and process timing model for this invention.

[0059] Figure 3 A flowchart for generating a preselected topology layout based on multidimensional constraints is provided for this invention.

[0060] Figure 4 This is a flowchart illustrating the in-situ dynamic parameter identification process of the present invention.

[0061] Figure 5 This is a flowchart illustrating the modification of characteristic parameters and verification of dynamic balance state in this invention.

[0062] Figure 6 This is a flowchart of the process for generating a smoothed sweep velocity coupling map according to the present invention;

[0063] Figure 7 This is a flowchart illustrating the adaptive control process for the operating conditions of this invention.

[0064] Figure 8 This is a comparison diagram of the tool magazine full-circumference speed limit map and the actual speed response curve in an embodiment of the present invention;

[0065] Figure 9 This is a comparison diagram of the servo motor output torque response under heavy load conditions in an embodiment of the present invention. Detailed Implementation

[0066] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0067] See attached document Figure 1 , Figure 1 This is a flowchart illustrating an adaptive scheduling and control method for tool magazines based on multidimensional dynamic constraints according to an embodiment of the present invention. This method is mainly applied to tool magazines in CNC machine tools equipped with servo drive systems, particularly chain-type or disc-type tool magazines. Figure 1 As shown, the method specifically includes the following steps:

[0068] S100 analyzes the machining program and tool database to establish a multi-dimensional feature model of the task tools. The system extracts the geometric dimensions, estimated quality, and installation attributes of the task tools, and constructs a feature vector containing physical attributes and status markers. At the same time, it counts the switching frequency between tools, quantifies the temporal correlation of tool calls, and generates a temporal correlation weight matrix.

[0069] S200 generates a pre-selected tool magazine topology layout based on temporal correlation weights and geometric interference conditions. With the optimization objectives of minimizing tool change travel and static load balancing, it calculates the optimal arrangement sequence of tools on the physical tool positions and outputs tool arrangement guidelines to guide tool loading, avoiding the risk of off-center loading caused by the concentration of static gravity torque.

[0070] The S300 executes a multi-band excitation motion sequence to identify in-situ dynamic parameters. After the tool is loaded, it controls the servo motor to execute a preset excitation trajectory and collects current, speed, and position feedback signals. Using the sensing data of the servo system, it calculates the equivalent moment of inertia, viscous friction coefficient, and gravitational torque distribution characteristics of the current system.

[0071] S400 corrects multi-dimensional feature space parameters and performs global load safety verification. It uses the identified actual physical parameters to correct the initially estimated tool mass data; it assesses load safety based on the identified gravitational torque amplitude. If the load exceeds the preset ratio of the servo motor's rated torque, it restricts operation or prompts for adjustment; if it is within the safe range, it calculates the global speed derating factor.

[0072] S500, combined with the corrected physical parameters and global derating coefficient, generates a full-path position-velocity coupling control graph. Taking into account the centrifugal force constraint of the tool at curves, spatial obstacle avoidance distance, and motor acceleration and deceleration performance, it calculates the local velocity upper limit of the tool magazine at each position point on the motion path, and constructs a velocity coupling graph describing the safe operating boundary.

[0073] The S600 compares the remaining processing time with the shortest motion time in real time and performs adaptive speed control. During processing, it predicts the remaining time window of the current process and matches it with the shortest motion time planned based on the speed map; accordingly, it dynamically adjusts the speed scaling factor to reduce mechanical load and energy consumption while meeting the processing cycle time.

[0074] See attached document Figure 2 , Figure 2 This is a flowchart illustrating the process of establishing an initial feature space and process timing model according to an embodiment of the present invention. The process aims to transform unstructured machining program code into a computer-processable digital model, specifically including the following sub-steps:

[0075] S101, the CNC system reads the CNC machining program of the workpiece to be machined and the machine tool's pre-loaded tool management database. The system identifies all the tools required in the machining program, forming a task tool set. For any tool in the task tool set... The system establishes a feature vector that includes geometric and physical properties and state markers. :

[0076] ;

[0077] In the formula, This is the unique logical index number of the tool within the system; The maximum rotational envelope radius of the tool is taken as the larger value between the tool cutting edge radius and the tool holder flange radius, and is used to establish the cylindrical safety envelope domain for subsequent interference detection. The total length of the tool refers to the farthest distance from the reference line of the tool holder's tapered surface to the tool tip. This is an initial estimate of the tool mass. Since it is difficult to weigh the tool directly before machining, the system automatically matches the corresponding equivalent density coefficient based on the tool type (such as face mill, drill, boring bar) recorded in the tool management database. And combined with the maximum profile diameter of the tool With length An equivalent cylindrical model is constructed to generate an approximate value. The calculation basis for this estimate is as follows:

[0078] ;

[0079] In the formula, A volume correction coefficient, preset according to the tool type, is set to a value range of 0.4 to 0.9 to correct volume deviations between the cylindrical model and the actual complex shape. For example, for solid drill bits, this coefficient is close to 0.8 to 0.9; for hollow pin shanks, this coefficient is 0.4 to 0.6. The confidence level of the parameter is initialized to 0, indicating that the current mass data is only an estimate and has not yet been verified by dynamics. Although this initial estimate has physical biases, it serves as the starting point for iterative optimization and will be corrected in subsequent steps using actual motion data.

