Automatic cutting processing system based on aluminum alloy production

Abstract

This invention discloses an automated cutting system based on aluminum alloy production, belonging to the field of aluminum alloy profile processing technology. The system includes a mechanical execution layer, a sensing and detection layer, and an intelligent control layer. Addressing the problems of lag in feed control and lack of mechanical basis for attitude selection in cutting complex cross-section profiles, this invention uses a multi-axis attitude optimization control unit to calculate the relationship between the profile cross-section moment of inertia and cutting force, selecting the optimal attitude with minimal deformation. A collaborative controller drives a cross-section topology analysis unit to generate a cross-section feature sequence based on the optimal attitude. A feed rate adaptive control unit establishes a forward-looking window based on this sequence, predictively adjusting the rate before the saw blade contacts the wall surface or cavity features. This invention eliminates response lag through feedforward control and reduces profile deformation through mechanical optimization, significantly improving the cutting quality and accuracy of aluminum profiles.

Description

Technical Field

[0001] This invention relates to the field of aluminum alloy profile processing technology, and in particular to an automatic cutting and processing system based on aluminum alloy production. Background Technology

[0002] Aluminum alloy profiles used in building curtain walls and window systems typically have complex multi-cavity cross-sectional structures, containing multiple hollow areas, walls of varying thicknesses, and reinforcing ribs. Existing automated cutting systems for this type of profile often employ a fixed saw blade feed rate or simply segment the profile based on its overall width. However, as the saw blade passes through solid walls and hollow areas sequentially during its cutting stroke, the cutting resistance undergoes a non-linear abrupt change. For example, the instantaneous disappearance of cutting force when cutting from the wall into the cavity causes saw blade oscillation, while the re-contact with the wall from the cavity generates impact loads. Current technologies primarily rely on feedback regulation based on motor current detection, but this feedback mechanism inherently suffers from hysteresis. By the time the system senses load changes and responds, defects such as burrs and tears caused by impact or vibration have often already occurred, severely impacting the yield of anodized or powder-coated profiles.

[0003] Furthermore, while high-end machining centers equipped with multi-axis sawing stations possess the capability to complete the same cutting task using various posture combinations (such as tilting saw blades or rotating profiles), they often lack mechanically-based optimization in actual path planning. Aluminum profile cross-sections typically exhibit asymmetry, with significant differences in bending stiffness along different directions. If the principal direction of the cutting force coincides with the weak axis of the profile cross-section, it can easily lead to elastic deformation or end warping during processing, thus affecting the dimensional accuracy of compound angle cutting. Current control systems typically treat feed control and posture planning as independent modules, lacking coordinated optimization between the two, making it difficult to meet the processing requirements of high-precision aluminum profiles. Summary of the Invention

[0004] The main objective of this invention is to provide an automatic cutting and processing system based on aluminum alloy production. By combining cross-sectional topological feature prediction and attitude mechanics optimization, this system solves the processing quality problems caused by feed control lag and improper cutting attitude selection in the prior art.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] An automated cutting and processing system based on aluminum alloy production includes a mechanical execution layer, a sensing and detection layer, and an intelligent control layer;

[0007] The mechanical execution layer includes an adaptive clamping module for clamping and positioning the profile, a multi-degree-of-freedom gripper module for transporting the profile and adjusting its spatial orientation, and a multi-axis sawing module for performing cutting actions.

[0008] The sensing and detection layer includes an encoder group for acquiring motion axis position information and a current sensor for acquiring motor load signals.

[0009] The intelligent control layer includes a cross-section topology analysis unit, a cutting force vector analysis unit, a multi-axis attitude optimization control unit, a feed rate adaptive control unit, and a cooperative controller;

[0010] The cutting force vector analysis unit is configured to: calculate the direction and relative magnitude of the cutting force in the profile section coordinate system based on the rotation direction, cutting direction and current posture parameters of the saw blade, combined with the material properties and cutting process parameters of the profile, and output the cutting force vector;

[0011] The multi-axis attitude optimization control unit is configured to: for each candidate attitude, extract representative wall thickness values ​​from the profile section CAD contour data according to the cutting direction determined by the candidate attitude, call the cutting force vector analysis unit to calculate the cutting force vector under the candidate attitude, and then calculate the profile deformation evaluation index under different feasible attitude combinations based on the profile section moment of inertia information, the cutting force vector and the geometric constraints of the cutting task, and select the attitude with the smallest deformation evaluation index as the optimal cutting attitude;

[0012] The collaborative controller is configured to: acquire the optimal cutting posture parameters, calculate the cutting direction vector of the saw blade cutting into the profile section, and transmit the vector to the section topology analysis unit;

[0013] The cross-sectional topology analysis unit is configured to: scan the profile cross-sectional contour along the cutting path according to the cutting direction vector, identify and generate a cross-sectional feature sequence containing feature point types and corresponding cutting depths;

[0014] The feed rate adaptive control unit is configured to: establish a forward view window based on the cross-sectional feature sequence, and pre-adjust the feed rate of the multi-axis sawing module before the saw blade reaches the feature point according to the type of cross-sectional feature to be encountered in the forward view window.

[0015] As a preferred technical solution of the present invention, the feature point types in the cross-sectional feature sequence include at least the outer contour entry point, wall cutting point, wall contact point, wall thickness change point, and outer contour departure point; the cross-sectional topology analysis unit establishes an index structure from the cutting depth value to the feature point for the feed rate adaptive control unit to query.

[0016] As a preferred embodiment of the present invention, the feed rate adaptive control unit includes a forward view window management submodule and a rate decision submodule; the forward view window management submodule dynamically calculates the forward view distance based on the current feed rate, and defines the range of the forward view window with the current cut depth position as the starting point; the rate decision submodule executes the following control strategy based on the feature point type within the forward view window: when a wall cut-through point is detected, the feed rate is reduced to a first preset proportion of the standard rate; when a wall contact point is detected, the feed rate is reduced to a second preset proportion of the standard rate.

[0017] As a preferred embodiment of the present invention, the first preset ratio is 40% to 60% of the standard rate, and the second preset ratio is 25% to 40% of the standard rate; when the rate decision submodule detects that multiple feature points exist at the same time, it selects the strategy with the largest deceleration amplitude to execute.

[0018] As a preferred technical solution of the present invention, the method for the multi-axis attitude optimization control unit to generate a feasible attitude set is as follows: a combination traversal is performed on the discrete value sequences of the saw blade tilt angle, the saw blade rotation angle, and the gripper rotation angle of the multi-degree-of-freedom gripper module to filter out combinations that satisfy the geometric constraints of the cutting task; the method for the multi-axis attitude optimization control unit to extract a representative wall thickness for each candidate attitude is as follows: the direction vector of the saw blade cutting into the profile section is determined according to the angle parameters of the candidate attitude; a ray-polygon intersection operation is performed on the CAD contour of the profile section along this direction to obtain the wall thickness values ​​of all solid wall segments through which the ray passes; the maximum value among them is taken as the representative wall thickness; the cutting force vector analysis unit... The tangential cutting force is calculated based on the cutting resistance coefficient of the profile material, the representative wall thickness, the feed per tooth, and the number of saw teeth participating in the cutting simultaneously. The radial cutting force is converted from the tangential cutting force based on the radial force ratio coefficient. The cutting force vector is synthesized in the saw blade coordinate system and transformed from the saw blade coordinate system to the profile section coordinate system through a coordinate rotation transformation matrix corresponding to the saw blade tilt angle, saw blade rotation angle, and gripper rotation angle. The deformation evaluation index is the ratio of the absolute value of the projection component of the cutting force vector in the strong axis direction of the profile section after transformation to the profile section coordinate system to the minimum moment of inertia of the profile section, wherein the strong axis direction is the direction parallel to the principal axis of inertia corresponding to the maximum moment of inertia of the profile section.

[0019] As a preferred embodiment of the present invention, the multi-axis attitude optimization control unit is further configured to perform clamping point torque verification: calculate the lever arm length of the cutting force relative to the nearest clamp in the adaptive clamping module and the resulting bending moment value, and eliminate attitude schemes whose bending moment value exceeds a preset threshold.

[0020] As a preferred embodiment of the present invention, the cooperative controller is provided with timing logic that enables the feed rate adaptive control unit to remain in a waiting state until the multi-axis attitude optimization control unit determines the optimal attitude and the cross-section topology analysis unit completes the cross-section feature sequence update, at which point the feed rate adaptive control unit is activated to intervene in the control.

[0021] As a preferred embodiment of the present invention, the feed rate adaptive control unit further includes a smooth transition submodule, which uses an S-shaped speed curve or a trapezoidal acceleration / deceleration curve to smoothly transition the current rate to the target rate.

[0022] As a preferred embodiment of the present invention, the intelligent control layer further includes a load feedback correction submodule. This module estimates the theoretical current value based on the local wall thickness information and cutting process parameters corresponding to the current cutting depth in the cross-sectional feature sequence and the motor load model. It then compares the real-time current value collected by the current sensor with the theoretical current value. When the measured current exceeds the theoretical threshold, the feed rate is corrected and slowed down.

[0023] As a preferred embodiment of the present invention, the mechanical execution layer further includes a loading chamber module and a unloading conveying module; the multi-axis sawing module is equipped with a seven-axis linkage sawing station, including a saw blade feed axis, a saw blade lifting axis, a saw blade lateral movement axis, a saw blade tilting axis, a saw blade rotation axis, and an auxiliary axis.

[0024] Compared with the prior art, the present invention has the following beneficial effects:

[0025] 1. This invention achieves predictive control of the feed rate through cross-sectional topology analysis and a forward-looking window mechanism. The system can decelerate in advance before the saw blade cuts through the wall or contacts a new wall surface, eliminating the lag of traditional feedback control, effectively suppressing saw blade oscillation and impact caused by sudden changes in cutting force, significantly reducing burrs and tear marks on the cut surface, and improving the processing quality of complex cavity profiles.

[0026] 2. This invention utilizes multi-axis attitude optimization control, combined with cutting force vector analysis, to automatically plan the optimal cutting attitude based on the moment of inertia characteristics of the profile cross-section. This mechanism ensures that the principal direction of the cutting force is as far away as possible from the strong axis direction of the profile cross-section, thereby minimizing the force component acting on the cross-section along the strong axis direction and causing bending deformation around the weak axis. This minimizes elastic deformation and vibration during the cutting process, thereby improving the dimensional and angular accuracy of composite angle cutting of irregular profiles.

[0027] 3. This invention establishes an information closed loop for attitude planning and feed control. The collaborative controller dynamically updates the cross-sectional feature sequence based on the optimized attitude parameters, ensuring that the feed control strategy always precisely matches the actual physical cutting direction. This achieves intelligent collaboration throughout the entire process, from process planning to execution, guaranteeing the robustness and consistency of the machining process. Attached Figure Description

[0028] Figure 1 This is a schematic diagram of the overall architecture of the automatic cutting and processing system provided in an embodiment of the present invention;

[0029] Figure 2 This is an overall structural layout diagram of the mechanical execution layer in an embodiment of the present invention.

[0030] Explanation of reference numerals in the attached diagram: 100, Mechanical execution layer; 110, Loading magazine module; 120, Multi-DOF gripper module; 130, Adaptive clamping module; 140, Multi-axis sawing module; 150, Unloading and conveying module; 200, Sensing and detection layer; 210, Encoder group; 220, Current sensor; 300, Intelligent control layer; 310, Cross-section topology analysis unit; 320, Feed rate adaptive control unit; 330, Cutting force vector analysis unit; 340, Multi-axis attitude optimization control unit; 350, Cooperative controller. Detailed Implementation

[0031] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0032] Please see Figure 1-2This invention provides an automated cutting and processing system based on aluminum alloy production. The system comprises three layers: a mechanical execution layer 100, a sensing and detection layer 200, and an intelligent control layer 300. The mechanical execution layer 100 carries out the physical processing actions and includes a loading magazine module 110, a multi-degree-of-freedom gripper module 120, an adaptive clamping module 130, a multi-axis sawing module 140, and a material unloading and conveying module 150. The sensing and detection layer 200 is responsible for collecting real-time system operating status data and providing information input for control decisions. It includes an encoder group 210, a current sensor 220, and a profile detection sensor. The intelligent control layer 300 serves as the control center of the entire system and includes a cross-section topology analysis unit 310, a feed rate adaptive control unit 320, a cutting force vector analysis unit 330, a multi-axis attitude optimization control unit 340, and a collaborative controller 350, responsible for implementing the intelligent cutting control strategy involved in this invention. The improvements of this invention focus on the software level of the intelligent control layer 300, without involving structural changes to the mechanical execution layer 100 device. Deployment can be achieved by upgrading the existing device control software.

[0033] Regarding the mechanical execution layer 100, the loading magazine module 110 is used to store the long strip aluminum alloy profiles to be processed. This module has an automatic one-to-one feeding function, and its standard configuration can accommodate seven profiles with a maximum profile width of 250mm. After receiving the feeding command from the control system, the loading magazine module 110 pushes the profiles one by one to the gripping station to complete the loading action. The multi-degree-of-freedom gripper module 120 is used to realize the handling and posture adjustment of the profiles during processing. This module has feed motion along the length of the profile, vertical lifting motion, lateral traversing motion, and rotational motion around the profile axis. The rotational motion angle range is 0° to 350°. All of the above movements are driven by servo motors, which have high positioning accuracy and smooth movement. The adaptive clamping module 130 includes multiple sets of vertical clamps. Each vertical clamp can automatically adjust its position in the horizontal direction. Its automatic horizontal setting range is 5mm to 70mm to adapt to the clamping requirements of profiles of different widths and ensure that the profiles are stably and reliably positioned in the processing area. The multi-axis sawing module 140 is the core execution component for implementing the cutting action. This module is equipped with a seven-axis linkage sawing station. The specific composition of the seven-axis linkage is as follows: five axes are the inherent motion axes of the sawing station itself, namely the saw blade feed axis (controlling the feed motion of the saw blade along the cutting direction), the saw blade lifting axis (controlling the position of the saw blade in the vertical direction), the saw blade lateral axis (controlling the horizontal lateral position of the saw blade), the saw blade tilt axis (controlling the tilt angle of the saw blade around the horizontal axis), and the saw blade rotation axis (controlling the rotation angle of the saw blade around the vertical axis); the other two axes are linkage axes that coordinate the positioning of the workpiece, namely the gripper longitudinal feed axis (controlling the feeding position of the multi-degree-of-freedom gripper module 120 along the length direction of the profile) and the gripper rotation axis (controlling the rotation angle of the multi-degree-of-freedom gripper module 120 around the axis of the profile). The two aforementioned linkage axes, while physically belonging to the multi-degree-of-freedom gripper module 120, are integrated into the CNC interpolation controller of the sawing station for unified management. Together with the five axes of the sawing station itself, they form a seven-axis linkage system, thereby achieving synchronous and coordinated interpolation motion between the saw blade movement and the workpiece posture adjustment, supporting the execution of complex angle cutting tasks. The saw blade diameter is 550mm (a 650mm saw blade is also optional), and the sawing motor power is 7kW, supporting cutting at any angle within the range of 45° to 135°. The unloading and conveying module 150 is used to gently transfer the cut profile finished product from the processing area to the unloading area. During the transfer, the force applied to the finished product is controlled within a range that will not cause surface damage. The unloading action is completed by the cooperation of the unloading chamber and the unloading conveyor belt.