[0080] In step S102, the system reads the CNC machining program (G-code) line by line and extracts the function instructions related to tool changing operations through syntax analysis. The system identifies the address characters used for pre-selecting tools (such as T-codes) and the auxiliary function codes that trigger the robot to perform tool changing actions (such as M06 codes). The system filters out cutting parameter instructions unrelated to tool changing and arranges all tool changing instructions in the order of program execution, generating a discrete tool call sequence. This sequence objectively reflects the order and frequency of use of different tools in the machining process.

[0081] S103, Based on the tool call sequence generated in step S102, the system calculates the correlation between any two tools in the process flow and constructs a temporal correlation weight matrix. This matrix is ​​a A symmetric matrix, where This represents the total number of tools used in the task. The construction process follows an accumulation principle: the system omits the tool call sequence, and for each adjacent tool switch operation (i.e., from tool) in the sequence... Switch to tool Find the row and column intersection element corresponding to the two knives in the matrix. and The values ​​of each element in the matrix are incremented by 1. The magnitude of the element values ​​directly quantifies the frequency of switching between the two corresponding tools during machining. For elements on the diagonal of the matrix, since they do not involve changes in physical position, the system forcibly sets their values ​​to 0. Through this step, the linear time logic in the machining program is transformed into weighted data that quantifies the physical adjacency requirements between tools, providing a mathematical basis for subsequent topology optimization aimed at minimizing travel. The conventional matrix initialization and traversal accumulation algorithm is a standard technique in the field of data processing, and its detailed pseudocode implementation will not be elaborated here.

[0082] See attached document Figure 3 , Figure 3 This is a flowchart illustrating the process of generating a pre-selected topology layout based on multi-dimensional constraints according to an embodiment of the present invention. The process primarily addresses how to determine the optimal sequence of storage positions within a limited physical tool space, based on the geometric and physical properties of the tool and its process relevance. Specifically, it includes the following sub-steps:

[0083] S201, the system first defines a mapping function. This function maps each logical index number in the task tool set to a physical tool position number in the tool magazine. To ensure physical storage feasibility, the system considers any two adjacent physical tool positions... and A geometric interference constraint is established. This constraint requires that the sum of the envelope radii of two adjacent tools plus a preset safety margin must be less than or equal to the physical center distance between these two tool positions. The constraint formula is as follows:

[0084] ;

[0085] In the formula, Indicates physical tool position The tool index stored above; The maximum rotational envelope radius of the tool; The minimum clearance reserved to compensate for manufacturing errors and thermal deformation is set to a range of 2mm to 10mm, depending on the size of the tool magazine. For a chain-type tool magazine, this parameter is the chain pitch; for a disc-type tool magazine, it is the chord length on the circumference of the tool holder distribution circle. When searching for layout schemes, the system automatically eliminates any solutions that do not satisfy this inequality, thus preventing the risk of rigid collisions between physical tools at the algorithmic level.

[0086] S202 addresses tool magazine systems employing horizontal axis rotation or asymmetrical path arrangements (such as racetrack-type chain structures), where the torque generated by tool gravity fluctuates significantly with the rotation angle. To prevent motor brake overload slippage or positional slippage after power failure due to severe imbalance of gravity loads on the left and right sides, the system establishes a full-cycle static gravity torque evaluation model. The system defines each physical tool position. vertical height It's about the rotation angle of the tool magazine motor. The function. The peak static moment index of the layout scheme. The calculation is as follows:

[0087] ;

[0088] In the formula, This represents the total number of tool positions in the tool magazine. To estimate the mass of the cutting tool; It is the acceleration due to gravity; Geometrically, this term is equivalent to the length of the lever arm of the gravity vector about the center of rotation. The model iterates through all angular positions of the tool magazine during one revolution, calculating the maximum absolute value of the gravitational torque. The system sets an upper limit for the allowable static torque. Its value range is set to 60% to 85% of the rated holding torque of the servo motor, and feasible solutions must meet this requirement. .

[0089] S203, based on the aforementioned geometric and static constraints, the system employs a combinatorial optimization algorithm to find the optimal layout. The system constructs a comprehensive objective function that includes tool change stroke efficiency and static load balancing. :

[0090] ;

[0091] In the formula, The temporal correlation weight matrix generated in the preceding step S103 characterizes the tool. and Switching frequency; The physical distance between the two tools in the current layout (expressed as the tool position interval); and This is a weighting coefficient, with a value range of 0.1 to 0.9, and it satisfies... It is used to balance the two optimization goals of fast tool change and stable load.

[0092] To solve the objective function, the system employs heuristic search algorithms such as genetic algorithms or particle swarm optimization. Taking genetic algorithms as an example, the system encodes the knife position arrangement sequence into chromosomes and iteratively evolves them using standard operators such as selection, crossover, and mutation until convergence. The goal is to minimize the number of knife positions. The specific implementation process of the genetic algorithm operators is a well-known technique in computational mathematics and will not be elaborated upon here.