[0034] The sensor detection layer 200 provides the necessary data foundation for the intelligent control layer 300. Encoder groups 210 are installed on each motion axis of the system to acquire real-time position information of each axis. For the feed control scheme of this invention, the cutting depth position information fed back by the encoder on the saw blade feed axis is particularly critical, as its data accuracy directly affects the accuracy of the feedforward control. Current sensors 220 are installed in the drive circuit of the sawing motor to collect the operating current value of the sawing motor in real time. This current value is physically positively correlated with the cutting load borne by the saw blade and can be used as a characteristic signal of the cutting load for load feedback correction. Profile detection sensors are installed at the system feed end to detect whether the profile has reached the designated position and to confirm the reference position of the profile, providing a spatial reference for subsequent gripping and positioning.

[0035] The intelligent control layer 300 is the concentrated embodiment of the technical contributions of this invention. The following sections will provide a detailed description of each functional unit in the intelligent control layer 300 and their collaborative mechanisms.

[0036] Before detailing each unit, it is necessary to first clarify the method of taking the wall thickness parameter at different working stages in this invention and the handling of its data dependency, as this relates to the execution timing design between attitude optimization and cross-sectional topology analysis. The intelligent control layer 300 of this invention involves two scenarios where wall thickness information is used at different stages: firstly, the attitude optimization stage before the cutting task is executed, where wall thickness information is needed for cutting force estimation and deformation evaluation index calculation; secondly, the feed control stage during the cutting task execution, where wall thickness information is needed for feedforward rate adjustment and motor current prediction. The accuracy requirements and data sources for wall thickness information differ fundamentally between these two stages.

[0037] During the attitude optimization phase, the system needs to calculate the cutting force vector and evaluate the deformation index for multiple candidate attitudes to select the optimal attitude. At this time, the optimal attitude has not yet been determined, and the cross-section topology analysis unit 310 has not yet generated the final cross-section feature sequence (because this sequence depends on the finally determined cutting direction). Therefore, the wall thickness parameters used for attitude optimization cannot be obtained from the cross-section feature sequence. This invention solves this problem by using the following method: For each candidate attitude, the multi-axis attitude optimization control unit 340 independently calculates the cutting direction vector corresponding to the attitude based on the angle parameters of the candidate attitude, and then performs a ray-polygon intersection operation on the profile cross-section CAD contour data along this direction, extracting a representative wall thickness value from the intersection result (taking the maximum wall thickness value among all solid wall segments traversed along this direction). This lightweight geometric operation only requires the coordinate data of the cross-section CAD contour and the direction vector of the candidate attitude, and does not depend on the generation result of the cross-section feature sequence, thereby eliminating the circular dependency of "attitude optimization depends on wall thickness → wall thickness depends on cross-section feature sequence → cross-section feature sequence depends on optimal attitude". Since the purpose of the attitude optimization stage is to make relative comparisons between different candidate solutions rather than to accurately calculate the absolute cutting force value, it is reasonable and conservative in engineering to use the maximum wall thickness as the representative wall thickness for cutting force estimation.

[0038] During the feed control phase, the optimal attitude has been determined and the cross-sectional feature sequence has been generated. At this point, the wall thickness parameters used by the feed rate adaptive control unit 320 and the load feedback correction submodule are the precise local wall thickness values ​​recorded point by point according to the cutting depth in the cross-sectional feature sequence. .

[0039] The cross-section topology analysis unit 310 is used to preprocess and analyze the cross-sectional structure of the aluminum profile to be cut before cutting. This unit receives the profile's cross-sectional contour data as input, and after a series of analysis processes, outputs a sequence of cross-sectional feature data reflecting the distribution of cross-sectional geometric features along the cutting direction. The cross-sectional contour data comes from the host computer's CAD data interface or eluCad software, and the imported cross-sectional contour information includes the coordinate data of the profile's outer contour boundary and the coordinate data of all internal cavity contour boundaries. After acquiring the cross-sectional contour data, the cross-section topology analysis unit 310 determines the cutting plane according to the angle parameters specified by the current cutting task. These angle parameters include two degrees of freedom: tilt angle and rotation angle. These two angle parameters uniquely determine the spatial relationship between the plane where the saw blade is located and the profile axis, thereby determining the projection trajectory of the cross-sectional cutting line traversed by the saw blade when cutting into the profile on the profile cross-section.

[0040] After determining the cross-sectional cutting line, the cross-sectional topology analysis unit 310 scans and analyzes the cutting line along the saw blade's entry direction, sequentially identifying and marking various cross-sectional feature points encountered along the way. These feature points include outer contour entry points, wall penetration points, wall contact points, wall thickness abrupt change points, and outer contour exit points. The outer contour entry point is the position where the saw blade first contacts the outer surface of the profile, marking the start of the cutting process; the wall penetration point is the position where the saw blade completely cuts through a wall and is about to enter the cavity region, where cutting resistance disappears instantaneously, making it one of the key points for control strategy attention; the wall contact point is the position where the saw blade contacts the wall again after passing through the cavity, where impact loads are generated; the wall thickness abrupt change point is the position where the wall thickness changes significantly, including transitions from thin to thick and from thick to thin walls; and the outer contour exit point is the position where the saw blade completely leaves the profile, marking the end of the cutting process.

[0041] All the identified cross-sectional feature points are arranged in ascending order of their coordinate values ​​along the cutting depth direction, forming a cross-sectional feature sequence. Each element in the sequence records three pieces of information: the feature point type identifier, the corresponding cutting depth value, and the local wall thickness value at that feature point. The local wall thickness value is only valid for wall-related feature types; for feature points in cavity regions, this value is recorded as zero. The generated cross-sectional feature sequence is stored in the buffer area of ​​the control system, and a fast index structure from the cutting depth value to the feature point is established accordingly. This index structure is designed so that the feed rate adaptive control unit 320 can perform efficient queries based on the cutting depth value during operation without traversing the entire sequence, thus meeting the real-time control requirements for query speed.