[0093] In step S204, once the algorithm converges and obtains the optimal mapping relationship, the CNC system outputs a visual tool arrangement guide through the display terminal, clearly indicating the specific physical tool position number that each tool should be installed in. After confirming the physical loading is complete, the system generates a virtual mapping table in the controller's non-volatile memory. This table establishes a one-to-one correspondence between the physical tool position number, tool ID, and geometric physical attributes, serving as the data reference for subsequent execution of S300 in-situ dynamic parameter identification and S600 adaptive motion control.

[0094] See attached document Figure 4 , Figure 4 This is a flowchart illustrating the in-situ dynamic parameter identification process according to an embodiment of the present invention. The process primarily addresses how to accurately acquire the current physical load characteristics of the tool magazine using the sensing capabilities of the servo drive system itself, without relying on external sensors. Specifically, it includes the following sub-steps:

[0095] S301, the system establishes a dynamic model describing the rotational motion of the tool magazine. Considering the chain-type or non-circularly symmetrical tool magazine structure to which this invention is applicable, its gravitational load torque is not a simple sine function, but a nonlinear function that exhibits complex periodic variations with position. Therefore, the system establishes a dynamic equilibrium equation containing higher-order harmonic terms:

[0096] ;

[0097] In the formula, This is the electromagnetic torque output by the servo motor, which is calculated in real time from the torque current component of the driver. Let be the equivalent rotational inertia of the system to be identified; and These are angular acceleration and angular velocity, respectively. It is the coefficient of viscous friction; This is the Coulomb friction torque term, which is usually represented by a sign function model.

[0098] In particular, The position-dependent gravitational load torque. Considering the composite trajectory characteristics of the chain tool magazine, which includes both straight and circular segments, the system employs a generalized periodic function derived from Fourier series expansion for parametric modeling.

[0099] ;

[0100] In the formula, The harmonic cutoff order is set to a range of 3 to 7 to adequately cover the low-frequency torque fluctuation characteristics introduced by the non-circular trajectory. and The gravitational moment distribution coefficient is to be identified. The establishment of this model transforms the complex physical structural characteristics into a set of identifiable mathematical parameters.

[0101] S302, To ensure sufficient observability of the parameters to be identified in the above equations, the system controls the tool magazine servo motor to execute a preset combined excitation motion sequence. This sequence contains two motion components of different frequencies: the first stage is a low-frequency, large-amplitude scanning motion, with a frequency range set from 0.1Hz to 1.0Hz. The motor drives the tool magazine to rotate slowly past the peak point of gravitational potential energy with a low acceleration, mainly used to excite and identify the gravitational term coefficient. The second stage involves high-frequency, low-amplitude flutter motion, with a frequency range set from 5Hz to 20Hz. The motor rapidly switches between forward and reverse rotation within a localized range, generating significant acceleration and deceleration dynamic torque, primarily used to excite and identify the inertia term. .

[0102] In the S303, during the execution of the excitation motion, the CNC system acquires the current of the servo drive, the position encoder value, and the speed feedback signal in real time via a high-speed fieldbus. Since directly calculating the acceleration from the second-order difference of the position signal introduces significant quantization noise, causing the identification algorithm to diverge, the system incorporates a state observer technique. The specific construction of the state observer (such as a Kalman filter or a Romberg observer) can be directly implemented by those skilled in the art based on control theory textbooks; it is a well-known technique in the field and will not be elaborated upon here. The system uses this observer to output high signal-to-noise ratio angular acceleration estimates in real time based on position and torque commands. Thus obtaining the contents Synchronous discrete time series dataset.

[0103] S304, the system rewrites the dynamic equation in step S301 into a linear regression form. .in, Corresponding electromagnetic torque A data vector containing trigonometric function terms for acceleration, velocity, and position. To include all physical parameters to be identified The system uses a recursive least squares method to iteratively calculate the collected dataset and obtain the optimal estimate of the parameter vector. Finally, the system separates the actual total moment of inertia. The gravitational moment curve for the entire period is reconstructed based on the identified gravitational moment coefficient, and its maximum amplitude is calculated. This serves as the physical benchmark for subsequent security verification.

[0104] See attached document Figure 5 , Figure 5 This is a flowchart illustrating the process of correcting characteristic parameters and verifying dynamic balance according to an embodiment of the present invention. The process mainly utilizes the real physical parameters identified in the preceding steps to perform closed-loop correction on the system's internal prediction model and establish a global safe operating benchmark. Specifically, it includes the following sub-steps:

[0105] S401, the system utilizes the actual total moment of inertia identified in step S304. The system then performs a posteriori correction on the feature vectors based on geometric volume estimation in step S101. First, the system calculates the sum of the estimated total inertia of all current tools and the inertia of the idle tool magazine, and then compares this sum with the identified actual total inertia. The inertia estimation error is obtained through comparison. This error is not uniformly distributed, but rather weighted according to the mass and installation radius of each tool. (For confidence level marking...) For tools with a value of 0, the system sets its quality parameter to 0. Update according to the error distribution ratio and label the confidence level. Set to 1. Although this step cannot precisely measure the absolute mass of each tool, it statistically eliminates the cumulative deviation of the total system load, ensuring the accuracy of the inertia parameter in the subsequent control model.