[0042] It is important to note that the cross-sectional feature sequence output by the cross-sectional topology analysis unit 310 is not a static and unchanging result; it is closely related to the direction in which the saw blade cuts into the profile cross-section. When the cutting angle or the spatial attitude of the profile changes, the direction in which the saw blade cuts into the cross-section also changes. Different cutting directions mean that the saw blade will traverse the walls and cavities in the cross-section along different paths, resulting in substantial differences in the identification results and arrangement order of feature points. Therefore, before each cutting task, the cross-sectional topology analysis unit 310 needs to re-execute the analysis process based on the attitude parameters finally determined by the multi-axis attitude optimization control unit 340, from the cutting plane definition step to the feature sequence generation and storage step, generating a cross-sectional feature sequence that perfectly matches the current cutting direction.

[0043] The feed rate adaptive control unit 320 is the core functional module for improving the quality of the cut surface. During the sawing process, it dynamically adjusts the feed rate of the saw blade based on the cross-sectional feature sequence data and the real-time position information of the saw blade, realizing feedforward-based cutting parameter optimization control based on prior knowledge of the cross-sectional topology. This unit consists of a forward window management submodule, a rate decision submodule, a smooth transition submodule, and a load feedback correction submodule. These submodules work together to complete the real-time adjustment of the feed rate.

[0044] The forward view window management submodule maintains a "forward view window" mechanism. Its core idea is to predict the cross-sectional features the saw blade will encounter within a short distance of its future travel, allowing the control system to react in advance rather than passively adjusting after actually encountering the feature point. The forward view window is defined as an interval extending a preset distance forward from the current saw blade cutting depth; this preset distance is called the forward view distance. The forward view distance is not a fixed constant but is dynamically calculated based on the current saw blade feed rate and the system control cycle. Its calculation principle ensures that, under the current feed rate conditions, the system has sufficient time to complete the entire response process from detecting the feature point to the rate adjustment taking effect. Specifically, the forward view distance... The calculation formula is:

[0045]

[0046] in, This represents the current feed rate of the saw blade. The minimum target rate corresponding to the forward window (take the lowest value among all rate adjustment strategies, i.e. the second preset ratio corresponding to the standard rate). This is the maximum deceleration allowed by the system, a value determined by the servo drive performance and mechanical inertia of the multi-axis sawing module 140. This refers to the number of control cycles from when the controller detects a feature point to when the deceleration command actually takes effect. The duration of a single control cycle. This is a safety margin constant. The physical meaning of the above formula is as follows: the first term is the kinematic braking distance required for the saw blade to decelerate from the current rate to the minimum target rate; the second term is the distance the saw blade continues to travel during the system detection and response delay; and the third term is a fixed safety margin to prevent parameter drift. When the feed rate is high, the look-ahead distance is increased accordingly to allow for a more ample deceleration distance; when the feed rate decreases, the look-ahead distance can be appropriately shortened.

[0047] Within each control cycle, the forward window management submodule first obtains the current cutting depth position value of the saw blade from the encoder group 210, then determines the coverage range of the forward window on the cutting depth coordinate based on the position value and the current forward distance, and then uses the fast index structure of the cross-sectional feature sequence to query all cross-sectional feature points falling within the range, and passes the query results to the rate decision submodule for processing.

[0048] The rate decision submodule makes feed rate decisions based on feature point information within the forward view window. This submodule maintains a set of control rules that map different feature types detected in the forward view window to corresponding rate adjustment strategies. When no feature points fall within the current view window, the current cutting section is determined to be a uniform wall region, and the current rate is maintained or restored to the standard cutting rate corresponding to the material and wall thickness. When a wall penetration point is detected within the current view window, it indicates that the saw blade is about to completely penetrate the current wall and enter the cavity. At this point, a pre-deceleration operation is performed, reducing the rate to a first preset proportion of the standard rate to avoid saw blade swaying and oscillation caused by the sudden disappearance of cutting force at the moment of wall penetration. When a wall contact point is detected within the current view window, it indicates that the saw blade is about to re-contact the wall after passing through the cavity. At this point, further deceleration is performed, reducing the rate to a second preset proportion of the standard rate to buffer the impact load generated upon re-contact with the wall, reducing cut surface tearing and burrs. When a sudden change in wall thickness is detected, and the wall thickness changes from thin to thick, a pre-deceleration operation is performed. The deceleration magnitude is determined based on the ratio of the wall thickness change before and after the change; the greater the wall thickness change, the greater the deceleration magnitude. When the wall thickness changes from thick to thin, the current speed is maintained, and preparations are made to moderately increase the speed after passing the feature point. When an outer contour departure point is detected, an end-of-line deceleration strategy is executed to avoid end impact and burrs caused by the sudden disappearance of cutting force when the saw blade cuts the profile. In actual operation, multiple feature points of different types may exist simultaneously in the forward view window. In this case, the speed decision submodule adopts the most conservative strategy principle, that is, it selects the strategy with the largest deceleration magnitude from all triggered strategies and executes it to ensure that the cut surface quality is not compromised under any circumstances.

[0049] The aforementioned first and second preset ratios are parameter values ​​preset by the system and can be adjusted according to the characteristics of the profile material, the condition of the saw blade, and the processing quality requirements. In one specific embodiment, the first preset ratio is set to 40% to 60% of the standard rate, and the second preset ratio is set to 25% to 40% of the standard rate. However, this range is not a limitation on the scope of protection of this invention, and those skilled in the art can make reasonable adjustments according to actual working conditions.

[0050] The function of the smooth transition submodule is to eliminate mechanical shock during rate switching. When the rate decision submodule outputs a new target rate value, the smooth transition submodule does not directly send the target value to the servo drive. Instead, it smoothly transitions the current rate to the target rate according to a pre-tuned acceleration / deceleration curve. This acceleration / deceleration curve preferably adopts an S-shaped velocity curve. The characteristic of an S-shaped curve is that its acceleration change is continuous and its jerk is limited, thus effectively suppressing mechanical vibration caused by sudden rate changes. The parameters of the S-shaped curve include maximum acceleration and maximum jerk, which are pre-tuned based on the mechanical inertia characteristics of the multi-axis sawing module 140 and the dynamic response capability of the servo system. In another embodiment, the smooth transition submodule can also use a trapezoidal acceleration / deceleration curve or other velocity transition curves with smooth characteristics, as long as the mechanical shock suppression requirements are met.