[0106] S402, the system directly calls the maximum amplitude of the gravitational torque identified in step S304. This is used to assess the safety of the current tool arrangement in a gravitational field. The system sets two key safety thresholds: the rated holding threshold... (Corresponding to the rated output torque of the motor) and peak overload threshold (Corresponding to the short-time peak torque of the motor). The verification logic is as follows:

[0107] when Exceed When the load is between 85% and 95%, the system determines that the current layout has a serious risk of off-center load, which may cause the motor to overheat or the brake to slip. At this time, the system immediately issues an alarm signal, prohibits the tool magazine from starting, and highlights the location of the heavy tool that is recommended to be removed on the human-machine interface.

[0108] when In When the load is between 60% and 85%, the system determines that the current layout is under heavy load. Although it has not exceeded the physical limit, in order to extend the mechanical life, the system automatically switches the operating mode to derating mode.

[0109] when Below When the speed reaches 60%, the system determines it to be in standard operating condition and allows full-speed operation.

[0110] S403, for operating conditions determined to be in derating mode, the system calculates a global speed derating factor. ( The physical meaning of this coefficient is that, within the motor's limited output torque capacity, after deducting the torque required to overcome gravitational loads and frictional resistance, the remaining torque determines the maximum allowable dynamic response capability of the system. Global speed derating factor. The calculation formula is as follows:

[0111] ;

[0112] In the formula, The average frictional torque to be identified; The system is designed with a nominal maximum angular acceleration; The safety margin factor is set to a range of 0.7 to 0.95. This formula, based on Newton's second law, ensures that the electromagnetic torque demand of the servo motor is always limited to the peak torque during maximum acceleration startup under the most severe gravity load conditions. Within the safe envelope. Calculated This will serve as a top-level constraint, directly impacting the velocity map generation process in subsequent steps S500, thus preventing servo overload alarms caused by blindly pursuing speed from the outset.

[0113] See attached document Figure 6 , Figure 6 This is a schematic flowchart illustrating the generation of a smoothed sweep velocity coupling map according to an embodiment of the present invention. This process primarily addresses how to dynamically limit the speed based on local geometric features during the variable-speed motion of a chain-type or non-circular tool magazine to prevent mechanical impact and tool slippage. Specifically, it includes the following sub-steps:

[0114] S501, the system first discretizes the entire motion trajectory of the tool magazine (0 to 360 degrees or the corresponding chain length range) into tiny position units. The step size is set to 0.1 mm to 2.0 mm (for straight segments) or 0.1 degrees to 1.0 degrees (for circular segments). For each discrete position... The system extracts the corresponding path curvature radius based on the mechanical design drawings of the tool magazine. For a chain-type tool magazine, its trajectory consists of straight line segments and circular arc segments: in the straight line segments, In the arc segment where the sprockets mesh, It is approximately equal to the pitch circle radius of the sprocket.

[0115] Based on the position-dependent radius of curvature, the system calculates the upper limit of dynamic velocity constrained by the tool clamping force. And limited by the geometric velocity limit for spatial obstacle avoidance

[0116] Regarding the upper limit of dynamic speed Its core purpose is to prevent the heavy-duty cutting tool from being thrown out due to excessive centrifugal force in the arc segment, overcoming the spring locking force. The calculation formula is as follows.

[0117] ;

[0118] In the formula, The nominal spring locking force of the tool holder chuck; The current maximum tool mass (or the actual mass of the tool at the current position) after identification and correction in step S401. To mitigate the risk of spring detachment, a safety factor ranging from 1.5 to 3.0 is set to account for uncertainties related to spring fatigue decay. In the straight section, due to... The upper limit of this speed approaches infinity mathematically, and is limited by the maximum speed of the motor.

[0119] Regarding the geometric velocity limit The system monitors in real time the minimum Euclidean distance between the scanning envelope of the tool tip and stationary parts of the machine tool (such as the column and sheet metal). The calculation formula is as follows:

[0120] ;

[0121] In the formula, This parameter, representing the shortest braking time required for the servo motor to decelerate to zero from its current rated speed to trigger an emergency stop, is determined by the motor's torque-speed characteristic curve and the system's total inertia. This formula ensures that the tool will not collide due to inertial slippage when an emergency stop is triggered at any position.

[0122] S502, the system uses the global speed derating factor calculated in step S403. Introduce and synthesize the original velocity limit value at each position. :

[0123] ;

[0124] For the special working condition of a chain-type tool magazine transitioning from a straight section to a circular arc, due to the abrupt change in the radius of curvature (from... Mutation ),lead to A step drop occurs. To prevent the servo motor from generating excessive jerk at the corner entry point, the system sets a look-ahead transition zone in the area between the straight section and the curved section. The length of this zone is set to 0.5 to 1.5 times the chain pitch. Within the transition zone, the system forces the high-speed limit value of the straight section to decay linearly or in an S-shaped curve to the low-speed limit value of the curved section, thus smoothing the abrupt change in speed boundaries.