[0051] The load feedback correction submodule is retained in the system as a supplement to feedforward control. The core control strategy of this invention is feedforward control based on prior knowledge of the cross-sectional topology. However, considering the potential differences between the actual profile product and the CAD model—for example, the profile may contain additional reinforcing ribs not shown in the CAD data, or there may be deviations in wall thickness due to manufacturing tolerances—pure feedforward control may not cover all operating conditions. Therefore, the load feedback correction submodule continuously monitors the real-time current value of the sawing motor and compares it with the theoretical current value estimated based on the wall thickness information at the current cutting depth. Theoretical current value. The prediction method is based on a physical mapping model of cutting force and motor load, and its calculation formula is as follows:

[0052]

[0053] in, The no-load current of the sawing motor can be obtained by actual measurement when the motor is running idle; This is the cutting drag coefficient corresponding to the profile material, in units of... ; The cutting depth in the cross-sectional feature sequence The local wall thickness value corresponding to the location, in mm; This refers to the feed per tooth, in mm. The current feed rate, The saw blade rotation speed, This refers to the number of teeth on the saw blade. This represents the average number of saw teeth that participate in the cutting process simultaneously. This refers to the saw blade diameter, in mm. This represents the torque constant of the sawing motor, expressed in Nm / A. This represents the mechanical efficiency of the transmission chain from the saw blade spindle to the motor. The derivation logic of the above formula is: tangential cutting force (Unit: N) Through the saw blade radius (Unit: mm) is converted into cutting torque acting on the spindle. (Unit is) Then, after conversion of transmission efficiency and motor torque constant, the increase in motor current required to maintain this cutting torque is obtained. Because The unit is and The unit is (Right now There is a 1000-fold difference in dimensions between the two, therefore the formula introduces... coefficients will Unified conversion to This ensures that the calculated theoretical current value is consistent with the measured current value in terms of dimensions. It should be noted that the wall thickness in this formula... It is the precise local wall thickness value obtained by querying point by point according to the cutting depth from the generated cross-sectional feature sequence during the cutting execution phase, and compared with the representative wall thickness used in the attitude optimization phase. They differ in data sources and levels of precision.

[0054] When the measured current exceeds a certain threshold of the theoretical current value, it is determined that there are unmodeled structural features or parameter deviations in the cross-sectional feature sequence. The load feedback correction submodule then immediately corrects and slows down the feed rate until the measured current falls back to a reasonable range, and then resumes the rate determined by the feedforward control. This mechanism makes feedforward control and feedback control complementary, maintaining the timely response advantage of feedforward control while enhancing the system's robustness to uncertain operating conditions through the feedback loop.

[0055] The cutting force vector analysis unit 330 is used to analyze the direction and relative magnitude of the cutting force exerted by the saw blade on the profile during the cutting process, based on the parameters of the cutting task and the profile cross-sectional information. This unit receives wall thickness inputs at different accuracy levels at different working stages: during the attitude optimization stage, it receives representative wall thicknesses extracted by the multi-axis attitude optimization control unit 340 for each candidate attitude. During the feed control phase, the point-by-point wall thickness in the cross-sectional feature sequence can be received. The cutting force estimation model for this unit is as follows:

[0056] First, calculate the tangential cutting force. For the working condition of a circular saw blade cutting aluminum alloy profiles, the tangential cutting force is... The calculation formula is:

[0057]

[0058] in, This refers to the cutting drag coefficient corresponding to the aluminum alloy grade of the profile, in units of... This coefficient reflects the material's ability to resist cutting deformation. For the commonly used 6063-T5 aluminum alloy, The typical value range is 300 to 500. ; The wall thickness is the input value, which, during the attitude optimization stage, is the representative wall thickness extracted by the multi-axis attitude optimization control unit 340 from the profile section CAD contour data for the current candidate attitude. (Unit: mm) During the feed control phase, this refers to the local wall thickness at the current cutting depth in the cross-sectional feature sequence. (Unit: mm); The feed per tooth is calculated using the following formula: ,in This represents the current feed rate (mm / s). The saw blade rotation speed (r / s) This refers to the number of teeth on the saw blade. This is the average number of saw teeth participating in the cutting simultaneously, and this value is related to the saw blade diameter and the current cutting depth.

[0059] Next, calculate the radial cutting force. Radial cutting force The tangential cutting force is expressed by the radial force ratio coefficient. The conversion yields:

[0060]

[0061] in, This is the radial force ratio coefficient, used for sawing aluminum alloys. The typical value range is 0.3 to 0.5, and the specific value is related to the saw blade tooth shape, rake angle and wear condition.

[0062] In the saw blade's own coordinate system, the tangential cutting force Radial cutting force acting along the tangent of the sawtooth. The force acts radially toward the center of the saw blade, and the combined force is the cutting force vector in the saw blade coordinate system. .

[0063] Subsequently, coordinate transformation is performed. To clarify the coordinate transformation relationship, this invention defines the following two coordinate systems and their reference pose relationships. The profile section coordinate system has its origin at the centroid of the profile section, with the X-axis along the horizontal direction of the profile section, the Y-axis along the vertical direction of the profile section, and the Z-axis along the axis of the profile. These three coordinates form a right-handed coordinate system; this coordinate system is fixed to the profile and reflects the inherent geometric characteristics of the profile section. The saw blade coordinate system has its origin at the center of the saw blade. The axis is along the radial direction of the saw blade (pointing towards the cutting direction). The axis is along the saw blade axis (perpendicular to the saw blade plane). The axes are along the tangent direction of the saw teeth, and the three coordinates form a right-handed coordinate system. The reference pose is defined as: the saw blade tilt angle. °, saw blade rotation angle °, gripper rotation angle At °, the saw blade coordinate system and the profile cross-section coordinate system are parallel and in the same direction on their respective axes, meaning the saw blade is in a vertical orientation and the profile is in an upright position. In this reference orientation, the saw blade cuts perpendicularly into the profile cross-section from directly above along the negative Y-axis. Any actual cutting posture can be considered as the result of applying three rotational transformations sequentially to this reference orientation. This is combined with the current saw blade tilt angle. and rotation angle and the rotation angle of the gripper The cutting force vector is transformed from the saw blade coordinate system to the profile section coordinate system through three rotational transformations:

[0064]

[0065] in, To combine the rotation transformation matrices, the correspondence between each rotation component and the physical motion axis is as follows: Rotation angle around the Y-axis of the profile section coordinate system The rotation matrix (corresponding to the movement of the saw blade's tilt axis, the saw blade tilting around the transverse axis) angle), The rotation angle around the Z-axis of the profile section coordinate system The rotation matrix (corresponding to the motion of the saw blade's rotation axis, the saw blade rotating around the profile axis) angle), Rotation angle around the X-axis of the profile section coordinate system The rotation matrix (corresponding to the inverse transformation of the gripper's rotation axis, the profile is rotated by the gripper) After the angle is adjusted, the cutting force vector is applied. The reverse rotation restores the profile from the rotated coordinate system to the profile's inherent cross-section coordinate system. The order of the three rotations is the inverse transformation of the saw blade tilting, the saw blade rotation, and finally the gripper rotation. This order reflects the elimination of the attitude deflection introduced by each motion axis during the transformation of the force vector from the saw blade coordinate system to the profile cross-section coordinate system.

[0066] This estimation process does not pursue absolute accuracy; its purpose is to provide a reliable quantitative basis for relative comparisons between different attitude schemes. The cutting force vector analysis unit 330 will analyze the cutting force vector, which includes information on direction and relative magnitude. The output is sent to the multi-axis attitude optimization control unit 340 as one of the input conditions for attitude optimization calculation.