[0125] S503, the above-generated While the curve numerically meets safety requirements, it may contain high-frequency fluctuations that cannot be executed by a physical motor. The system... The sequence is applied using a kinematic inverse smoothing filter algorithm. This algorithm performs a backward recursive scan from the end of the path to the beginning, based on the maximum braking angular acceleration of the motor. Limit the velocity value at the previous point. Smoothed velocity profile. satisfy:

[0126] ;

[0127] The distance from the walking distance is determined. This processing ensures that the velocity descent slope at any position in the graph is within the range of the motor's active braking capability. The final generated... It is stored as a lookup table, serving as the benchmark data source for dynamic amplitude limiting in subsequent real-time control.

[0128] See attached document Figure 7 , Figure 7 This is a flowchart illustrating the implementation of adaptive control based on operating conditions according to an embodiment of the present invention. The process aims to solve the dynamic matching problem between the tool magazine movement and the machine tool's machining cycle time, reducing the average load rate of the mechanical system through flexible scheduling in the time domain. Specifically, it includes the following sub-steps:

[0129] In S601, during the execution of the current machining operation by the CNC machine tool, the control system uses the look-ahead control module to pre-read multiple lines of machining program code in the instruction buffer. The system parses the next tool number to be called and locks the physical target position of the tool according to the virtual mapping table generated in the previous steps. Simultaneously, the system calculates the remaining machining time before the end of the current operation. Regarding the calculation of the remaining time, the system does not simply rely on code line count statistics, but uses a time simulation technique based on virtual interpolation: the system simulates an interpolator to preprocess all instruction segments in the buffer, accumulating the planned motion time, spindle pause time, and auxiliary function execution time of each segment. For instruction blocks containing fixed loops or subroutines, the system calls a preset standard cycle database to obtain their estimated time. This calculation process belongs to the existing conventional techniques in the field of CNC system trajectory planning and can obtain the remaining machining time with high accuracy. .

[0130] S602, based on the position-velocity coupling map generated in S503, the system plans the motion path from the current tool magazine position to the target position. To ensure optimal efficiency, the system calculates the time cost required to rotate to the target position clockwise and counterclockwise respectively. During the calculation, the system uses the smoothed velocity map as a constraint condition for the path integral. Shortest motion time. The calculation formula is as follows:

[0131] ;

[0132] In the formula, This represents the upper limit of the smoothed velocity at the corresponding discrete angular position; The position step size for discretization; and These are the start-up acceleration time and stop-deceleration time, obtained by looking up a table based on the servo motor's torque-speed characteristic curve. In the microprocessor implementation, this calculation is performed using a discretized numerical accumulation method, that is, summing the fastest allowable time slices corresponding to all differential angle units along the path. The system compares the two directions... The value determines the direction with the shorter execution time as the final execution path.

[0133] S603, the system will use the remaining processing time obtained in step S601. The shortest motion time calculated in step S602 Perform comparisons and calculate the adaptive speed scaling factor. The physical meaning of this factor is to stretch the tool magazine's motion cycle on the time axis to match the machine tool's machining cycle. Scaling Factor The calculation logic is as follows:

[0134] ;

[0135] In the formula, The buffer time constant reserved for the system is usually set to a range of 0.2 seconds to 1.5 seconds. It is used to cover the communication handshake delay, the mechanical positioning pin action time and the signal stabilization time, and to prevent tool change timeout due to small errors in time prediction.

[0136] When the calculated scaling factor is less than the preset minimum maintenance threshold Its value range is set to 0.1 to 0.4, and the system clamps its value to [a specific value]. This is to maintain the steady-state response performance of the servo system at low speeds and prevent crawling caused by excessively low speeds.

[0137] S604, based on a defined scaling factor and velocity coupling map, generates real-time dynamic speed commands from the servo driver. This command is dynamically adjusted as the position changes during movement, and its calculation relationship is as follows:

[0138] ;

[0139] In the specific implementation of the servo control loop, the aforementioned dynamic speed command is input to the speed loop controller as the dynamic saturation threshold of the output limiting module. The control system monitors the current position feedback signal in real time and outputs the corresponding limiting value according to the lookup table. If the spindle ratio on the machine tool side is manually changed during the tool magazine rotation, resulting in reduced remaining machining time... If the scaling factor is suddenly shortened, the system will immediately re-execute steps S603 and S604 via an interrupt mechanism to increase the scaling factor. This involves increasing the tool magazine speed to ensure tool change preparation is completed before the new time window closes. Through this step, the system achieves a dynamic balance between mechanical load and machining efficiency while ensuring physical safety.

[0140] In this embodiment, a heavy-duty gantry machining center was selected as the verification platform. The machine tool is equipped with a chain-type tool magazine with a capacity of 60 tools. The total length of the chain is approximately 5.8 meters, and it is driven by a servo motor with a rated power of 2.5kW and a rated torque of 12Nm (peak 30Nm).