[0067] The multi-axis attitude optimization control unit 340 automatically selects the optimal solution that minimizes the deformation of the profile during cutting from a variety of feasible attitude combinations that meet the cutting geometry requirements, based on the principle of mechanical optimization, before the actual execution of the cutting task. Taking a 45° bevel cutting task as an example, the multi-axis attitude combination to achieve this cutting geometry is not unique. The saw blade can be tilted at 45° while the profile remains upright, or the saw blade can be kept vertical while the profile is rotated 45° by the multi-degree-of-freedom gripper module 120, or the saw blade can be tilted at one angle while the gripper rotates at another angle, as long as the combination satisfies the geometric constraints. Although these different attitude combinations are geometrically equivalent, the profile cross-section is often asymmetrical, and its bending stiffness varies significantly in different directions. Different attitude combinations will cause the cutting force to act in different directions on the profile cross-section, thus causing different degrees of elastic deformation.

[0068] The working process of the multi-axis attitude optimization control unit 340 is as follows. First, a set of feasible attitudes is generated. Specifically, within the angle range allowed by the multi-axis sawing module 140, discrete value sequences are set for the saw blade tilt angle, saw blade rotation angle, and gripper rotation angle, respectively. These three sequences are combined and traversed, and then each combination is checked to see if it meets the geometric constraints of the current cutting task. Combinations that pass the check are included in the set of feasible attitudes. Each feasible attitude is represented by a vector containing three parameters: saw blade tilt angle, saw blade rotation angle, and gripper rotation angle.

[0069] After generating the feasible attitude set, the system extracts the section moment of inertia information from the profile section data. The section moment of inertia reflects the section's ability to resist bending deformation. For any section shape, there are two mutually orthogonal principal axes of inertia, corresponding to the maximum moment of inertia of the section. and minimum moment of inertia . The corresponding axial direction is the strong axis direction of the cross section, and the cross section has the strongest resistance to bending along this direction; The corresponding axis is the weak axis direction of the cross-section, along which the cross-section's resistance to bending is weakest. It is important to note that, according to the Euler-Bernoulli beam bending theory in mechanics of materials, when a cross-section undergoes bending deformation about a principal axis of inertia, the component of the external force causing this bending is perpendicular to that bending axis—that is, along the other principal axis of inertia. Therefore, the component of the external force causing the profile to bend around the weak axis ( The force component of bending is not along the weak axis, but rather along the strong axis. For asymmetrical multi-cavity cross-sections commonly found in architectural aluminum profiles, and The differences between them are often quite significant.

[0070] For each candidate scheme in the feasible attitude set, the multi-axis attitude optimization control unit 340 calculates an evaluation index to characterize the degree of profile deformation under that scheme. The calculation process is as follows: First, based on the three parameters of saw blade tilt angle, saw blade rotation angle, and gripper rotation angle in the candidate scheme, the direction vector of the saw blade cutting into the profile section under that candidate attitude is determined; then, a ray-polygon intersection operation is performed on the CAD contour data of the profile section along this direction, and the maximum wall thickness value is taken as the representative wall thickness from the intersection results of the ray passing through all solid wall segments. ; then, The angle parameters of the candidate scheme are input into the cutting force vector analysis unit 330 to calculate the cutting force vector under the candidate attitude. The aforementioned ray-polygon intersection operation is performed directly based on the coordinate data of the cross-section CAD profile and the direction vector of the candidate attitude, without relying on the cross-section feature sequence output by the cross-section topology analysis unit 310. This ensures that the attitude optimization process can be executed independently before the cross-section feature sequence is generated, eliminating the data circular dependency between the two. The engineering basis for selecting the maximum wall thickness as the representative wall thickness is that the maximum wall thickness corresponds to the peak cutting force during the cutting process. Using the peak force for deformation evaluation ensures that the comparison results between the candidate schemes are conservative and reliable.

[0071] Obtaining the cutting force vector Then, the cutting force is decomposed into components along the strong axis of the cross section. and components along the weak axis ,in The calculation method is the absolute value of the projection component of the cutting force vector onto the unit vector in the strong axis direction, that is:

[0072]

[0073] Then follow Calculate deformation evaluation index The mechanical meaning of this index is based on the Euler-Bernoulli beam theory: the force component along the strong axis. The load perpendicular to the weak axis is the effective load that causes the profile to bend around the weak axis, while Let be the moment of inertia of the cross section about the weak axis, representing the cross section's ability to resist the bending; the ratio of the two is... It is proportional to the maximum bending deflection of the profile about the weak axis. The smaller the value, the smaller the force component applied to the cross-section along the strong axis under this attitude scheme, the smaller the elastic bending deformation of the profile around the weak axis, and the better the preservation of cutting accuracy. From the physical meaning of optimization, minimizing... This means that the principal direction of the cutting force should be as far away from the strong axis direction of the cross section as possible (i.e., as close as possible to the weak axis direction), so that the effective force component projected onto the strong axis direction that causes bending around the weak axis is minimized.

[0074] In addition to deformation analysis in the cross-sectional direction, the multi-axis attitude optimization control unit 340 also performs clamping point torque verification. For cases where the cutting position is far from the vertical clamp in the adaptive clamping module 130, even if the deformation index in the cross-sectional direction is excellent, the bending moment generated after the cutting force acts through a long lever arm may still cause non-negligible deformation or vibration at the profile end. Therefore, for each candidate scheme, the lever arm length of the cutting force relative to the vertical clamp closest to the cutting position, and the resulting bending moment value, are calculated. When the bending moment value exceeds a preset threshold, the attitude scheme is marked as requiring clamping position adjustment or not recommended. In one embodiment, when the system determines that the clamping torque of a certain attitude scheme exceeds the standard, it can automatically issue a clamping position adjustment command to the adaptive clamping module 130, moving an additional vertical clamp closer to the cutting position, thereby allowing the attitude scheme to meet the torque verification requirements again. In another embodiment, the system directly excludes the scheme from the feasible attitude set and selects a suboptimal scheme instead.

[0075] Finally, the system selects deformation evaluation indicators from feasible solutions that have passed the clamping point torque verification. The option with the smallest value is selected as the optimal posture scheme, and its parameter vector is output to the multi-axis motion controller to control the saw blade tilt axis, saw blade rotation axis, and gripper rotation axis to move to the corresponding positions. At the same time, the optimal posture parameters are also output to the cooperative controller 350, which triggers the update process of the cross-sectional feature sequence.

[0076] The cooperative controller 350 is a key scheduling module connecting the multi-axis attitude optimization control unit 340 and the feed rate adaptive control unit 320. Its existence enables the attitude optimization and feed control functions to no longer be isolated from each other, but to form a cooperative working relationship with a closed information loop. The working mechanism of the cooperative controller 350 is reflected in the following aspects.

[0077] Upon receiving a new cutting task, the cooperative controller 350 first invokes the multi-axis attitude optimization control unit 340 to initiate attitude optimization calculations. During this stage, the multi-axis attitude optimization control unit 340 independently executes an evaluation process for each candidate attitude: "direction calculation → ray intersection extraction to extract representative wall thickness → cutting force estimation → deformation index calculation." This process is directly based on the profile cross-section CAD contour data and candidate attitude parameters, without relying on the cross-sectional feature sequence output by the cross-section topology analysis unit 310. Before the optimal attitude is determined, the feed rate adaptive control unit 320 remains in a waiting state, and the cross-section topology analysis unit 310 does not perform a complete cross-sectional feature sequence generation operation. The design logic of this timing constraint is that the generation of the cross-sectional feature sequence depends on the direction in which the saw blade cuts into the profile cross-section, and this direction is determined by the final selected cutting attitude. If the cross-sectional feature sequence is generated before the attitude is determined, there will be a mismatch between the sequence and the actual cutting direction, and the accuracy of the feedforward control cannot be guaranteed.