[0141] 1. Initial state and parameter configuration

[0142] The machining task is milling and drilling of battery trays for new energy vehicles. The process involves 3 types of heavy-duty tools (D125 face milling cutter, 8.2kg each), 12 types of medium-duty tools (D30 reamer, 3.5kg each), and 20 types of light-duty tools (taps and drills, <1.0kg each).

[0143] When executing step S100, the system reads the tool parameters and identifies that if the three heavy face milling cutters are loaded in the default order, they will be arranged in a continuous manner, resulting in local mass concentration of the chain. The estimated static unbalanced torque is as high as 11.5 Nm, which is close to 95% of the rated torque of the motor.

[0144] 2. Topology optimization and parameter identification

[0145] The system executes step S200, replanning the tool positions using a genetic algorithm. The optimized scheme places three heavy-duty end mills at tool positions 1, 21, and 41 respectively (120 degrees apart), reducing the static unbalanced torque to 2.3 Nm.

[0146] Then, step S300 is executed for in-situ identification. The system detects that due to long-term use of the chain, the average viscous friction coefficient has increased. This is approximately 25% higher than the factory-specified value. The identified total system moment of inertia... for (Converted to the motor shaft).

[0147] 3. Velocity map generation and control

[0148] Based on the analysis model generated by MATLAB (corresponding to step S500), the system constructed a position-velocity coupled spectrum.

[0149] 4. Adaptive processing

[0150] During step S600, the machine tool is performing a 4-minute planar milling operation. The next tool is a drill bit, located 15 tool positions away from the current position.

[0151] The system calculated the shortest physical motion time to be 5.2 seconds. Since the remaining processing time is much longer than 5.2 seconds, the system automatically adjusts the speed scaling factor. The value is set to 0.15. The tool magazine glides silently to the target position at an extremely low speed (approximately 5 m / min) and under extremely low current load, avoiding unnecessary high-speed start-stop shocks.

[0152] Experimental verification and effect comparison

[0153] To verify the effectiveness of this solution, comparative tests were conducted on the aforementioned platforms.

[0154] Test Groups:

[0155] Control group: The traditional trapezoidal velocity planning was used, with the global maximum velocity set at 30 m / min and the acceleration set to a fixed value.

[0156] Experimental group: The adaptive control scheme based on dynamic constraints described in this invention is adopted.

[0157] Test conditions: Continuously execute a mixed tool change cycle of heavy tool-light tool-heavy tool 100 times, and use a high-frequency data acquisition card to record the current of the servo driver, the position error, and the acceleration sensor data installed on the tool magazine bracket.

[0158] The specific test data is shown in the table below:

[0159] Table 1. Test data record of tool magazine operation dynamics and energy efficiency comparison.

[0160] Key Performance Indicators Unit Control group (traditional PID + fixed planning) Experimental group (adaptive control of the present application) Optimization amplitude / variation Peak impact torque Nm 29.8 (close to saturation) 18.4 -38.2% Torque ripple coefficient (Ripple) % 14.5% 3.2% -77.9% Chain structure vibration peak g 0.82 0.21 -74.4% Average positioning steady-state error mm 0.45 0.08 -82.2% Single cycle average energy consumption J 412.5 285.3 -30.8% Heavy tool radial displacement mm 0.15 <0.02 Significant improvement Average tool change action time consumption s 4.2 5.8 +38.1% (does not affect the beat) Overshoot times (5%) Times 12 0 -

[0161] According to Table 1 and appendix Figure 8 Appendix Figure 9 Based on the data analysis, this technical solution draws the following conclusions:

[0162] First, the dynamic stability was significantly improved. The peak impact torque of the experimental group decreased by 38.2%, and the peak structural vibration decreased by 74.4%. This is directly attributed to the smoothed velocity profile generated in step S500, especially the local deceleration strategy at the sprocket rotation angle (e.g., Figure 8 As shown in the figure, the jet impact caused by the abrupt change in curvature was eliminated. The control group, lacking awareness of its geometric position, maintained high speeds while cornering, resulting in extreme mechanical stress.

[0163] Second, control accuracy and safety are enhanced. The steady-state error has been reduced from 0.45mm to 0.08mm, and the radial displacement (risk of loosening) of the heavy tool during movement has been almost eliminated. This is because the parameter identification and correction in step S300 corrects the system model, making the feedforward control more precise; simultaneously, based on the clamping force constraint in step S501... This ensures that the cutting tool always operates within the physical safety boundaries.

[0164] Third, energy efficiency optimization and lifespan extension. Although the average action time of the experimental group increased (from 4.2s to 5.8s), this increase came from the adaptive scheduling of step S600, which actively reduced speed by utilizing the masking time during processing. As a result, energy consumption was reduced by 30.8%, and the reduction in torque ripple meant a significant reduction in heat loss and wear of the motor and reducer.

[0165] See attached document Figure 8 , Figure 8 This is a comparison chart of the tool magazine full-circumference speed limit map generated according to an embodiment of the present invention and the actual speed response curve.

[0166] The horizontal axis in the figure represents the physical tool position number of the tool magazine (1 to 60), and the vertical axis represents the allowable chain running speed (m / min). The dashed line in the figure represents the constant speed limiting strategy commonly used in the prior art, with a set value of 30m / min that does not change with position.