[0078] After the multi-axis attitude optimization control unit 340 completes the calculation and outputs the optimal attitude parameters, the cooperative controller 350 receives this parameter information and calculates the cutting direction vector when the saw blade actually cuts into the profile section based on the combination of the gripper rotation angle, saw blade tilt angle, and rotation angle. This cutting direction vector directly determines the direction of the section cutting line and the identification results of feature points in the section topology analysis unit 310. The cooperative controller 350 then transmits this cutting direction vector to the section topology analysis unit 310, triggering it to execute the entire analysis process from defining the cutting plane to storing the feature sequence and establishing the index. Since this analysis is performed under the finally determined cutting direction, the generated section feature sequence perfectly matches the actual cutting direction.

[0079] After the cross-sectional feature sequence is updated, the cooperative controller 350 sends an activation signal to the feed rate adaptive control unit 320, which switches from a waiting state to an active state, officially intervening in the real-time control of the feed rate during the cutting process. At this point, the complete information chain from attitude optimization to cross-sectional analysis to feed control is established, and the system can begin executing the cutting action.

[0080] Under the unified scheduling of the collaborative controller 350, the complete workflow of the system for each cutting task is as follows: receiving cutting task information, including the target cutting angle, profile identification, cross-section CAD data, etc.; calling the multi-axis attitude optimization control unit 340 to independently extract the representative wall thickness for each candidate attitude based on the cross-section CAD contour data and perform operations such as feasible attitude generation, calling the cutting force vector analysis unit 330 to perform cutting force vector analysis, cross-section moment of inertia calculation, deformation evaluation index calculation, and clamping moment verification, and output the optimal attitude plan; the collaborative controller 350 calculates the cutting-in direction vector according to the optimal attitude parameters and transmits it to the cross-section topology analysis unit 310; the cross-section topology analysis unit 310 generates a cross-section feature sequence based on the cutting-in direction and establishes an index; the multi-axis motion controller drives each axis to move to the position specified by the optimal attitude; the collaborative controller 350 activates the feed rate adaptive control unit 320; performs the cutting action, during which the feed rate adaptive control unit 320 continuously performs feedforward control according to the cross-section feature sequence and the real-time position of the saw blade, and the load feedback correction sub-module intervenes for compensation when necessary; after cutting, the finished product is transferred to the discharge area by the blanking conveying module 150.

[0081] To further illustrate the technical solution of the present invention, the working process of the system is described below through a specific application scenario.

[0082] This scenario is for 45° bevel cutting of a multi-chamber aluminum alloy crossbeam profile for building curtain walls. The cross-section of the profile is an asymmetric "day" shape structure, including two main chambers and several heat insulation chambers inside. The total width of the profile is 180 mm, the main wall thickness is 2 mm, and the inner rib wall thickness is 1.4 mm. Due to the asymmetry of the profile cross-section, the moment of inertia values corresponding to the strong axis direction and the weak axis direction are significantly different.

[0083] After the system receives this cutting task from the eluCad software, the collaborative controller 350 calls the multi-axis attitude optimization control unit 340 to start working. The multi-axis attitude optimization control unit 340 generates a set of feasible attitudes according to the geometric requirements of 45° bevel cutting, which at least includes the following three types of representative schemes: Scheme A is that the saw blade is tilted at 45° and the gripper does not rotate, Scheme B is that the saw blade remains vertically tilted at 0° and the gripper rotates at 45°, and Scheme C is that the saw blade is tilted at 22.5° and the gripper rotates at 22.5° at the same time. For each candidate scheme, the multi-axis attitude optimization control unit 340 first calculates the cutting-in direction vector according to the angle parameters of the scheme, and then performs ray-polygon intersection operation on the cross-section CAD contour along this direction to extract the representative wall thickness (In this example, due to different cutting-in directions for each scheme, the wall segments traversed by the rays are different, and the Although they differ, they all take the maximum wall thickness value in their respective directions. The cutting force vector analysis unit 330 calculates the components of the cutting force vector in the profile section coordinate system for each scheme based on the representative wall thickness and angle parameters of each scheme. For scheme A, the calculated angle between its cutting force direction and the strong axis direction of the profile section is small, which means that the projection component of the cutting force in the strong axis direction is relatively small. A relatively large component will cause significant bending deformation of the profile around its weak axis, according to the deformation evaluation index. The value is relatively large. For schemes B and C, the angle between the cutting force direction and the weak axis direction of the section is small, and the projection component in the strong axis direction is relatively large. A smaller load results in a reduced effective load causing bending deformation around the weak axis. The value is correspondingly low. After completing the evaluation index calculation and clamping torque verification of all candidate solutions, the system selects... The solution with the smallest value that passes the torque check is selected as the optimal attitude. Assume that solution B is selected as the optimal solution in this scenario.

[0084] The collaborative controller 350 receives the attitude parameters of scheme B and calculates the actual direction of the saw blade cutting into the profile section based on the condition that the gripper rotates 45°. Since the profile rotates 45° under the action of the multi-degree-of-freedom gripper module 120, although the saw blade maintains a vertical orientation and cuts downwards, the cutting direction relative to the profile section is deflected by 45° compared to when the profile is not rotated. The collaborative controller 350 transmits this deflected cutting direction vector to the section topology analysis unit 310. The section topology analysis unit 310 generates a complete section feature sequence based on this finally determined cutting direction and identifies section feature points. Due to the change in cutting direction, the order in which the saw blade passes through the walls and cavities is completely different from the sequence when the profile is upright. The newly generated section feature sequence accurately reflects the actual encounter order of the saw blade along each wall and cavity when the profile is rotated 45°.

[0085] After each axis moves to its designated position, the co-controller 350 activates the feed rate adaptive control unit 320, and cutting begins. During the cutting process, the saw blade cuts into the outer contour of the profile, and the forward view window management submodule continuously tracks the cutting depth of the saw blade. When the first wall cutting point of the inner rib wall appears in the forward view window, the rate decision submodule issues a deceleration command in advance, and the smooth transition submodule reduces the feed rate from the standard value to the rate corresponding to the first preset ratio according to an S-shaped curve. After the saw blade cuts through the wall and enters the inner cavity area, if a wall contact point of the next layer of wall is detected in the forward view window during the brief cavity travel, the rate decision submodule further reduces the rate to the second preset ratio. The saw blade contacts and cuts into the next layer of wall at a lower rate. When the cutting process stabilizes and no new feature points appear in the forward view window, the rate gradually returns to the standard cutting rate. Throughout the cutting process, the saw blade undergoes a similar pre-deceleration and recovery process each time it passes through an alternating section between the wall and the cavity. Changes in cutting force are buffered in advance, preventing damage to the cut surface quality from impact loads. Finally, when the saw blade approaches the point where it leaves the outer contour of the profile, the system executes an end-of-line deceleration strategy to complete the final cutting action at a low speed.

[0086] After cutting, the unloading conveyor module 150 gently transfers the finished product to the unloading area. Throughout the process, the profile cut surface is smooth and clean, with very few burrs and tear marks. For profiles that require subsequent anodizing or powder coating surface treatment, the grinding process can be eliminated or significantly reduced, improving production efficiency and reducing processing costs.