[0167] The solid line represents the smoothed sweep velocity coupling map generated in step S500 in this embodiment of the invention. .

[0168] The trend of the curve in the graph shows that:

[0169] In the straight segment region (e.g., the tool position range of 5-15, 25-35), due to the radius of curvature... Approaching infinity, limited by the motor's maximum speed, the system automatically increases the speed limit to 35m / min, utilizing the physical advantages of straight segments to improve efficiency.

[0170] In the sprocket meshing area (i.e., the curved area, such as the tool positions 18-22 and 38-42), due to centrifugal force constraint... It plays a dominant role, and the curve exhibits a significant concave characteristic, with the local velocity limited to between 12 m / min and 15 m / min.

[0171] The curve shape indicates that this scheme changes the traditional one-size-fits-all conservative speed limit mode, allowing for faster speeds on straight sections where physical conditions permit and slower speeds on higher-risk curves, thus achieving a fine-grained allocation of speed in the spatial dimension.

[0172] See attached document Figure 9 , Figure 9 This is a comparison diagram of the output torque response of a servo motor under heavy load conditions according to an embodiment of the present invention.

[0173] This test simulates the extreme condition of continuous switching between heavy-blade, light-blade, and heavy-blade operation. The horizontal axis represents time, and the vertical axis represents the output torque of the servo motor.

[0174] The dashed line in the figure represents the torque response of the existing technology. It can be seen that during the acceleration phase (0.5 to 1.5 seconds) and the cornering phase (2.5 to 3.5 seconds), the torque waveform exhibits severe oscillations, with the peak value repeatedly approaching the physical limit of 30 Nm. This high-frequency oscillation reflects the conflict between insufficient rigidity of the mechanical system and overly harsh control commands.

[0175] The solid line represents the torque response of this embodiment of the invention. This is achieved thanks to the adaptive speed scaling factor in step S600. With the adjustment of the torque curve and the accurate inertia parameters identified in step S300, the torque curve planned by the system is smooth and continuous. Even at the maximum load, the peak torque is controlled at around 20 Nm, with no obvious overshoot.

[0176] The comparative data confirms that this solution effectively reduces the dynamic peak load of the mechanical system and lowers the risk of motor overheating and mechanical component fatigue failure through flexible scheduling in the time domain (i.e., actively reducing speed during non-emergency times).

[0177] In summary, this invention solves the vibration and safety hazards of heavy-duty chain tool magazines during high-speed operation without sacrificing the overall machining efficiency of the machine tool by using multidimensional dynamic constraints and adaptive timing scheduling, and achieves peak shaving and valley filling of mechanical load.

[0178] 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 rapid tool change method for an automotive subframe machining center, characterized in that, Includes the following steps: S1, analyze the machining program and tool database, establish a multi-dimensional feature space containing the physical properties of the task tools, and quantify the tool call order to construct a temporal correlation weight matrix; S2, Based on the time-series correlation weight matrix and geometric interference conditions, generate a tool magazine pre-selected topology layout with the goal of minimizing tool change stroke and static load balancing; S3 controls the tool magazine to execute multi-frequency excitation motion sequences, collects servo feedback signals to identify in-situ dynamic parameters, and calculates the system's equivalent moment of inertia and gravitational torque distribution characteristics. S4, use the identification results to correct the multidimensional feature space, and perform global load verification based on the identified gravitational torque amplitude to calculate the global speed derating factor. S5. Combining the corrected multidimensional feature space parameters, the global speed derating coefficient, and the local geometric curvature of the tool magazine path, a full-path position-speed coupling control map defining the local speed upper limit is generated. S6. In real time, compare the remaining processing time of the current process with the shortest motion time planned based on the map, calculate the adaptive speed scaling factor, and generate dynamic speed commands to drive the tool magazine to run.

2. The method according to claim 1, characterized in that, Step S1 involves establishing a multi-dimensional feature space that includes the physical properties of the task tool, specifically including: For each task tool, its geometric envelope dimensions are extracted, and a preset equivalent density coefficient and volume correction coefficient are matched according to the tool type. The initial estimate of the tool mass is calculated based on the cylindrical model, and a feature vector containing geometric attributes, the initial mass estimate, and parameter confidence labels is constructed. The construction of the temporal correlation weight matrix in step S1 specifically includes: extracting the tool change instruction sequence in the machining program, counting the frequency of switching between any two tools in the process flow, and filling the corresponding row and column positions of the matrix with the frequency as the weight value to form a symmetric matrix that quantifies the physical adjacency requirements between tools.

3. The method according to claim 1, characterized in that, The step S2 of generating the tool magazine pre-selected topology layout specifically includes: Establish geometric interference constraints for adjacent tool positions, requiring that the sum of the rotational envelope radii of two adjacent tools plus a preset safety margin be less than or equal to the physical tool position center distance; Establish a full-cycle static gravity moment evaluation model, traverse all angular positions of the tool magazine rotation, and calculate the maximum absolute value of the gravity moment; With minimizing the total cost of the weighted tool change stroke and the peak value of the static gravity torque calculated by the evaluation model as the comprehensive optimization objectives, a heuristic search algorithm is used to solve for the optimal arrangement sequence of tools on the physical tool positions.

4. The method according to claim 1, characterized in that, The in-situ dynamic parameter identification in step S3 specifically includes: A generalized servo dynamic equilibrium equation describing the rotational motion of the tool magazine is established. This equation includes an equivalent rotational inertia term, a viscous friction term, and a periodic gravitational load torque term expanded using Fourier series. Control the servo motor to perform a combined excitation motion that includes low-frequency large-amplitude scanning and high-frequency small-amplitude dithering; An angular acceleration estimate is extracted from the position feedback signal using a state observer, and a discrete time series dataset is constructed by combining it with the current and torque signals. The dataset was solved using the recursive least squares method to separate the actual total moment of inertia, friction coefficient, and gravitational torque distribution coefficient.

5. The method according to claim 1, characterized in that, Step S5, which generates a full-path position-velocity coupled control map, specifically includes: The motion trajectory of the tool magazine is discretized into position units, and the path curvature radius corresponding to each position unit is extracted. For each position unit, based on the nominal locking force of the tool magazine chuck and the corrected actual tool mass, the upper limit of the dynamic velocity constrained by centrifugal force is calculated, so that the centrifugal force generated by the tool is less than the ratio of the locking force to the safety factor. For each position unit, based on the minimum distance between the tool tip scanning envelope and the stationary parts of the machine tool, the upper limit of geometric velocity constrained by spatial obstacle avoidance is calculated, so that the sliding distance of the tool after triggering an emergency stop is less than the minimum distance; The smaller value between the upper limit of dynamic velocity and the upper limit of geometric velocity is selected as the original velocity limit value for this position unit.

6. The method according to claim 5, characterized in that, The step S5 of generating the full-path position-velocity coupled control map also includes: In the tool magazine trajectory, a look-ahead transition zone is set in the area where the straight segment transitions to the arc segment. Within the look-ahead transition zone, the speed limit value is forced to decay according to a preset curve to smooth the speed boundary jump caused by the curvature change. A kinematic inverse smoothing filter is applied to the original speed limit value sequence of the entire path. A reverse recursive scan is performed from the end point of the path to the starting point. The speed value of the previous position unit is limited according to the maximum braking angle acceleration of the servo motor, thereby generating a continuous and smooth position-velocity coupled control spectrum.

7. The method according to claim 1, characterized in that, Step S4, which uses the identification results to correct the multidimensional feature space, specifically includes: The total load inertia is obtained by calculating the difference between the actual total rotational inertia obtained from the identification and the inertia of the unloaded tool magazine. The total load inertia is compared with the sum of the estimated inertia of all tools in the multidimensional feature space to obtain the inertia estimation error. The inertia estimation error is allocated to the mass parameters of each tool according to the estimated mass weight of each tool, the data of the multidimensional feature space is updated, and the parameter confidence level is marked as verified.

8. The method according to claim 1, characterized in that, The global load verification in step S4 specifically includes: The full-cycle gravitational moment curve is reconstructed based on the identified gravitational moment distribution coefficient, and the maximum amplitude of the gravitational moment is obtained. If the maximum amplitude of the gravitational torque is within the preset ratio range of the rated torque of the servo motor, it is determined to be a heavy-load condition. The global speed derating factor is calculated based on the difference between the peak torque of the motor and the maximum amplitude of the gravitational torque. When generating the position-velocity coupled control map, the global velocity derating coefficient is used as the top-level constraint to uniformly scale the velocity upper limit of the entire path.

9. The method according to claim 1, characterized in that, The calculation of the adaptive speed scaling factor in step S6 follows the logic below: Use virtual interpolation technology to estimate the remaining processing time before the end of the current process; If the remaining processing time is greater than the sum of the shortest movement time and the reserved buffer time, then the adaptive speed scaling factor is set to the ratio of the shortest movement time to the effective remaining time, where the effective remaining time is the remaining processing time minus the reserved buffer time. If the calculated scaling factor is less than the preset minimum sustaining threshold, the scaling factor will be clamped at the minimum sustaining threshold to maintain the low-speed steady-state response of the servo system.

10. A quick tool change system for an automotive subframe machining center, characterized in that, include: The model building module is used to execute step S1, establishing a multi-dimensional feature space containing the physical properties of the task tool and a process temporal correlation weight matrix; The topology optimization module is used to execute step S2 and generate a tool magazine pre-selected topology layout that balances stroke efficiency and static load balancing. The parameter identification module is used to execute step S3, perform in-situ dynamic parameter identification, and calculate the system's equivalent moment of inertia and gravitational torque distribution characteristics. The parameter correction and verification module is used to execute step S4, correct the multidimensional feature space using the identification results and calculate the global speed derating coefficient. The map generation module is used to execute step S5, which generates a full path position-velocity coupling control map that defines the local velocity upper limit based on the corrected parameters, global velocity derating factor and path curvature. The adaptive control module is used to execute step S6, calculate the speed scaling factor based on the comparison result of the remaining processing time and the shortest motion time, and generate dynamic speed commands in combination with the map.

Images