[0087] Based on the descriptions of the above embodiments, it can be understood that the present invention solves the problem of feed rate control lag when cutting complex cross-section profiles by the coordinated cooperation of the cross-section topology analysis unit 310 and the feed rate adaptive control unit 320, solves the problem of deformation caused by the mismatch between the cutting force direction and the profile stiffness characteristics by the coordinated cooperation of the cutting force vector analysis unit 330 and the multi-axis attitude optimization control unit 340 based on the cross-section mechanical properties, and integrates the various units into an organic whole by the collaborative controller 350, realizing intelligent optimization of the entire process from cutting planning to cutting execution.

[0088] It should be noted that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An automated cutting and processing system based on aluminum alloy production, characterized in that, include: The mechanical execution layer includes an adaptive clamping module for clamping and positioning the profile, a multi-degree-of-freedom gripper module for moving the profile and adjusting its spatial orientation, and a multi-axis sawing module for performing cutting actions. The sensing and detection layer includes an encoder group for acquiring position information of the motion axis and a current sensor for acquiring motor load signals; The intelligent control layer includes a cross-section topology analysis unit, a cutting force vector analysis unit, a multi-axis attitude optimization control unit, a feed rate adaptive control unit, and a cooperative controller; The cutting force vector analysis unit is configured to: calculate the direction and relative magnitude of the cutting force in the profile section coordinate system based on the rotation direction, cutting direction and current posture parameters of the saw blade, combined with the material properties and cutting process parameters of the profile, and output the cutting force vector; The multi-axis attitude optimization control unit is configured to: for each candidate attitude, extract representative wall thickness values ​​from the profile section CAD contour data according to the cutting direction determined by the candidate attitude, call the cutting force vector analysis unit to calculate the cutting force vector under the candidate attitude, and then calculate the profile deformation evaluation index under different feasible attitude combinations based on the profile section moment of inertia information, the cutting force vector and the geometric constraints of the cutting task, and select the attitude with the smallest deformation evaluation index as the optimal cutting attitude; The collaborative controller is configured to: acquire the optimal cutting posture parameters, calculate the cutting direction vector of the saw blade cutting into the profile section, and transmit the vector to the section topology analysis unit; The cross-sectional topology analysis unit is configured to: scan the profile cross-sectional contour along the cutting path according to the cutting direction vector, identify and generate a cross-sectional feature sequence containing feature point types and corresponding cutting depths; The feed rate adaptive control unit is configured to: establish a forward view window based on the cross-sectional feature sequence, and pre-adjust the feed rate of the multi-axis sawing module before the saw blade reaches the feature point according to the type of cross-sectional feature to be encountered in the forward view window.

2. The automatic cutting and processing system based on aluminum alloy production according to claim 1, characterized in that, The feature point types in the cross-sectional feature sequence include at least the outer contour entry point, wall cutting point, wall contact point, wall thickness change point, and outer contour departure point; the cross-sectional topology analysis unit establishes an index structure from the cutting depth value to the feature points for the feed rate adaptive control unit to query.

3. The automatic cutting and processing system based on aluminum alloy production according to claim 2, characterized in that, The feed rate adaptive control unit includes a forward window management submodule and a rate decision submodule; the forward window management submodule dynamically calculates the forward distance according to the current feed rate and defines the range of the forward window with the current cut-in depth position as the starting point. The rate decision submodule maps the rate adjustment strategy according to the feature point type in the forward window.

4. The automatic cutting and processing system based on aluminum alloy production according to claim 3, characterized in that, The control strategy of the rate decision submodule includes: when a wall cutting point is detected, reducing the feed rate to a first preset ratio of the standard rate; when a wall contact point is detected, reducing the feed rate to a second preset ratio of the standard rate; the first preset ratio is greater than the second preset ratio.

5. The automatic cutting and processing system based on aluminum alloy production according to claim 1, characterized in that, The method for generating a feasible attitude set by the multi-axis attitude optimization control unit is as follows: the discrete value sequences of the saw blade tilt angle, the saw blade rotation angle and the gripper rotation angle of the multi-degree-of-freedom gripper module are combined and traversed to select the combination that satisfies the geometric constraints of the cutting task.

6. The automatic cutting and processing system based on aluminum alloy production according to claim 5, characterized in that, The method by which the multi-axis attitude optimization control unit extracts representative wall thickness for each candidate attitude is as follows: The direction vector of the saw blade cutting into the profile section is determined based on the angle parameters of the candidate posture. Ray-polygon intersection operation is performed on the CAD profile of the profile section along this direction to obtain the wall thickness value of the ray passing through all solid wall segments. The maximum value among them is taken as the representative wall thickness. The method for calculating the cutting force vector by the cutting force vector analysis unit includes: calculating the tangential cutting force based on the cutting drag coefficient of the profile material, the representative wall thickness, the feed per tooth, and the number of saw teeth participating in the cutting simultaneously; converting the tangential cutting force into the radial cutting force based on the radial force ratio coefficient; synthesizing the tangential cutting force and the radial cutting force in the saw blade coordinate system to obtain the cutting force vector; and transforming the cutting force vector from the saw blade coordinate system to the profile section coordinate system through a coordinate rotation transformation matrix corresponding to the saw blade tilt angle, saw blade rotation angle, and gripper rotation angle. The deformation evaluation index is the ratio of the absolute value of the projection component of the cutting force vector in the strong axis direction of the profile section after transformation to the profile section coordinate system to the minimum moment of inertia of the profile section, wherein the strong axis direction is the direction parallel to the principal axis of inertia corresponding to the maximum moment of inertia of the profile section.

7. The automatic cutting and processing system based on aluminum alloy production according to claim 6, characterized in that, The multi-axis attitude optimization control unit is also configured to perform clamping point torque verification: calculate the cutting force relative to the lever arm length of the nearest clamp in the adaptive clamping module and the resulting bending moment value, and eliminate attitude schemes with bending moment values ​​exceeding a preset threshold.

8. The automatic cutting and processing system based on aluminum alloy production according to claim 1, characterized in that, The cooperative controller is equipped with timing logic that allows the feed rate adaptive control unit to remain in a waiting state until the multi-axis attitude optimization control unit determines the optimal attitude and the cross-section topology analysis unit completes the cross-section feature sequence update, at which point the feed rate adaptive control unit is activated to intervene in the control.

9. The automatic cutting and processing system based on aluminum alloy production according to claim 1, characterized in that, The intelligent control layer also includes a load feedback correction submodule. The load feedback correction submodule estimates the theoretical current value based on the local wall thickness information and cutting process parameters corresponding to the current cutting depth in the cross-sectional feature sequence and the motor load model. It then compares the real-time current value collected by the current sensor with the theoretical current value. When the measured current exceeds the theoretical threshold, the feed rate is corrected and slowed down.

10. The automatic cutting and processing system based on aluminum alloy production according to claim 1, characterized in that, The feed rate adaptive control unit also includes a smooth transition submodule, which uses an S-shaped speed curve or a trapezoidal acceleration / deceleration curve to smoothly transition the current rate to the target rate.

Images